2005 Proceedings - ASNR

2005 

PROCEEDINGS 

43 rd Annual Meeting 

May 23-27, 2005 

Metro Toronto Convention Centre 

Toronto, Ontario, Canada 

American Society 

of Neuroradiology 

2005 Proceedings includes 

Interactive CD-ROM

ASNR 

American Society of Neuroradiology 

Dear Colleagues, 

The ASNR 43nd Annual Meeting and NER Symposium 2005 breaks new ground in attendance and programming. 

Patricia A. Hudgins, M.D., President-Elect/Program Chair, has collected topical and important educational and 

scientific sessions. Thje meeting covers the clinical and technological developments important for our practice and 

the Symposium provides an excellent review of head and neck imaging in cancer. 

The Annual Meeting has over 40 Focus Sessions developed in cooperation with the American Society of Functional 

Neuroradiology (ASFNR), American Society of Head and Neck Radiology (ASHNR), American Society of Pediatric 

Neuroradiology (ASPNR), American Society of Interventional and Therapeutic Neuroradiology (ASITN) and the 

American Society of Spine Radiology (ASSR), covering a wide range of topics of interest for both the sub-specialist 

in neuroradiology as well as the general neuroradiologist. I wish to extend a special thanks to the following 

Co-Chairs for their efforts in organizing the programming for the following specialty areas: 

American Society of Functional Neuroradiology (ASFNR) ..............................................Andrei I Holodny, M.D. 

American Society of Head and Neck Radiology (ASHNR)..............................................Roy A. Holliday, M.D. 

American Society of Interventional and Therapeutic Neuroradiology (ASITN) ............John D. Barr, M.D. 

American Society of Pediatric Neuroradiology (ASPNR) .................................................Marvin D. Nelson, M.D. 

American Society of Spine Radiology (ASSR) ...................................................................Gordon K. Sze, M.D. 

Other program highlights include the Advanced Imaging Seminars providing an in-depth look at advanced imaging 

techniques (perfusion, diffusion, spectroscopy, parallel and high field imaging), a Research Grant Writing Seminar, 

Expanded ELC Workshop and Lectures, National Library of Medicine (NLM) workshops, the successful breakfast, 

lunch and reception, How-To Sessions and new this year the ASNR Business Center sessions. Don’t miss the 

Wednesday morning Special Session: Surviving Change: Neuroradiology Practice. 

The annual meeting provides a unique opportunity to gain a better understanding of how the ASNR functions 

to assist the practice of neuroradiology during a time of rapid change. 

The meeting also provides excellent opportunities to renew old friendships and make new ones, join the Welcome 

Reception with Technical Exhibitors on Monday evening and see a portrait of neuroradiology as seen by the 

performing artists of the Second City Toronto Comedy Troupe. 

See Toronto, join a dinner tour and sample the optional tour programs during your stay. 

On behalf of the entire Executive Committee, welcome to Toronto, Ontario, Canada for the NER Foundation 

Symposium 2005 and ASNR 43rd Annual Meeting where advanced technology, clinical imaging and interventional 

neuroradiological excellence come together. 

Sincerely, 

Victor M. Haughton, M.D. 

ASNR President 

TORONTO 

ASNR 2005

How to use your Proceedings CD-Rom 

With free Acrobat Reader ® software, you can view and print Adobe PDF files. If your computer doesn’t already 

have Acrobat installed you can simply download either a Macintosh ® or Windows ® version from this CD-Rom. 

With this CD-Rom you can search the data in many different ways as shown in the examples below: 

User can link from the Table of Contents to pages throughout the 

CD-Rom as shown above by simply clicking on the links in the file. 

Convenient Bookmarks make it easy to jump to specific sessions 

throughout the week’s program. 

You can easily move to a page of your choice. 

To access the author search section simply click the Index of 

Program Participants link on Table of Contents page. 

The Thumbnail palette displays a small image of each page in the 

proceedings for quick viewing.

How to use your Proceedings CD-Rom (continued) 

With this CD-Rom you can also search by author: 

CD-Rom basic training 

Blue Text: Indicates links — click on blue text to go 

directly to the page indicated. 

View Bookmarks (on toolbar): To have 

bookmarks appear to the left of the document, 

click on the title you want in order to go directly 

to that page in the article. 

Go Back (on toolbar): To go directly back to 

the last page you were viewing click on the 

arrow in the toolbar. Or, from the Document 

pull-down menu select Go Back anytime you 

want to go directly back to the last page you 

were viewing. 

Hand Tool (on toolbar): To move the page on 

the screen, click on the hand button on the 

toolbar. Your cursor will change to a hand. Place 

the cursor on the page, hold down the left 

mouse button and move the mouse. 

Zoom in Tool (on toolbar): To enlarge a 

specific piece of text, click on the magnifying 

glass button on the toolbar. Your cursor will 

change to a magnifying glass. Place the cursor on 

the text and click. To return to normal view, click 

on the page buttons on the toolbar. 

User can link from the Index of Program Participants to 

pages throughout the CD-Rom as shown above by simply 

clicking on the links in the file. 

Previous Page (on toolbar): To go to the 

previous page. 

Next Page (on toolbar): To go to the 

next page. 

Page Buttons (on toolbar): Click on these buttons 

to change the size of the page on your screen. From left 

to right: 

Actual Size: Shows the page in 

its actual size. 

Fit in Window: Shows the entire page 

on your screen. 

Fit Width: Shows the page filling the width 

of your screen. 

Find Tool (on toolbar): Click on the binoculars 

button on the toolbar to find a specific word or 

phrase in the booklet. 

Help: If you have questions about using Acrobat 

Reader, click on the Help pull-down menu and select 

“Reader Guide.”

TABLE OF CONTENTS 

Table of Contents Addendum NeuroNews 

II..................................................................................2004 - 2005 ASNR Executive Committee 

II – III............................................2004 - 2005 ASNR Annual Meeting Arrangements Committees 

IV ....................................................................................................................ASNR/AJNR Staff 

V – VI ......................................................................ASNR 43rd Annual Meeting Invited Speakers 

VII....................................................................................................................All About Toronto 

VII ..........................................................................................................Walking Map of Toronto 

VIII – XI ..................................................................Metro Toronto Convention Centre Floor Plans 

XIII – XIV ............................................Scientific Posters and Scientific Exhibits (Printed and Electronic) 

XV – XVII ........................................................................................................Technical Exhibits 

XVIII – XX ....................................................................................................General Information 

XXI..........................................................................................Social Program and Optional Tours 

XXII ..................................................................................................................Guest Hospitality 

XXIII ........................................................................................................Future ASNR Meetings 

XXIII ....................................................................................ASNR Past Presidents and Founders 

XXIV – XXV ........................................................................................Past Meetings of the ASNR 

XXVI ..............................................................................2005 ASNR Gold Medal Award Recipient 

XXVII ............................................................................Past ASNR Gold Medal Award Recipients 

XXVII ............................................................................................2005 ASNR Honorary Member 

XXVIII ............................................................................Past ASNR Honorary Member Recipients 

XXVIII ....................................................................2004 ASNR Outstanding Presentation Awards 

XXIX ............................................................................2004 Regional Society Best Paper Awards 

XXIX ........2005 NER Foundation Award for Outstanding Contributions in Research Award Recipient 

XXIX ..........Past NER Foundation Award for Outstanding Contributions in Research Award Recipient 

XXX......................2005-2006 Berlex/NER Foundation Fellowship in Basic Science Research Award 

XXX – XXXI ......Past Berlex/NER Foundation Fellowship in Basic Science Research Award Recipients 

XXXII ..................................NER Foundation Scholar Award in Neuroradiology Research Recipients 

XXXII............................................Past NER Foundation Scholar Award in Neuroradiology Research 

XXXII ........2005 NER Foundation/Boston Scientific Fellowship in Cerebrovascular Disease Research 

XXXIII ......................NER Foundation Outcomes Research Grant Related to Neuroradiology Imaging 

XXXIII ............................................................................2005 Cornelius G. Dyke Memorial Award 

XXXIII – XXXIV ........................................Past ASNR Cornelius G. Dyke Memorial Award Recipients 

XXXV – XXXVI................................................ASNR 43rd Annual Meeting Accreditation Statement, 

Educational Objectives, and Target Audience 

XXXVI – XXXVIII ............................Electronic Learning Center (ELC) 2005 Workshops and Lectures 

XXXIX ....................................................................National Library of Medicine (NLM) Workshops 

XXXIX ......................................................................................................ASNR Business Center 

XXXIX..........................................................................................Research Grant Writing Seminar 

XL ....................................................................................................................How-To Sessions 

XLI – XLVI ........................................................................................Scientific Program Overview 

Page 1..............................................................................................................Monday Sessions 

Page 67............................................................................................................Tuesday Sessions 

Page 139 ....................................................................................................Wednesday Sessions 

Page 183 ........................................................................................................Thursday Sessions 

Page 255 ............................................................................................................Friday Sessions 

Page 285 ........................................................................................................Scientific Posters 

Page 385 ..........................................................................................Scientific Exhibits (Printed) 

Page 427................................................................................Electronic Scientific Exhibits (eSE) 

Page 447........................................................................................Index of Program Participants 

I 

Toronto, Canada Canada



II 

COMMITTEES 

2004-2005 ASNR Executive Committee 

Victor M. Haughton, MD 

President 

Patricia A. Hudgins, MD 

President-Elect/Program Chair 

Robert I. Grossman, MD 

Vice President 

M. Judith Donovan Post, MD, FACR 

Treasurer 

John R. Hesselink, MD, FACR 

Secretary 

James M. Provenzale, MD 

Member-at-Large 

Howard A. Rowley, MD 

Research Committee Chair 

Andrei I. Holodny, MD 

Functional Representative 

Vijay M. Rao, MD 

Head and Neck Representative 

Gary R. Duckwiler, MD 

Interventional Representative 

Marvin D. Nelson, MD 

Pediatric Representative 

Gordon K. Sze, MD 

Spine Representative 

Charles M. Strother, MD 

First Past-President 

Patrick A. Turski, MD 

Second Past-President 

William P. Dillon, MD 

Third Past-President 

J. Arliss Pollock, MD 

Clinical Practice Committee Chair 

James G. Smirniotopoulos, MD 

Education Committee Chair 

Charlotte H. Rydberg, MD 

Rules Committee Chair 

Mauricio Castillo, MD 

Publications Committee Chair 

Andrew W. Litt, MD 

AMA Delegate 

Kelly K. Koeller, MD 

ACR Representative 

Robert R. Lukin, MD 

ABR Representative 

Glenn S. Forbes, MD, FACR 

ABR Representative 

William G. Bradley, MD 

ACR Neuro/MRI Commission Liaison 

James B. Gantenberg, CHE 

ASNR Executive Director/CEO, 

Commercial Relations Committee Chair 

2004-2005 Meeting Arrangements Committees 

Scientific Program Committee 

Scientific Program Committee 



Roy A. Holliday, MD 

Head and Neck Representative 

John D. Barr, MD 

Interventional Representative 


Functional Representative 

Marvin D. Nelson, MD 

Pediatric Representative 


Spine Representative 

Michael Brant-Zawadzki, MD 

Hugh D. Curtin, MD 

Nancy I. Fischbein, MD 

Lawrence E. Ginsberg, MD 

R. Gilberto Gonzalez, MD, PhD 



John R. Hesselink, MD, FACR 

Emanuel Kanal, MD 

Edward E. Kassel, MD 

Gregory L. Katzman, MD 

Walter Kucharczyk, MD, FRCPC 

Laurie A. Loevner, MD 

Carolyn Cidis Meltzer, MD 

David J. Mikulis, MD 

Suresh K. Mukherji, MD 

Jay J. Pillai, MD 


Robert M. Quencer, MD 

Timothy P.L. Roberts, PhD 


David Saloner, PhD 

Hervey D. Segall, MD 

James G. Smirniotopoulos, MD 

A. Gregory Sorensen, MD 

Richard H. Wiggins, III, MD 

Robert D. Zimmerman, MD

COMMITTEES 

2004-2005 Meeting Arrangements Committees (continued) 

Audio Visual Committee 

Erin M. Simon, MD 

Chair 

Edward A. Michals, MD 

Consultant 

Robert M. Barr, MD 

Alexander B. Baxter, MD 

Jacqueline A. Bello, MD 

Richard M. Berger, MD 

Brian C. Bowen, PhD, MD 

Orest B. Boyko, MD, PhD 

Phillip W. Chao, MD 

Dale A. Charletta, MD 

Dianna Chooljian, MD 

Raquel Del Carpio-O’Donovan, MD 

Colin P. Derdeyn, MD 

David B. Granato, MD 

Rose Marie Holt, MD, PhD 

Anil Khosla, MD 


Robert A. Koenigsberg, DO, FACR 

Warren W. Lam, MD 

John I. Lane, MD 

Joel R. Meyer, MD 

Pratik Mukherjee, MD, PhD 

Walter L. Olsen, MD, PhD 

Tina Young Poussaint, MD 

Joshua S. Shimony, MD, PhD 

Harish N. Shownkeen, MD 

Murray A. Solomon, MD 

Jeffrey A. Stone, MD 

Vance E. Watson, MD 


Electronic Learning Center (ELC) Committee 


Chair 


Vice Chair, Syllabus Editor 


Vice Chair, Moderator and Technical Coordination 

Hervey D. Segall, MD 

Chair Emeritus 

Scientific Exhibits Committee 

Linda A. Heier, MD 

Chair 

Stephen Gebarski, MD 

Consultant 


Nancy J. Fischbein, MD 

Jill V. Hunter, MD 

Prabhakar P. Kesava, MD 

Leena M. Ketonen, MD 

Bernadette L. Koch, MD 


Jonathan S. Lewin, MD 

Gary M. Nesbit, MD 

Susan Palasis, MD 

Gregory W. Petermann, MD 



Evelyn M. Sklar, MD 

David M. Yousem, MD, MBA 

Technical Exhibits Committee 

Lawrence N. Tanenbaum, MD 

Chair 

Alan L. Williams, MD, FACR, MBA 

Vice Chair 

Walter S. Bartynski, MD 


Ray A. Brinker, MD 

Thomas S. Dina, MD 

Glen K. Geremia, MD 


John M. Mathis, MD, MSc 

John Perl, II, MD 

C. Douglas Phillips, MD 

Donald S. Willig, MD 

Gregg H. Zoarski, MD 

Local Arrangements Committee 


President 



Susan I. Blaser, MD 

Sylvester Chuang, MD 

Lyne N. DeTilly, MD 

Allan J. Fox, MD, FRCPC, FACR 


Walter Kucharczyk, MD FRCPC 


Karel G. TerBrugge, MD 

Robert Willinsky, MD 

III 




IV 

ASNR Staff 

ASNR STAFF 

American Society of Neuroradiology 

Headquarters Office 

2210 Midwest Road, Suite 207 

Oak Brook, IL 60523-8205 

Telephone: 630-574-0220 

Fax: 630-574-0661/630-574-1740 

Internet: http://www.asnr.org 

James B. Gantenberg, CHE 

Executive Director/CEO 

Ext. 224, jgantenberg@asnr.org 

Arthur An 

Educational Technologist 

Ext. 243, aan@asnr.org 

Angelo Artemakis 

Director of Communications and Media Management 

Ext. 227, aartemakis@asnr.org 

Ken Cammarata 

Director of Specialty Societies and Member Services 

Ext. 226, kcammarata@asnr.org 

Tina Cheng, CPA 

Director of Finance and Information Systems 

Ext. 225, tcheng@asnr.org 

Maurine Dennis, MPH, MBA 

Director of Clinical Practice Services 

Ext. 233, mdennis@asnr.org 

AJNR Staff 

AJNR STAFF 

American Journal of Neuroradiology 




Telephone: 630-574-1487 

Fax: 630-574-0326 

Internet: http://www.ajnr.org 

Bridget Donohue, MA 

Managing Editor 

Ext. 2, bdonohue@asnr.org 

Mary Harder 

Editorial Assistant 

Ext. 3, mharder@asnr.org 

Derrick J. Leaks, MA 

Editorial Assistant 

Ext. 4, dleaks@asnr.org 

Hayley Whittington 

Publications Assistant 

Ext. 1, hwhittington@asnr.org 

Pat Galle-Bingham 

Staff Accountant 

Ext. 222, pgalle-bingham@asnr.org 

Valerie Geisendorfer, CMP 

Meetings Coordinator 

Ext. 231, vgeisendorfer@asnr.org 

Bonnie Mack 

Membership/Society Coordinator 

Ext. 234, bmack@asnr.org 

Karen Mansfield 

Office Manager/Executive Assistant 

Ext. 221, kmansfield@asnr.org 

Catherine Neis, MTA 

Senior Logistics and Programming Specialist 

Ext. 232, cneis@asnr.org 

Barbara Schell, CAE 

Manager of Scientific Meetings 

Ext. 228, bschell@asnr.org 

Debbie Seitz 

CPC Administrative Assistant 

Ext. 236, dseitz@asnr.org 

Lora J. Tannehill, CMP 

Director of Scientific Meetings 

Ext. 229, ltannehill@asnr.org 

Additional Email Addresses 

Executive 

executive@asnr.org 

Clinical Practice 

cpc@asnr.org 

Communications 

communications@asnr.org 

Finance 

finance@asnr.org 

Meetings 

meetings@asnr.org 

Membership 

membership@asnr.org

INVITED SPEAKERS 

43rd Annual Meeting Invited Speakers 

James J. Abrahams, MD 

Yale University School of Medicine 

Walter H. Backes, PhD 

Maastricht University Hospital, The Netherlands 

Roland Bammer, PhD 

Stanford University 


Mid South Imaging and Therapeutics, 

Memphis, TN 


The Hospital for Sick Children 

Toronto, ON, Canada 

Brian C. Bowen, MD, PhD 

University of Miami School of Medicine 

William G. Bradley, MD, PhD, FACR 

University of California San Diego Healthcare 

Allan L. Brook, MD 

Montefiore Medical Center, Bronx, NY 

R. Nick Bryan, MD, PhD 

University of Pennsylvania Health System 

Patricia E. Burrows, MD 

Children’s Hospital, Boston, MA 


University of North Carolina School of Medicine 

John J. Connors, III, MD 

Miami Cardiac and Vascular Institute 

H. Christian Davidson, MD 

University of Utah 


Washington University, School of Medicine, 

St. Louis, MO 


University of California Medical Center, 

San Francisco 


University of California Los Angeles, 

School of Medicine 

William K. Erly, MD 

University of Arizona 

Aaron S. Field, MD, PhD 

University of Wisconsin Hospital and Clinics 

Alisa D. Gean, MD 

University of California Medical Center, 

San Francisco 

Bassem A. Georgy, MD 

Valley Radiology Consultants, Escondido, CA 

Floyd H. Gilles, MD 

Children’s Hospital, Los Angeles 

J. Rob Gimbel, MD, FACC 

East Tennessee Heart Consultants, Knoxville, TN 


MD Anderson Cancer Center, Houston, TX 

Charles M. Glasier, MD 

Arkansas Children’s Hospital, Little Rock, AK 

Xavier Golay, PhD 

National Neuroscience Institute, Singapore 


Massachusetts General Hospital 

P. Ellen Grant, MD 



New York University Medical Center 

Patrick Gullane, MB, FRCSC, FACS 

University Health Network, Toronto, ON, Canada 

H. Ric Harnsberger, MD 


Anton N. Hasso, MD FACR 

University of California Irvine Medical Center 

Gary Lee Hedlund, DO 

Primary Children’s Medical Center, Salt Lake City, UT 

Eric C. Holland, MD 

Memorial Sloan-Kettering Cancer Center 



Jill V. Hunter, MD 

Texas Children’s Hospital, Houston, TX 

Blake A. Johnson, MD 

Center for Diagnostic Imaging, St. Louis, MO 

Michele H. Johnson, MD 


Blaise V. Jones, MD 

Children’s Hospital Medical Center, Cinncinnati, OH 


University of Pittsburgh Medical Center 

Sasan Karimi, MD 



Mount Sinai Hospital, Toronto, ON, Canada 

V 




VI 

INVITED SPEAKERS 

43rd Annual Meeting Invited Speakers (continued) 

Spyros Kollias, MD 

Institute of Neuroradiology, University Hospital, 

Zurich, Switzerland 

Andrew Ku, MD 

Allegheny General Hospital, Pittsburgh, PA 

Christiane Kuhl, PhD 

University of Bonn, Germany 

Alan Leviton, MD, MS 

The Children’s Hospital, Boston, MA 

Lisa H. Lowe, MD 

Children’s Mercy Hospital, Kansas City, MO 

Helmi Lutsep, MD 

Oregon Health and Sciences University 

Joseph A. Maldjian, MD 

Wake Forest University School of Medicine 

Michael P. Marks, MD 

Stanford University Medical Center 


Oregon Health and Science University 

Brian O’Sullivan, MD, FRCPI 

Princess Margaret Hospital, 

Toronto, ON, Canada 

Ashok Panigrahy, MD 

The Children’s Hospital, Los Angeles, CA 

Jeffrey R. Petrella, MD 

Duke University Medical Center 


Medical College of Georgia 


Radiological Associates, Sacramento, CA 



Klaas P. Pruessmann, PhD 

University and ETH Zurich, Switzerland 

Janet S. Rasey, PhD 

University of Washington 

Deborah L. Reede, MD 

The Long Island College Hospital 


University of Toronto, ON, Canada 

Caroline D. Robson, MB, ChB 

The Children’s Hospital, Boston, MA 

Brian D. Ross, MD 

University of Michigan 

Jeffrey S. Ross, MD 

Cleveland Clinic Foundation 



Eric J. Russell, MD, FACR 

Northwestern University, Chicago, IL 

James T. Rutka, MD, PhD 

The Hospital for Sick Children, Toronto, ON, Canada 

Yutakes Sato, MD 

University of Iowa Hospitals and Clinics, Iowa City, IA 

Pamela W. Schaefer, MD 


Kurt P. Schellhas, MD 

Center for Diagnostic Imaging, St. Louis, MO 

Eric D. Schwartz, MD 

Hospital of the University of Pennsylvania 

Deborah R. Shatzkes, MD 



The Children’s Hospital of Philadelphia 

James Smirniotopoulos, MD 

Uniformed Services University, Bethesda, MD 

Wendy R.K. Smoker, MD, FACR 

University of Iowa College of Medicine 

O. Carter Snead, III, MD 

The Hospital for Sick Children, Toronto, ON, Canada 



Philip E. Stieg, MD 

Cornell University 

Jeffrey L. Sunshine, MD, PhD 

University Hospitals of Cleveland 

Marshall S. Sussman, PhD 

University of Toronto, ON, Canada 


Yale University 

J. Michael Tyszka, PhD 

California Institute of Technology 

John L. Ulmer, MD 

Medical College of Wisconsin 

Alyssa T. Watanabe, MD 

Long Beach Memorial Medical Center 

Bruce M. Wenig, MD 

Beth Israel Medical Center, New York, NY 



Edward D. Wirth, III, MD, PhD 

Geron Corporation 

Wade H.M. Wong, DO 

University of California, San Diego 

Wayne F. Yakes, MD 

Vascular Malformation Center, Engelwood, CO 


The John Hopkins Medical Institute 

Robert A. Zimmerman, MD 

The Children’s Hospital of Philadelphia 

Robert D. Zimmerman, MD 

New York Presbyterian Hospital

SEATTLE, WA 

About Toronto, Ontario, Canada 

Toronto, the seat of government for the Province of 

Ontario, is also the cultural, entertainment, and 

financial capital of Canada. Toronto is a great city for 

visitors. It boasts a thriving arts community, reliable 

services and one of the safest urban environments in 

the world. 

The personality of Toronto, reflected in its slogan, 

“The World Within a City,” ® is represented by a 

mosaic of colorful cultures from around the world. 

Individuals often retain their cultural identities 

complete with traditions, languages and customs. 

You’ll see these personalities in the vibrant, quirky 

neighborhoods and experience it in the diversity of 

the arts, theatres, and dining. 

Walking Map of Toronto, Ontario, Canada 

Toronto offers ASNR 43rd Annual Meeting attendees 

an endless choice of activities. It boasts world-class 

museums such as the Royal Ontario Museum, 

fascinating and historic architecture including the CN 

Tower, Old and New City Hall, and lovely public spaces 

like Queen’s Park and Nathan Phillips Square. Music, 

film, theatre, dance and cultural events are 

everywhere. Adding to Toronto’s landscape are 150 

pieces of public art and monuments in addition to the 

over 2,000 moveable works of fine art on display in 

public buildings such as City Hall. Many of the city’s 

cultural attractions, ethnic and trendy neighborhoods, 

shopping areas such as the Eaton Centre, and 

Toronto’s waterfront are just a short walk or taxi ride 

away from the Fairmont Royal York Hotel, the ASNR 

headquarters hotel. 

ASNR would like to thank NORTHSTAR Travel Media, LLC for the use of their Toronto, ON, Canada downtown area map. 

VII 




VIII 

Level 100 

(As of 3/24/05) 

F 

101 

Headquarters 

Office 

ELEVATOR 

Theatre 

102 

Breakout Session 

Room 1 

FOYER 

EXIT STAIRS 

Directory 

Theatre 

Entrance 

Electronic103 

B 

Learning 

Center 

and NLM 

Workshops103 

A 

WR 

COATS 

ESCALATOR UP 

WR 

Front Street 

EXIT STAIRS 

Reception Hall 10 

A B C 

CME 

Pavilion 

WR 

y 

d 

a 

e 

R 

r 

e 

k 

a 

e 

p 

S 

m 

o 

o 

R 

Ov 

Sea 

Food 

E

EXIT STAIRS 

eception Hall 104 

B C D 

WR 

y 

d 

a 

e 

R 

r 

e 

k 

a 

e 

p 

S 

m 

o 

o 

R 

Overflow 

Seating for 

Food Service 

ELEVATOR 

ESCALATOR UP 

Directory Directory 


Room 2 

Constitution Hall 

FIRST AID 

MTCC 

108 

EXIT STAIRS EXIT STAIRS 

PRE-FUNCTION AREA 

105 106 107 

10 9 8 

7 

P 

M 

A 

R 

6 5 4 3 2 


Room 3 

D 

R 

A 

O 

B 

D 

Level 100 

(As of 3/24/05) 

WR WR 

KITCHEN 

Toronto, Canada Canada 

ELEVATOR 

IX



X 

Level 200 

(Entrance Level) 

(As of 3/24/05) 

Elevator 

Express Onsite Self 

Registration 

AJNR 

Meeting and 

Announcements 

M ESSA GE 

BOARD 

NER 

FOUNDATION 

BOARD 

Internal Street 

2 

201 A 201 C 201 E 

202 A 

202 C 

Reg 

Office 

203 A 203 C 

Elevator 

Committee Committee Committee Committee Committee Committee 

Room Room Room Food Service Room Room 

Research Session/ 

ASNR Business 

Center 

Committee 

Room 

Committee Committee Committee 

Room Food Service Room 

Committee 

Room 

Executive 

Committee 

Office 

201 B 201 D 201 F 

202 B 202 D 

203 B 

203 D 

204 

Summit 

Room 

Express Onsite Self 

Registration 

AJNR 

205 

Breakout 

Session 4 

M 2 

Meeting and 

Announcements 

M ESSA GE 

BOARD 

NER 

FOUNDATION 

BOARD 

M ESSAG E 

BOARD 

Directory 

Directory 

Restau rant 

Th eatre 

Entra nce 

Front Street

FRONT STREET 

EXIT STAIRS 

12'h drape 

8'h drape 

EXIT STAIRS 

EXIT STAIRS 

E-ACCESS/MESSAGING CENTER 

F.H.C. 

ESCALATOR UP/DOWN 

Entrance 

ESCA LATOR UP/DOWN 

F.H.C 

Entrance 

F.H.C. 

F.H.C. 

F.H.C. 

F.H.C. 

F.H.C. 

ELEV ATOR 

OPEN 

OPEN 

Technical 

8'h drape 

8'h drape 

F.H.C. 

100 

Micrus 

F.H.C. 

Scientific Exhibits 

101 

Bracco 

Diagnos. 

103 

4-D 

Neuro. 

200 

201 

401 301 

Kyphon 

Inc. 

GE 

Healthcare 

Philips 

Medical 

Systems 

Cardinal 

Health 

ELEVATORS 

104 

Shelley 

Med. Tech. 

106 

Cook Inc. 

107 

MicroVention 

110 

111 

407 406 307 306 207 

VSM 

Night- Lippincott 

Springer Integra 

MedTech 

Hawk Williams 

Spinal 

409 

Kopp 

Develop. 

411 

311 

8'h drape 

SE66 SE82 

SE67 SE83 

SE65 SE81 

SE68 SE84 

SE64 SE80 

SE69 SE85 

SE63 SE79 

SE70 SE86 

SE61 SE78 

SE71 SE87 

SE60 SE77 

SE72 SE88 

SE59 SE76 

SE73 SE89 

SE58 SE75 SE91 

SE90 

SE49 

SE50 

SE48 

SE51 

SE47 

SE52 

SE46 

SE53 

SE45 

SE54 

SE44 

SE55 

SE43 

SE56 

SE42 

SE57 

SE16 SE33 

SE17 SE34 

SE15 SE32 

SE18 SE35 

SE14 SE31 

SE19 SE36 

SE13 SE30 

SE20 SE37 

SE12 SE29 

SE21 SE38 

SE11 SE27 

SE22 SE39 

SE10 SE26 

SE23 SE40 

SE9 

SE25 

SE24 SE41 

SE1 

SE2 

SE3 

SE4 

SE5 

SE6 

Food Service 

TeraRecon 

Siemen 

Medical 

Hitachi 

Medical 

Vital 

Images, 

Inc. 

SE7 

SE8 

217 

Electronic 


(eSE) 

117 116 

Virtual 

Berlex 

Rad. 

Imaging 

118 

Elsevier 

120 

Amirsys 

Boston 

Scientific 

417 317 

Arthrocare 

Stryker 

419 

421 

P001 P028 P029 P056 P057 P084 P085 P112 P113 P141 P142 P169 P170 

P002 P027 P030 P055 P058 P083 P086 P111 P114 P140 P143 P168 P171 P184 



P005 P024 P033 P052 P061 P080 P089 

P108 P117 P137 P146 P165 P174 P181 




P009 P020 P037 P048 P065 P076 P093 P104 P121 P133 P150 P161 

P010 P019 P038 

P047 P066 P075 P094 P103 P123 P132 P151 P160 

P011 P018 P039 P046 P067 P074 P095 P102 P125 P131 P152 P159 

P012 P017 

P045 

P040 

P068 P073 P096 P101 P126 P130 P153 P158 

P013 P016 P041 P044 P069 P072 P097 P100 P127 P129 P154 P157 

P014 P015 P042 P043 P070 P071 P098 P099 

P128 P155 P156 

8'h drape 

T.F.H.C. 

222 123 

National Advanced 

Library Imaging 

322 223 

Advanced Invivo 

Bio 

324 225 

ASNR SNR 

T.F.H.C. 

8'h drape 

8'h drape 

TRUCK ENTRANCE 

RAMP 

Sc 

Posters 

8'h drape 

FREEMAN 

RETRIEVAL 

LEAD 

Food Service 

F.H.C. 

F.H.C. 

CONC ESSIO N SIGN 

F.H.C. 

F.H.C. 

F.H.C. 

F.H.C. 

CONC ESSIO N SIGN 

F.H.C. 

F.H.C. 

EXIT WALKWAY ABOVE 

CN TOWER BRIDGE 

EXHIBIT HALL “A” EXHIBIT HALL “B” 

Level 300 

(As of 3/24/05) 

XI 




XII 

Exhibit Hall B — Level 300 

(As of 3/24/05) 

SE1 

SE2 

SE3 

SE4 

SE5 

SE6 

SE7 

SE8 

P001 

P002 

P003 

P004 

P005 

P006 

P007 

P008 

P009 

P010 

P011 

P012 

P013 

P014 

P028 

P027 

P026 

P025 

P024 

P023 

P022 

P021 

P020 

P019 

P018 

P017 

P016 

P015 

Scientific Posters, Scientific Exhibits 

and Electronic Scientific Exhibits (eSE) 

SE16 

SE14 

SE16A 

SE15 

SE17 

SE18 

SE13 

SE20 

SE12 

SE21 

SE11 

SE22 

SE10 

SE9 

Note: A missing number indicates an abstract has been withdrawn. 

SE23 

SE24 

P029 

P030 

P031 

P032 

P033 

P034 

P035 

P036 

P037 

P038 

P039 

P040 

P041 

P042 

P056 

P055 

P054 

P053 

P052 

P051 

P050 

P049 

SE33 

SE34 

SE32 

SE35 

SE31 

SE36 

SE30 

SE37 

SE29 

SE38 

SE27 

SE39 

SE26 

SE40 

SE25 

P048 

P047 

P046 

P044 

P043 

SE41 

P057 

P058 

P059 

P060 

P061 

P062 

P063 

P064 

P065 

P066 

P067 

P068 

P069 

P070 

P084 

P083 

P082 

P081 

P080 

P079 

P078 

P077 

P076 

P075 

P074 

P073 

P072 

P071 

SE49 

SE50 

SE48 

SE51 

SE47 

SE52 

SE46 

SE53 

SE45 

SE54 

SE44 

SE55 

SE43 

SE56 

SE42 

SE57 

P085 

P086 

P087 

P088 

P089 

P090 

P091 

P092 

P093 

P094 

P095 

P096 

P097 

P098 

P112 

P111 

P110 

P109 

P108 

P107 

P106 

P105 

P104 

P103 

P102 

P101 

P100 

P099 

SE64 

SE65 

SE63 

SE66 

SE62 

SE67 

SE61 

SE68 

SE60 

SE69 

SE59 

SE70 

SE58A 

SE71 

SE58 

P113 

P114 

P115 

P116 

P117 

P118 

P119 

P120 

P121 

P123 

P125 

P126 

P127 

SE72 

P141 

P140 

P139 

P138 

P137 

P136 

P135 

P134 

P133 

P132 

P131 

P130 

P129 

P128 

SE80 

SE81 

SE79 

SE81A 

SE78 

SE82 

SE77 

SE83 

SE76 

SE84 

SE75 

SE85 

SE74 

SE86 

SE73 

P143 

P144 

P145 

P146 

P147 

P148 

P149 

P150 

P151 

P152 

P153 

P154 

P155 

SE87 

P169 

P168 

P167 

P166 

P165 

P164 

P163 

P162 

P161 

P160 

P159 

P158 

P157 

P156 

SE91 

SE90 

SE89 

SE88 

Stand Alone eSE Exhibits Shared eSE Exhibits 

eSE Presentations 1-31 can be accessed from any shared eSE computer. 

eSE Presentations 32-41 can be accessed on assigned computer only. See sign in eSE area for details. 

P170 

P171 

P172 

P173 

P174 

P175 

P176 

P177 

P184 

P183 

P182 

P181 

P180 

P179 

P178

Printed and Electronic Scientific Exhibits (eSE) 


SCIENTIFIC EXHIBITS 

Adult Brain ................1-39 

Functional................40-43 

Head and Neck ........44-70 

Interventional ..........71-73 

Pediatrics ................75-81 

Socioeconomic ..............82 

Spine ......................83-91 

SE1 

SE2 

SE3 

SE4 

SE5 

SE6 

SE7 

SE8 

SE16 

SE15 

SE17 

SE14 

SE16A 

SE18 

SE13 

SE20 

SE12 

SE21 

SE11 

SE22 

SE10 

SE9 

SE23 

SE24 

ELECTRONIC SCIENTIFIC EXHIBITS (eSE) 

SE33 

SE34 

SE32 

SE35 

SE31 

SE36 

SE30 

SE37 

SE29 

SE38 

SE27 

SE39 

SE26 

SE40 

SE25 

SE41 

Shared/(Stand Alone) 

Adult Brain ................1-10 

Functional ....................11 

(32, 35) 

Head and Neck ........12-25 (36-37) 

Interventional ................26 (38) 

Pediatrics ................27-29 

Spine ......................30-31 

(39-41) 

SE49 

SE50 

SE48 

SE51 

SE47 

SE52 

SE46 

SE53 

SE45 

SE54 

SE44 

SE55 

SE43 

SE56 

SE42 

SE57 

SE64 

SE65 

SE63 

SE66 

SE62 

SE67 

SE61 

SE68 

SE60 

SE69 

SE59 

SE70 

SE58A 

SE71 

SE58 

SE72 


SE80 

SE81 

SE79 

SE81A 

SE78 

SE82 

SE77 

SE83 

SE76 

SE84 

SE75 

SE85 

SE74 

SE86 

SE73 

SE87 

SE91 

SE90 

SE89 

SE88 

(As of 3/24/05) 

Stand Alone eSE Exhibits Shared eSE Exhibits 

eSE Presentations 1-31 can be accessed from any shared eSE 

computer. eSE Presentations 32-41 can be accessed on assigned 

computer only. Refer to sign in eSE area for further details. 

XIII 




XIV 


(As of 3/24/05) SCIENTIFIC POSTERS 

P001 

P002 

P003 

P004 

P005 

P006 

P007 

P008 

P009 

P010 

P011 

P012 

P013 

P014 

P028 

P027 

P026 

P025 

P024 

P023 

P022 

P021 

P020 

P019 

P018 

P017 

P016 

P015 

P029 

P030 

P031 

P032 

P033 

P034 

P035 

P036 

P037 

P038 

P039 

P040 

P041 

P042 

Scientific Posters Exhibits 

P056 

P055 

P054 

P053 

P052 

P051 

P050 

P049 

P048 

P047 

P046 

P044 

P043 

P057 

P058 

P059 

P060 

P061 

P062 

P063 

P064 

P065 

P066 

P067 

P068 

P069 

P070 


P084 

P083 

P082 

P081 

P080 

P079 

P078 

P077 

P076 

P075 

P074 

P073 

P072 

P071 

P085 

P086 

P087 

P088 

P089 

P090 

P091 

P092 

P093 

P094 

P095 

P096 

P097 

P098 

P112 

P111 

P110 

P109 

P108 

P107 

P106 

P105 

P104 

P103 

P102 

P101 

P100 

P099 

P113 

P114 

P115 

P116 

P117 

P118 

P119 

P120 

P121 

P123 

P125 

P126 

P127 

P130 

P129 

P128 

Adult Brain ................1-81 

Functional..............82-102 

Head and Neck ....103-120 

Interventional ......121-148 

Pediatrics ............149-172 

Socioeconomic ............173 

Spine ..................174-184 

P141 

P140 

P139 

P138 

P137 

P136 

P135 

P134 

P133 

P132 

P131 

P143 

P144 

P145 

P146 

P147 

P148 

P149 

P150 

P151 

P152 

P153 

P154 

P155 

P169 

P168 

P167 

P166 

P165 

P164 

P163 

P162 

P161 

P160 

P159 

P158 

P157 

P156 

P170 

P171 

P172 

P173 

P174 

P175 

P176 

P177 

P184 

P183 

P182 

P181 

P180 

P179 

P178

401 301 

Cardinal 

Health 

407 406 307 306 

VSM 

MedTech 

409 

Kopp 

Develop. 

411 

Vital 

Images, 

Inc. 

417 317 

Arthrocare 

419 

421 

Night- 

Hawk 

Technical Exhibits 

Philips 

Medical 

Systems 

Hitachi 

Medical 

Stryker 

Lippincott 

Williams 

311 

GE 

Healthcare 

201 

207 

Springer Integra 

Spinal 

Food Service 

Boston 

Scientific 

322 

Advanced 

Bio 

324 

ASNR 

223 

Invivo 

225 

SNR 

217 

Exhibit Hall A — Level 300 

(As of 3/24/05) 

200 

Kyphon 

Inc. 

101 

Bracco 

Diagnos. 

103 

4-D 

Neuro. 

107 

MicroVention 

111 

222 123 

National 

Library 

Siemens 

Medical 

Berlex 

Imaging 

100 

Micrus 

104 

Shelley 

Med. Tech. 

106 

Cook Inc. 

110 

TeraRecon 

117 116 

Advanced 

Imaging 

Virtual 

Rad. 

118 

Elsevier 

120 

Amirsys 

XV 




XVI 

Technical TECHNICAL Exhibits (As of 3/24/05) EXHIBITS 

Metro Toronto Convention Centre — Exhibit Hall A 

Monday, May 23 — Welcome Reception ........................................................6:00 pm – 7:30 pm 

Tuesday, May 24 through Thursday, May 26 ..................................................9:30 am – 4:00 pm 

Wednesday, May 25 — Evening of Comedy & Cuisine......................Reception 6:15 pm – 7:30 pm 

Performance 8:00 pm - 9:00 pm 

4-D Neuroimaging ......................................Booth 103 

9727 Pacific Heights Boulevard 

San Diego, CA 92121 

Advanced Biomaterial Systems, Inc. ..........Booth 322 

100 Passaic Avenue 

Chatham, NJ 07928 

Advanced Imaging Research ......................Booth 123 

4700 Lakeside Avenue, Suite 400 

Cleveland, OH 44114 

Amirsys, Inc...............................................Booth 120 

25092 Morro Court 

Laguna Hills, CA 92653 

ArthroCare Spine........................................Booth 417 

680 Vaqueros Avenue 

Sunnyvale, CA 94085 

ASNR/NER Foundation ................................Booth 324 



Berlex Imaging ..........................................Booth 117 

P.O. Box 100 

Montville, NJ 07045 

Boston Scientific........................................Booth 217 

47900 Bayside Parkway 

Fremont, CA 94538 

Bracco Diagnostics, Inc. ............................Booth 101 

107 College Road East 

Princeton, NJ 08543 

Cardinal Health, Special Procedures ..........Booth 401 

1500 Waukegan Road 

McGaw Park, IL 60085 

Cook Incorporated......................................Booth 106 

750 Daniels Way, P.O. Box 489 

Bloomington, IN 47402-0489 

Elsevier ....................................................Booth 118 

1 Goldthorne Avenue 

Toronto, ON M8Z 5S7, Canada 

GE Healthcare............................................Booth 201 

3000 N. Grandview Boulevard, W-402 

Waukesha, WI 53188 

Hitachi Medical Systems America, Inc. ......Booth 311 

1959 Summit Commerce Park 

Twinsburg, OH 44087 

Integra Spinal Specialties ..........................Booth 207 

12001 Network Blvd., Building F 

San Antonio, TX 78249 

Invivo ........................................................Booth 223 

N27 W23676 Paul Road 

Pewaukee, WI 53072

TECHNICAL EXHIBITS 

Technical Exhibits (As of 3/24/05) 

Kopp Development Inc. ..............................Booth 409 

785 NE Dixie Highway 

Jensen Beach, FL 34957 

Kyphon Inc. ..............................................Booth 200 

1221 Crossman Avenue 


Lippincott, Williams & Wilkins ..................Booth 307 

21 Ripley Crescent 

Brampton, ON L6Y 4S8, Canada 

MicroVention, Inc.......................................Booth 107 

75 Columbia, Suite A 

Aliso Viejo, CA 92656 

Micrus Corporation ....................................Booth 100 

610 Palomar Avenue 


National Library of Medicine (NLM)..................Booth 222 

University of Washington, Box 357155 

Seattle, WA 98195 

NightHawk Radiology Services ..................Booth 406 

250 Northwest Boulevard, Suite 202 

Coeur d’Alene, ID 83814 

Philips Medical Systems ............................Booth 301 

3000 Minuteman Road 

Andover, MA 01810 

Shelley Medical Imaging Technologies ......Booth 104 

157 Ashley Crescent 

London, ON N6E 3P9, Canada 

Siemens Medical Solutions ........................Booth 111 

1230 Shorebird Way 

Mountain View, CA 94043 

Springer ....................................................Booth 306 

175 Fifth Avenue 

New York, NY 10010-7858 

Stryker ......................................................Booth 317 

4100 East Milham 

Kalamazoo, MI 49001 

TeraRecon, Inc...........................................Booth 110 

2955 Campus Drive #325 

San Mateo, CA 94403 

Virtual Radiologic Consultants ....................Booth 116 

5995 Opus Parkway, Suite 200 

Minneapolis, MN 55343-9058 

Vital Images, Inc. ......................................Booth 411 

3300 Fernbrook Lane North, Suite 200 

Plymouth, MN 55447 

VSM MedTech Ltd. ....................................Booth 407 

9 Burbridge Street 

Coquillam, BC V3K 7B2, Canada 

World Federation of Neuroradiological 

Societies (WFNRS) ....................................Booth 225 

Meeting Dates: 

Adelaide, Australia – March 19-24, 2006 

XVII 




XVIII 

General GENERAL Information INFORMATION 

MEETING REGISTRATION 

Registration will take place in the North Building 

Lobby (Level 200) of the Metro Toronto Convention 

Centre. The registration desk will be open during 

the following hours: 

Friday, May 20 ......................5:00 pm - 8:00 pm 

Saturday, May 21 ..................8:00 am - 6:00 pm 

Sunday, May 22 ....................6:30 am - 6:00 pm 

Monday, May 23 ....................6:30 am - 6:00 pm 

Tuesday, May 24 ....................6:30 am - 6:00 pm 

Wednesday, May 25 ................6:30 am - 6:00 pm 

Thursday, May 26 ..................6:30 am - 6:00 pm 

Friday, May 27 ....................6:30 am - 11:45 am 

SPEAKER READY ROOM LOCATION & HOURS 


— Room 104B (Level 100) 

Friday, May 20........................5:00 pm - 8:00 pm 

Saturday, May 21 ....................8:00 am - 6:00 pm 

Sunday, May 22 through 

Thursday, May 26 ................6:00 am - 6:00 pm 

Friday, May 27 ....................6:00 am - 11:45 am 

NAME BADGES 

Please wear name badges at all times while you are 

attending the scientific sessions, social programs, 

and technical exhibits. Badge colors are identified 

as follows: 

ASNR, ASFNR, ASHNR, ASITN, ASPNR, 

or ASSR Member ......................................Blue 

Non-Member ..............................................Green 

Fellow/Trainee ................................................Tan 

Other Professional ......................................Yellow 

Guest ........................................................Peach 

Exhibitor ......................................................Gold 

Staff ........................................................Purple 

CME PAVILION 


— Room 104A (Level 100) 

Saturday, May 21 ..................1:00 pm - 9:00 pm 



Friday, May 27 ......................6:30 am - 1:00 pm 

COMMITTEE/SPECIALTY/REGIONAL SOCIETY MEETINGS 

Please refer to the Daily Postings on the Meetings & 

Announcements Board located in the Metro Toronto 

Convention Centre North Building Lobby (Level 200). 

MEETINGS & ANNOUNCEMENTS BOARD 

The Meetings & Announcements Board is located in 

the North Building Lobby (Level 200) of the Metro 

Toronto Convention Centre. Please refer to the Daily 

Postings on the Meetings & Announcements Board for 

information on committee meetings. 

MESSAGE CENTER 

The Message Center is located on Level 300 near the 

escalators in Hall B. They will have computer terminals 

available to registered attendees that can be used to 

access/send external email and to leave internal 

messages for other attendees and/or exhibitors. 

METRO TORONTO CONVENTION CENTRE INFORMATION 


255 Front Street West 

Toronto, ON M5V 2W6, Canada 

Phone: 416-585-8000

General GENERAL Information INFORMATION 

EMERGENCY SERVICE PROCEDURE 

Within the Metro Toronto Convention Centre: 

In the event of a medical emergency, please contact 

Security Control immediately. Attendees may contact 

Security Control by dialing extension 8160 from any 

house phone located in the facility. Emergency personnel 

will be dispatched immediately to your location. 

The caller should provide the following: 

1. Determine name of specific meeting room 

or exhibit hall where the situation has occurred. 

2. Identify yourself as an ASNR attendee, reference 

your exact location, and provide details on the 

nature of the emergency situation. 

3. Provide a brief but concise description of the 

problem, be prepared to answer any questions 

that the operator may ask you, and remain on 

the line. 

Contacting Security Control will greatly minimize 

response time in the event an emergency medical 

unit needs to report to the Convention Centre. 

Security personnel can quickly assess the situation 

and bring emergency personnel directly to the 

scene, saving precious minutes. For this reason, the 

Metro Toronto Convention Centre management 

requests that attendees not contact 911 directly. 

NEAREST HOSPITAL 

St. Michael’s Hospital 

30 Bond Street 

Toronto, ON M5B 1WB, Canada 

Phone: 416-360-4000 

NEAREST PHARMACY 

Shoppers Drugmart - Royal Bank Plaza 

(Underground - near the Fairmont Royal York) 

200 Bay Street 

Toronto, ON, M5J 2J3 Canada 

Phone: 416-865-0001 

Hours: Monday - Friday, 7:00 am - 6:30 pm 

Saturday, 10:00 am - 4:30 pm 

Sunday, CLOSED 

NEAREST PHARMACY (OPEN 24-HOURS) 

Shoppers Drugmart at Bay & Gerrard 

700 Bay Street 

Toronto, ON, M5G 1Z6, Canada 

Phone: 416-979-2424 

HOTEL INFORMATION 

The Fairmont Royal York Hotel 


Toronto, ON M5J 1E3 Canada 

Phone: 416-368-2511 

Fax: 416-368-9040 

THE FAIRMONT ROYAL YORK HOTEL 

BUSINESS CENTER SERVICES 

The Xerox Centre at The Fairmont Royal York provides 

business travelers with the best printing technology 

available. It also offers an “office away from the 

office” for guests of The Fairmont Royal York. The 

quiet, efficient facility will let you be more productive 

during your stay. 

Hours: Mon. - Fri. 7:00 am - 10:00 pm 

Sat. and Sun. 10:00 am - 6:00 pm 

PHOTOCOPY SERVICE 

The Printing House (TPH) 


Toronto, ON Canada 

Phone: 416-865-1660 

Fax: 416-865-9997 

Hours: Mon. - Fri. 8:00 am - 6:00 pm 

Sat. and Sun. 10:00 am - 4 pm 

XIX 




XX 

GENERAL INFORMATION 

General Information (continued) 

FOOD SERVICE 

ASNR Food Service will be served in Exhibit Hall A during 

technical exhibition hours or in Room 105/106 Foyer. 

Continental Breakfasts, Morning and Afternoon Coffee 

Service and Box Lunches are provided complimentary 

throughout the week. Please refer to the schedule below. 

Continental Breakfast 


Friday, May 27 ..................Room 105/106 Foyer 

(Level 100) 

Additional seating in Room 104D 

How-To Session Breakfasts 


Thursday, May 26..............Room 105/106 Foyer 

(Level 100) 

Morning Breaks 

Sunday, May 22 ............Theatre Foyer (Level 100) 

Monday, May 23 ....Room 105/106 Foyer (Level 100) 

and Theatre Foyer (Level 100) 

Tuesday, May 24 through 

Thursday, May 26 ....................................Hall A 

Friday, May 27 ....................Room 105/106 Foyer 

(Level 100) 

Box Lunches 


Monday, May 23..................Room 105/106 Foyer 

(Level 100) 



For Breaks and Lunches, use Room 104D for additional seating Saturday 

through Monday & Friday when Technical Exhibition is closed. 

How-To Session Lunches 


Wednesday, May 25 ..........Room 105/106 Foyer 

(Level 100) 

Afternoon Breaks 

Saturday, May 21 through 

Sunday, May 22..........Theatre Foyer (Level 100) 

Monday, May 23....Room 105/106 Foyer (Level 100) 

Theatre Foyer (Level 100) 



How-To Session Receptions 

Sunday, May 22 

(5:30 pm - 7:00 pm)............Room 105/106 Foyer 

(Level 100) 

Tuesday, May 24 

(6:00 pm - 7:30 pm)............Room 105/106 Foyer 

(Level 100) 

MEETING LOCATION: Metro Toronto Convention Centre 

NOTE: All scientific sessions and exhibits are 

located at the Metro Toronto Convention Centre. 

Registration 

North Building Lobby (Level 200) 

CME Pavilion Terminals 

Room 104A (Level 100) 

E-Access/Messaging Center 

Level 300 (outside Hall B) 

How-To Breakfast/Lunch/Reception Sessions 

Room 105/106 

Focus/Scientific Paper Sessions 

Room 105/106, Theatre, Room 107 and Room 205 

ELECTRONIC LEARNING CENTER (ELC) 

Workshops 

Room 103 (Level 100) 

Lectures 


NATIONAL LIBRARY OF MEDICINE (NLM) 

Workshops 


EXHIBITS 

Scientific Exhibits, Electronic Scientific Exhibits, 

Scientific Posters 

Hall B (Level 300) 

Technical Exhibits 

Hall A (Level 300) 

MISCELLANEOUS 

Past-Presidents’ and Executive Committee Office 

Room 203D (Level 200) 



Meetings & Announcements Board 

North Building Lobby (Level 200) 

Coat Check 

Level 100 by escalators 

Hours of Operation: 

Saturday, May 21 ................12:00 pm - 6:00 pm 



Friday, May 27 ......................6:30 am - 1:00 pm

OPTIONAL TOURS 

Optional Tour Desk Hours 

The Fairmont Royal York Hotel – Tudor 8 (Main Mezzanine Level) 

Saturday, May 21 ..............................................................................................12:00 pm – 4:00 pm 

Sunday, May 22 through Thursday, May 26............................................................8:00 am – 2:00 pm 

SOCIAL PROGRAM 

Social Program 

An exciting social program has been planned for registrants and their registered guests during the 

ASNR 43rd Annual Meeting. 

Welcome Reception with 

Evening of Comedy and Cuisine 

Technical Exhibitors 

Wednesday, May 25 

Monday, May 23 

6:15 pm - 7:30 pm - Reception - Hall A (Level 300) 

6:00 pm – 7:30 pm 

8:00 pm – 9:00 pm - Second City Performance 

Hall A (Level 300) 

Theater (Level 100) 



The Welcome Reception with Technical Exhibitors offers 

the perfect opportunity for a preview of this year’s 

Technical Exhibit, the ASNR’s annual showcase for the 

newest products and services for the field of 

neuroradiology. Enjoy complimentary pre-dinner hors 

d’oeuvres and beverages while you learn about the newest 

technology. Connect with old friends, make new ones and 

meet representatives from the companies participating in 

this year’s technical exhibition. 

This casual social setting allows plenty of time for informal 

discussion with the company representatives. So bring your 

product and service challenges and come in search of 

solutions to the place where advanced technology and 

diagnostic and interventional neuroradiological excellence 

come together. 

The Scientific Exhibit (poster, scientific and electronic 

scientific exhibits) will also be available for viewing 

throughout the evening’s reception. 

Ticket required for admission: A ticket to the Welcome 

Reception with Technical Exhibitors is included in the fee for 

registration categories that include Monday, May 23 and in 

the Guest Hospitality fee. 

Combine the cuisine of Toronto’s many ethnic cultures, the 

comedy of the world-renowned Second City Toronto and the 

camaraderie of fellow neuroradiologists from all over the 

world and what do you get? A spectacular evening of fun, 

food and frivolity! During a pre-performance reception, enjoy 

an appetizer buffet highlighting the tastes and cuisines 

representing the many cultures from around the world that 

have found a home in Toronto. Then settle in for a laughterfilled 

performance provided by the Second City Toronto 

comedy troupe. This internationally famous institution noted 

for improvisation has been the launching pad for many of the 

comedy world’s best and brightest. Second City Toronto 

alumni include Dan Aykroyd, Gilda Radner, John Candy, 

Catherine O’Hara, Mike Myers and Martin Short, to name 

just a few. Skilled at finding humor in the absurdity of 

everyday life, their satire can be appreciated by everyone. 

This evening’s show will feature “The Best of Second City” 

sketches, a look at the field of neuroradiology through the 

eyes of these talented comedians, and a chance for audience 

interaction as they create sketches from ideas shouted from 

the audience. 

Ticket required for admission: A ticket to the Evening of 

Comedy and Cuisine is included in the fee for registration 

categories that include Wednesday, May 25 and in the Guest 

Hospitality fee. 

XXI 




XXII 

GUEST HOSPITALITY 

Guest Hospitality 

GUEST HOSPITALITY 


Tudor 8 (Main Mezzanine Level) 

The Guest Hospitality area, available to ASNR registered 

guests, provides complimentary food service, visitor 

information and special presentations. Teens and 

younger individuals who are with registered guests, but 

are not themselves registered, may also visit the 

hospitality room. Afternoon snacks and beverages will 

be offered on Saturday, May 21. Continental breakfast, 

afternoon snacks and beverages will be available from 

Saturday, May 21 - Thursday, May 26 and continental 

breakfast only on Friday, May 27. 

Guest Hospitality is a great place to start your 

mornings and plan the remainder of your day. It’s 

an ideal location to see old friends and meet 

new acquaintances. 

A representative from the Toronto area will be 

available to acquaint you with suggestions on what to 

see and do while in Toronto. Be sure to stop by to 

pick up visitor information and brochures available in 

Guest Hospitality on citywide attractions, downtown 

maps, and shopping and restaurant guides to assist 

you in planning your week. 

GUEST HOSPITALITY HOURS 


Tudor 8 (Main Mezzanine Level) 

Saturday, May 21 ................12:00 pm – 4:00 pm 


Thursday, May 26................8:00 am – 2:00 pm 

Friday, May 27 ....................8:00 am – 11:30 am 

SPECIAL PROGRAMS IN GUEST HOSPITALITY 

A highlight of the Guest Hospitality program is a 

schedule of complimentary entertaining presentations 

on a variety of interesting topics. Presentations will take 

place in Tudor 7 at the The Fairmont Royal York Hotel. 

“Welcome to Toronto” Orientation 

Toronto offers a theatre scene that rivals London and New 

York, shopping with an international flair, miles of 

waterfront, beaches and walking trails, over 20 ethnically 

diverse neighborhoods and many fun attractions. So much to 

see and do, but so little time. 

This session will help participants make the most of their 

stay in Toronto. The presentation will provide tips on how to 

get those sought after theatre tickets, the hottest dining 

spots, current exhibits at Toronto’s world class museums, 

those unique and out of the way boutiques and how to get 

from “here to there.” 

Presentation given by: Tourism Toronto 

This program will be offered twice. 

Sunday, May 22 10:00 am – 11:00 am 

Tuesday, May 24 10:00 am – 11:00 am 

Lights, Camera, Action! 

Are you a film buff? Have you ever wondered what goes on 

behind the scenes when filming a movie? What happens 

when a film crew uses your home? How do they film those 

chase scenes? How pampered are those movie stars? Now is 

your chance to find out! Toronto is becoming known as 

Hollywood North because of the many feature films that have 

been made there including My Big Fat Greek Wedding and, 

more recently, Confessions of a Teenage Drama Queen. 

Rhonda Silverstone, Film Commissioner, Toronto Film and 

Television Office, and Donna Zuchlinski, Provincial Film 

Commissioner, Ontario, will give you an insider’s view of what 

it takes to get from concept to the big screen. During their 

presentation, they will share stories of the challenges, the 

strangest requests and funniest moments they’ve experienced 

as the mavens of Toronto’s Hollywood North. 

Monday, May 23 10:00 am – 11:00 am 

Picasso and Ceramics 

The Gardiner Museum of Ceramic Art is the only museum in 

Canada entirely devoted to ceramics. Situated in the heart of 

downtown Toronto, across from the Royal Ontario Museum, the 

Gardiner is one of Toronto's outstanding cultural destinations. 

Although the museum will be closed for renovation during the 

ASNR meeting, ASNR Guest Hospitality participants can still 

have an experience from this “jewel box” of ceramic treasurers. 

Sue Jeffries, Gardiner Museum Curator of Contemporary 

Ceramics, will present “Picasso and Ceramics”, highlights from 

the Museum’s most recent exhibition. 

Pablo Picasso was the most influential artist of the 20th 

century. His universally recognized name connotes energy, 

daring and originality. Although seldom described as a ceramic 

artist, Picasso devoted considerable energy to the medium over 

a period of twenty years at the height of his fame, producing an 

estimated 4,500 ceramic works. Ms. Jeffries’ presentation will 

highlight his appropriation of historic ceramic traditions and 

his major innovations when he took these forms and reinvigorated 

them in his own inimitable way. 

Thursday, May 26 10:00 am – 11:00 am

FUTURE MEETINGS 

Future ASNR Annual Meetings 

2006 

44th Annual Meeting 

April 29 – May 5 

San Diego Convention Center 

San Diego, California 

2007 


June 9 – 15 

Hyatt Regency Chicago Hotel 

Chicago, Illinois 

2008 


May 31 – June 6 

Morial Convention Center 

New Orleans, Louisiana 

2009 


May 16 – 22 

Vancouver Convention & 

Exhibition Centre 

Vancouver, British Columbia, 

Canada 

ASNR Past Presidents and Founders 

PAST PRESIDENTS 

1962-64 Juan M. Taveras, MD* 

1964-65 Mannie M. Schechter, MD* 

1965-66 Donald L. McRae, MD* 

1966-67 Ernest H. Wood, MD* 

1967-68 Harold O. Peterson, MD* 

1968-69 Colin B. Holman, MD 

1969-70 Giovanni Di Chiro, MD* 

1970-71 D. Gordon Potts, MD 

1971-72 Norman E. Chase, MD 

1972-73 Fred J. Hodges, III, MD 

1973-74 T. Hans Newton, MD 

1974-75 Hillier L. Baker, Jr., MD 

1975-76 Irvin I. Kricheff, MD 

1976-77 Norman E. Leeds, MD 

1977-78 Sadek K. Hilal, MD* 

1978-79 Stephen A. Kieffer, MD 

1979-80 David O. Davis, MD 

1980-81 George Wortzman, MD 

1981-82 Gabriel H. Wilson, MD 

1982-83 Arthur E. Rosenbaum, MD 

1983-84 O. Wayne Houser, MD 

1984-85 Samuel M. Wolpert, MD 

1985-86 R. Thomas Bergeron, MD 

1986-87 Derek C. Harwood-Nash, MD* 

1987-88 Michael S. Huckman, MD 

1988-89 Anne G. Osborn, MD 

1989-90 Joseph F. Sackett, MD 

1990-91 Anton N. Hasso, MD, FACR 

1991-92 R. Nick Bryan, MD, PhD 

1992-93 David Norman, MD 

2010 


May 15 – 21 

Hynes Convention Center 

Boston, Massachusetts 

1993-94 Glenn Forbes, MD 

1994-95 Robert M. Quencer, MD 

1995-96 Robert R. Lukin, MD 

1996-97 Burton P. Drayer, MD 

1997-98 Richard E. Latchaw, MD 

1998-99 A. James Barkovich, MD 

1999-00 Eric J. Russell, MD, FACR 

2000-01 William S. Ball, Jr., MD 

2001-02 William P. Dillon, MD 

2002-03 Patrick A. Turski, MD 

2003-04 Charles M. Strother, MD 

FOUNDING MEMBERS 

Norman E. Chase, MD 

Giovanni Di Chiro, MD* 

William N. Hanafee, MD 

Fred J. Hodges, III, MD 

Colin B. Holman, MD 

Norman E. Leeds, MD 

Eugene V. Leslie, MD 

Donald L. McRae, MD* 

Thomas H. Newton, MD 

Harold O. Peterson, MD* 

D. Gordon Potts, MD 

Mannie M. Schechter, MD* 

Juan M. Taveras, MD* 

Ernest H. Wood, MD* 

*deceased 

XXIII 




XXIV 

PAST MEETINGS 

Past ASNR Annual Meetings 

Organizational Meeting 

May 19, 1962 

Keene’s English Chophouse 

New York 

Second Business Meeting 

October 5, 1962 

Shoreham Hotel 

Washington, DC 

First Annual Meeting 

October 7, 1963 

Queen Elizabeth Hotel 

Montreal 

Second Annual Meeting 

September 23, 1964 

Waldorf Astoria 

New York 

Third Annual Meeting 

June 11, 1965 

Dennis Hotel 

Atlantic City 

Fourth Annual Meeting 

June 15-16, 1966 

Sheraton-Park Hotel 


Fifth Annual Meeting 

May 15, 1967 

Columbia University 

New York 

Sixth Annual Meeting 

September 27-28, 1968 

Jung Hotel 

New Orleans 

Seventh Annual Meeting 

May 13-19, 1969 

Joint Meeting with American Association 

of Neurological Surgeons 

Sheraton-Cleveland Hotel 

Cleveland 

Eighth Annual Meeting 

February 12-13, 1970 

Washington Hilton 

Washington 

Ninth Annual Meeting 

May 27-29, 1971 

Fairmont Hotel 

San Francisco 

Tenth Annual Meeting 

February 21-24, 1972 

Maria-lsabel Sheraton 

Mexico City 

Eleventh Annual Meeting 

May 26-28, 1973 

Statler Hilton 

Boston 

Twelfth Annual Meeting 

March 14, 1974 

(In conjunction with X Symposium 

Neuroradiologicum) 

Convention Center 

Punta del Este, Uruguay 

Thirteenth Annual Meeting 

June 3-7, 1975 

Bayshore Inn 

Vancouver 

Fourteenth Annual Meeting 

May 18-22, 1976 

Peachtree Plaza 

Atlanta 

Fifteenth Annual Meeting 

March 27-31, 1977 

Hamilton Princess Hotel 

Bermuda 

Sixteenth Annual Meeting 

February 26-March 2, 1978 

Hyatt Regency 

New Orleans 

Seventeenth Annual Meeting 

May 20-24, 1979 

Hotel Toronto 

Toronto 

Eighteenth Annual Meeting 

March 16-21, 1980 

Century Plaza 

Los Angeles 

Nineteenth Annual Meeting 

May 5-9, 1981 

Marriott Hotel 

Chicago 

Twentieth Annual Meeting 

October 10-16, 1982 

(In conjunction with XII Symposium 


Washington Hilton 


Twenty-First Annual Meeting 

June 5-9, 1983 

St. Francis Hotel 

San Francisco

Past ASNR PAST Annual Meetings MEETINGS 

(continued) 

Twenty-Second Annual Meeting 

June 2-7, 1984 

Westin Copley Place Hotel 

Boston 

Twenty-Third Annual Meeting 

February 18-23, 1985 

Marriott Hotel 

New Orleans 

Twenty-Fourth Annual Meeting 

January 19-23, 1986 

Sheraton Harbor Island Hotel 

San Diego 

Twenty-Fifth Annual Meeting 

(Silver Anniversary) 

May 10-15, 1987 

New York Hilton 

New York 

Twenty-Sixth Annual Meeting 

May 15-20, 1988 

Chicago Hilton & Towers 

Chicago 

Twenty-Seventh Annual Meeting 

March 19-24, 1989 

The Peabody Orlando 

Orlando 

Twenty-Eighth Annual Meeting 

March 19-23, 1990 

Century Plaza Hotel & Tower 

Los Angeles 

Twenty-Ninth Annual Meeting 

June 9-14, 1991 

The Washington Hilton and Towers 


Thirtieth Annual Meeting 

May 31-June 5, 1992 

Adam’s Mark 

St. Louis 

Thirty-First Annual Meeting 

May 17-20, 1993 

Vancouver Trade and Convention Centre 

Vancouver 

Thirty-Second Annual Meeting 

May 3-7, 1994 

Opryland Hotel and Conference Center 

Nashville 

Thirty-Third Annual Meeting 

May 23-27, 1995 

Sheraton Chicago Hotel and Towers 

Chicago 

Thirty-Fourth Annual Meeting 

June 23-27, 1996 

Washington State Convention and Trade Center 

Seattle 

Thirty-Fifth Annual Meeting 

May 18-22, 1997 


Toronto 

Thirty-Sixth Annual Meeting 

May 17-21, 1998 

(In conjunction with XVI Symposium 


Pennsylvania Convention Center 

Philadelphia 

Thirty-Seventh Annual Meeting 

May 23-28, 1999 

San Diego Convention Center 

San Diego 

Thirty-Eighth Annual Meeting 

April 4-8, 2000 

Hyatt Regency Atlanta 

Atlanta 

Thirty-Ninth Annual Meeting 

April 23-27, 2001 

Hynes Convention Center 

Boston 

Fortieth Annual Meeting 

May 13-17, 2002 

Vancouver Convention & Exhibition Centre 

Vancouver 

Forty-First Annual Meeting 

April 28 – May 2, 2003 

Marriott Wardman Park Hotel 


Forty-Second Annual Meeting 

June 7 – June 11, 2004 

Washington State Convention and Trade Center 

Seattle 

XXV 




XXVI 

AWARDS & HONORS 

Awards and Honors 

2005 ASNR Gold Medal Award 

The Gold Medal fosters the highest standards of the American Society of Neuroradiology, 

based on exceptional quality, service, and excellence, and not necessarily on fame. It 

emphasizes both professional and personal attributes… individuals who are superb 

neuroradiologists, clinicians, or scientists, and truly outstanding. The recipients are 

individuals who have extended themselves beyond furthering their own careers through 

contributions at all levels of professional strata, with an accent on consistency and duration 

of these outstanding contributions. 

Samuel M. Wolpert, MD 

In 1976, Dr. Wolpert published “Angiography of Posterior Fossa 

Dr. Samuel (Muli) Wolpert was Tumors.” The timing of the publication was unfortunate, since 

born in Johannesburg, South computerized axial tomography had recently been discovered; 

Africa in 1930, educated in and at the time, many thought that there would be no future 

public schools and went to the need for posterior fossa angiography. In 1977, he became 

University of the Witwatersrand Treasurer of the ASNR and during his tenure presented the idea 

Medical School, where he of an ASNR journal to the President of the ASNR, Dr. Sadek 

graduated in 1953. In South Hilal, and his Executive Committee. Dr. Wolpert and Dr. Leeds 

Africa, after completing his chaired the search committee for an editor-in-chief and, 

internship, he spent time as a eventually, Dr. Juan Taveras was chosen as Editor of the AJNR. 

resident in Internal Medicine, He invited Dr. Wolpert to be the first Associate Editor, and the 

Orthopaedic Surgery and finally initial issue of the AJNR appeared at the beginning of 1980. 

Radiology. After traveling to London, he trained at Guy's The history of neuroradiology has been one of Muli's interests, 

Hospital and St. Mary's Hospital, and received a D.M.R.D. and recently he edited a series of papers on Neuroradiologic 

(Diploma in Medical Radiodiagnosis) in 1960. He then returned Classics which appeared in the AJNR. He became President of 

to radiological hospital practice in South Africa. 

the ASNR in 1984. 

In 1963-1964, Dr. Mannie Schechter, Senior Neuroradiologist In 1992, Dr. Wolpert was the lead neuroradiological investigator 

at the Albert Einstein College of Medicine (AECOM) in New York in one of the first studies on the value of intravenous 

and a major influence on Dr. Wolpert's career, visited his home recombinant tissue plasminogen activator in the treatment of 

country of South Africa. Mannie offered Dr. Wolpert a two-year acute stroke. The Boston Floating Hospital, a pediatric hospital, 

NIH fellowship in Neuroradiology. Through this fortunate visit of is part of NEMC, and many of Dr. Wolpert's publications were on 

Dr. Schechter to South Africa, Muli emigrated to the United pediatric neuroradiology. This interest culminated in the 

States in December 1964, and commenced a two-year publication of “MRI in Pediatric Neuroradiology,” by Dr. Wolpert 

fellowship at AECOM in January 1965. 

and Dr. Patrick Barnes of Children's Hospital in 1992. Dr. 

Wolpert is a charter member of the Board of Directors of the 

Dr. Wolpert did not initially intend to specialize in 

American Society of Pediatric Neuroradiology (ASPNR). In June 

neuroradiology, but impressed by the quality of the 

2004, because of his interest in and publications on pediatric 

neurologists, neurosurgeons and neuropathologists at the 

neuroradiology, Dr. Wolpert received a Special Recognition Award 

AECOM, as well as by the charisma and competence of Mannie 

from the ASPNR. He has published 3 books, 114 papers and 24 

Schechter, he decided that this was going to be his ultimate 

chapters. He has had multiple speaking engagements, in the 

career choice. In 1967, Drs. Alice Ettinger and Robert Paul of 

United States and overseas, and was Vice-President of the New 

the New England Medical Center Hospitals (NEMC) in Boston 

England Roentgen Ray Society, as well as a member of numerous 

recruited Dr. Wolpert as a neuroradiologist. He stayed at NEMC 

radiological societies. He has been an examiner for the American 

for the next 30 years. 

Board of Radiology and a CAQ examiner in neuroradiology. 

He climbed the academic ladder at Tufts University School of 

Now living in Santa Fe and retired after a part-time 

Medicine, becoming a full professor of Radiology in 1974, and 

neuroradiology stint at the University of New Mexico Health 

of Neurology in 1979. He initiated a fellowship training 

Medical Center, Dr. Wolpert is devoting his time 

program, training 36 fellows between 1971 and 1997, and also 

(unsuccessfully) to reducing his golf handicap, oil painting (also 

started the Boston Neuroradiology Club. With Dr. Bennett 

unsuccessfully) and opera. (He has been a backstage docent at 

Stein's encouragement, Muli commenced embolizing AVMs 

the Santa Fe Opera for the last 5 years). Dr. Wolpert and his 

utilizing silastic emboli, which limited the hemorrhagic 

wife Cynthia have been married 48 years and have three 

complications that could ensue during surgical removal of the 

children, David (a physicist at NASA-Ames in California), 

AVMs. The two physicians were invited to discuss their 

Michelle (a bank officer in Houston), and Steven (a family 

experience at one of the national television morning shows. The 

practitioner in Phoenix). 

show was advertised as “Sam's balls cure stroke.”



Past ASNR Gold Medal Award Recipients 

1995 

Juan M. Taveras, MD* 

T. Hans Newton, MD 

1996 

Sadek K. Hilal, MD* 

Giovanni Di Chiro, MD* 

1997 

Derek C. Harwood-Nash, MB, ChB., 

DSc, FRCPC, FACR, RCRAD(SA)* 

2005 ASNR Honorary Member 

Professor Luc Picard 

1998 

Irvin I. Kricheff, MD 

D. Gordon Potts, MD 

1999 

Grant B. Hieshima, MD 

Michael S. Huckman, MD 

2000 

Hillier L. “Bud” Baker, Jr., MD 

2001 

O. Wayne Houser, MD 


Dr. Luc Picard, a neuroradiologist 

and a neurologist, graduated 

from the University of Nancy, 

France. He was initially trained in 

neurology and worked in close 

cooperation with an interventional 

neuroradiology pioneer in Paris, 

Professor René Djindjian. Having 

been appointed Associate 

Professor of Radiology in 1970, 

he was in charge of all the 

neuroradiological investigations of the University Hospital. 

Thanks to his work, he created a full-fledged independent 

Department of Neuroradiology, of which he was appointed 

Director in 1977. He performed his first series of 

embolizations in 1968, and created a laboratory for 

experimental neuroradiology at Nancy. Having been 

appointed to the new Chair of Professor of Neuroradiology, 

he began his effort for the development of interventional 

neuroradiology. In 1982, he organized the annual meeting 

of the Working Group in Interventional Neuroradiology 

(WIN) at Val d'Isère-France. He is a founding member of the 

World Federation of Interventional and Therapeutic 

Neuroradiology, of which he was the President from 1993 

until 1995. 

With the expansion of his administrative responsibilities, 

he has dedicated himself to the development of 

neuroradiology as a specialty. He was founding member 

of the European Society of Neuroradiology (ESNR) and of 

the French Society of Neuroradiology (SFNR), of which 

he was appointed General Secretary before becoming 

President. Founding member of the Journal of 

Neuroradiology in 1974, he became its Chief Editor from 

1978 until 2002. He is also a founding Editorial Board 

member of Interventional Neuroradiology. As a tribute to 

his vast experience, he has given more than 300 guest 

2002 

R. Thomas Bergeron, MD 

David O. Davis, MD 

2003 

Norman E. Leeds, MD, FACR 

Anne G. Osborn, MD, FACR 

2004 

Ralph Heinz, MD, FACR 

Stephen A. Kieffer, MD, FACR 

*deceased 

lectures (United States, South America, Japan, China 

and India). His publications in diagnostic and 

therapeutic neuroradiology number over 400. He has 

organized many meetings, and particularly the XVIIth 

Symposium Neuroradiologicum in Paris in 2002. In 

1999, he was elected a member of the French National 

Academy of Surgery, and was awarded “Chevalier de la 

Légion d'Honneur” in 2002. At the present time, he 

serves as Vice-President of the World Federation of 

Neuroradiological Societies (WFNRS). 

Toronto, Canada 

Toronto, Canada 

XXVII



XXVIII 



Past ASNR Honorary Member Recipients 

Torsten Almen, MD 

James W. Bull, MD* 

Graeme M. Bydder, MD, ChB 

M. Paul Capp, MD 

Sten Cronqvist, MD* 

B. G. Ziedses des Plantes, MD* 

George du Boulay, MD* 

Richard R. Ernst, MD 

Torgny V. B. Greitz, MD 

Godfrey N. Hounsfield, PhD* 

Yun Peng Huang, MD 

Ian Isherwood, MD 

Pierre Lasjaunias, MD, PhD 


ASNR 2004 Outstanding Presentation Awards 

ASNR is pleased to announce the winners of the Outstanding Presentation Awards given annually to the 

top paper or poster presentation from the prior Annual Meeting in general neuroradiology and the four 

neuroradiology specialties. A $1,000 award was given to each winner. 

General Neuroradiology 

“Evaluation of a Signal Intensity Mask in the 

Interpretation of Functional MR Imaging 

Activation Maps” 

Strigel, R. M., Haughton, V. M., Moritz, C., Field, A. 

S., Badie, B., Wood, D. A., Hartman, M., Rowley, H. A. 

University of Wisconsin at Madison 

Madison, WI 

Berlex Best Paper Award in General Neuroradiology 

“Assessment of Tissue Viability with Quantitative CT 

Perfusion in Acute Ischemic Stroke Patients Treated 

with Intraarterial Thrombolysis” 

Schaefer, P. W., Roccatagliata, L., Ledezma, C. J., 

Gonzalez, R. G., Koroshetz, W. J., Lev, M. H. 


Boston, MA 

Head and Neck Radiology 

“Physiologic Enhancement of the Labyrinth 

Demonstrated on Delayed FLAIR Imaging” 

Butman, J. A. 

Warren G. Magnuson Clinical Center, 

National Institutes of Health 

Bethesda, MD 

Paul C. Lauterbur, PhD 

Dennis LeBihan, MD, PhD 

Marco Leonardi, MD 

Erik Lindgren 

Claude H. Manelfe, MD 

Joseph Ransohoff, MD* 

Jesus Rodriguez-Carbajal, MD 

Lee F. Rogers, MD 

Prof. Lucy Balian Rorke 

Michael Radford Sage, MD, 

FRANZCR, FRCR, FRCPC (Lon), 

FRCPC (Ed), FHKCR (Hon) 

George Schuyler 

S. I. Seldinger, MD 

Fjodor Serbinenko, MD 

Mutsumasa Takahashi, MD 

Michel Ter Pogossian, MD* 

Galdino E. Valvassori, MD 

Marjo S. van der Knaap, MD 

Prof. Jacqueline Vignaud 

M. Gazi Yasargil, MD 

Ian R. Young, BSc, PhD 

*deceased 

Interventional Neuroradiology 

(The Michael Brothers Memorial Award) 

“Three-Dimensional Cerebral Angiography: 

Radiation Dose Comparison with Digital 

Subtraction Angiography” 

Schueler, B., Kallmes, D., Cloft, H. J. 

Mayo Clinic 

Rochester, MN 

Pediatric Neuroradiology 

(The Derek C. Harwood-Nash Award) 

“MR Imaging of the Fetal Cerebellar Vermis In Utero: 

Criteria for Abnormal Development, with Ultrasonographic 

and Clinicopathologic Correlation” 

Robinson, A. J. 1 , Blaser, S. 1 , Toi, A. 2 , Chitayat, D. 2 , 

Ryan, G. 2 , Pantazi, S. 2 , Gundogan, M. 2 Laughlin, S. 1 

1Hospital for Sick Children, Toronto, ON, CANADA, 

2Mount Sinai Hospital, Toronto, ON, CANADA 

Spine Radiology 

“Can Real Time Video Fluoroscopy of the 

Cervical Spine Differentiate the Normal from 

the Abnormal Spine?” 

Rothman, S. L. G. 1 , Go, J. L. 1 , Irvine, B. 2 , 

Kim, P. E. 1 , Zee, C. S. 1 

1University of Southern California, Los Angeles, CA 

2Private Practice, Redondo Beach, CA

AWARDS & Awards and Honors HONORS 

2004 Regional Society Awards 

The American Society of Neuroradiology is pleased to announce the recipients of the 2004 Regional Society 

Awards. These individuals were selected by the respective regional societies as having the best presentation at 

each society’s 2004 Annual Meeting. 

Eastern Neuroradiological Society (ENRS) 

(The Norman E. Leeds Award) 

“Characterization of Cerebral 

Aneurysms for Assessing Risk 

of Rupture Using Patient- 

Specific Computational 

Hemodynamic Models” 

Christopher M. Putman, MD 

Inova Fairfax Hospital 

Falls Church, VA 

Southeastern Neuroradiological Society 

(SENRS) 

“Are Protective Devices Necessary 

for Carotid Stenting?” 

Gregory J. Joseph, MD 

Presbyterian Hospital 

Charlotte, NC 

This award, created in 2004 in 

recognition of consistent excellence and 

lifelong accomplishment in basic or 

clinical neuroscience research, is given 

to an ASNR senior member over the age 

of 50 recognized in the neuroradiology 

field for distinguished long term achievement in 

basic or clinical research. 

The recipient of the award is: 

Dixon M. Moody, MD, FACR 

Wake Forest University, School of Medicine 

Winston-Salem, NC 

Dixon M. Moody, MD, FACR 

Dr. Moody is recognized worldwide for his contributions in 

radiologic-pathology correlation of brain microvascular 

anatomy and disease. A Charles A. Dana Foundation 

award recipient, he also has a long track record of NIH 

funding with the principal grant, 

originally (1984) a Jacob K. 

Javits Neuroscience Investigator 

Award, funded through 2008 – 

the 24th year of this project. 

Dr. Moody recently finished a 6year 

term on the Diagnostic 

Radiology Study Section of the 

National Institutes of Health 

(NIH). He served a 4-year term 

on the National Advisory Council 

of the National Institute of 

Neurological Diseases and Stroke (NINDS) of the NIH, 

which provides recommendations for the conduct and 

support of epidemiological and fundamental research on 

neurological diseases. 

Western Neuroradiological Society 

(WNRS) (The Gabriel H. Wilson Award) 

“Are There MR Imaging Features 

That Predict Regression of Lumbar 

Disk Herniation?” 


University of Arizona 

Tucson, AZ 

The Neuroradiology Education and Research (NER) Foundation Award for 

Outstanding Contributions in Research 

In studies designed by Dr. Moody and funded by his 

grants, he and members of his team have been 

instrumental in furthering our understanding of the cause 

(lipid microemboli during cardiopulmonary bypass) of 

brain complications related to heart surgery and of 

strategies for its prevention. He published a 

comprehensive study of the normal human cerebral 

microvascular pattern. Contrary to prevailing opinion at 

the time, Moody demonstrated that germinal matrix 

hemorrhage in premature neonates results from rupture of 

veins rather than arterioles/capillaries; Charcôt-Bouchard 

microaneurysms are very rare and are not the etiology of 

most spontaneous brain hemorrhages; precapillary 

arteriolar-venular anastomoses (thoroughfare channels), a 

germinal matrix vascular rete, or intraparenchymal 

arteriole-to-arteriole shunts are not normally present in 

humans from the limits of viability in pre-term neonates to 

adults; the pathological substrate for the radiological 

finding of leukoaraiosis is apoptosis-induced 

oligodendrocyte cell death associated with capillary dropout 

and occasionally stenosis of the deep cerebral veins 

(periventricular venous collagenosis); string vessels, 

probably degenerating capillary/arterioles, are found in 

significantly increased numbers in subjects with 

Alzheimer’s disease – reinforcing the theory that a vascular 

component may contribute to this degenerative 

neurological disease. 

The 2004 Neuroradiology Education and Research 

(NER) Foundation Award for Outstanding Contributions 

in Research award recipient was: 


XXIX 




XXX 


2005-2006 Berlex/NER Foundation Fellowship in Basic Science Research Award 

This fellowship, first awarded in 1986, was created by 

the ASNR with the support of Berlex Laboratories to 

stimulate the scientific development of promising 

young men and women, and to aid them in embarking 

on a career in academic radiology. It is specifically 

designed to provide educational opportunities for 

young radiologists who are not yet professionally 

established in the radiologic sciences to gain further 

insight into scientific investigation, and to develop 

competence in research. These fellowships are jointly 

sponsored by Berlex Laboratories, Inc. and the 

Neuroradiology Education and Research (NER) 

Foundation of the American Society of Neuroradiology. 

1986-87 

Jeremy B. Rubin, MD, 


“New Methods Using MRI to A se s Ventricular Shunt 

Function and Measure Intravenous Pre sure Non-invasively 

in Patients with Ventricular Shunt Catheters” 

1987-88 

No Award 

1988-89 

Apichai Jarenwattananon, MD, 

University of Wisconsin Medical Center 

“In-Vivo Sodium MRI (Na-MRI) in Canine 

Model of Status Epilepticus” 

Warren A. Stringer, MD, 

Loma Linda University Medical Center 

“Evaluation of the Relationships Between Cerebral 

Perfusion, Ventilation, and Intracranial Pressure by 

Xenon-enhanced Computed Tomography in Children 

with Cerebral Edema” 

1989-90 

Todd Lempert, MD, 

University of California at San Francisco 

“Evaluation of the Healing Response to Thrombogenic 

Coil Occlusion of Experimental Aneurysms” 

1990-91 

Lori L. Baker, MD, 


“Evaluation of MR Diffusion Imaging Versus Magnetic 

Susceptibility Enhanced Mapping of Perfusion Pool 

in Regional Cerebral Ischemia” 

The recipients of the 2005-06 fellowships are: 

Srinivasan Mukundan, Jr., PhD, MD 


“Toward the Development of a Nanoscale, Target- 

Specific Liposomal Platform Technology for Computed 

Tomography Based Molecular Imaging” 

Max Wintermark, MD 

University of California, San Francisco 

“Morphometric and Functional Characterization of 

Atherosclerotic Carotid Disease by Multidetector-Row 

CT-Angiography: A Comparative Study with Ex Vivo 

Histology and Imaging” 

Past Berlex/NER Foundation Fellowship in Basic Science Research Award 

1990-91 (continued) 

Lee H. Monsein, MD, 

The Johns Hopkins University School of Medicine 

“Primate Model of Reversible Regional 

Cerebral Ischemia” 

1991-92 

Steven N. Breiter, MD, The Johns Hopkins Hospital 

“Proton MRS in the Determination of Lactic Acid 

Concentration in Seizures, Both Human and Animal” 

Frank J. Lexa, VII, MD, University of Pennsylvania 

“MRI Demonstration of Axonal Transport 

in the Mammalian CNS” 

1992-93 

Michael A. Kraut, MD, PhD, 

The Johns Hopkins Hospital 

“Lactate Production and Metabolism 

in Cerebral Activation” 

Brian W. Chong, MD, 

University of California at San Diego 

“A Search for Hidden MRI Flow Patterns 

in Human Cranial Vessels” 

1993-94 

Thomas E. Conturo, MD, PhD, 

The Johns Hopkins Hospital and 

Johns Hopkins University 

“Mechanisms of the Phase Enhancement Effects of 

Bolus-Injected Paramagnetic Contrast Agents and 

Applications in Quantitative Cerebral Blood Volume 

and Flow Imaging” 

John P. Karis, MD, Barrow Neurological Institute 

“Epilepsy Localization: Advanced High Resolution 

MRI-PET FDG Correlation”


Past Berlex/NER Foundation Fellowship in Basic Science Research Award (Continued) 

1994-95 

Jerry Burke, MD, Bowman Gray School of Medicine 

“Serial Positron Emission Tomography and Functional 

MR Imaging of Stroke” 

Robert Fulbright, MD, 


“Functional MR Imaging of the Spine” 

1995-96 

Norman J. Beauchamp, MD, 


“The Natural History of ‘Areas of Risk of Infarction’ as 

Defined by Perfusion MRI and MR Spectroscopy” 

Anthony Masaryk, MD, 

University of Wisconsin-Madison 

“Analysis of Aneurysm Hemodynamics Using 

MRI/MRA Morphology and Flow Measurements 

Correlated with Hemodynamic Numerical Analysis 

and Simulation” 

1996-97 

Joseph T. Lurito, MD, PhD, 


“Functional MRI and Electrophysiologic Correlates of 

Sub-modality Specific Somatosensory Activation” 

Jeffrey L. Sunshine, MD, 


“Early Identification of Ischemic Penumbra by Diffusion 

and Perfusion MR in Acute Stroke” 

1997-98 

Huy M. Do, MD, 

University of Virginia Health Sciences Center 

“The Neuroprotective Effect of Intraarterial 

Nerve Growth Factor (HGF) in a Rabbit Embolic 

Stroke Model” 

1998-99 

William F. Marx, MD, University of Virginia 

“Endovascular Treatment of Experimental Aneurysms 

Using Biologically Modified Embolic Coils: 

Promotion of Permanent Occlusion via 

Intra-aneurysmal Fibroblast Delivery” 

1999-00 

Kevin R. Moore, MD, 

University of Utah Center for Advanced Medical 

Technology 

“Meg-Constrained High-Resolution Surface-Coil MR 

Imaging and MR Spectroscopy for Evaluating 

Medically Refractory Epilepsy” 

John G. Short, MD, University of Virginia 

“Induction of Spinal Interbody Fusion Using Gene 

Therapy Tissue Engineering Techniques” 

2000-01 

John Port, MD, PhD, 

The Johns Hopkins Medical Institution 

“Imaging Selective Attention Mechanisms” 

Eric Schwartz, MD, 


“Diffusion-based MR Imaging in a Rat Spinal Cord 

Following Injury and Transplantation” 

2001-02 

Pratik Mukherjee, MD, PhD, 

Mallinckrodt Institute of Radiology, 

Washington University School of Medicine 

“Comparison of Magnetic Resonance Imaging and 

Positron Emission Tomography in the Study of 

Cerebral Hemodynamics” 

2002-03 

John G. Dalle, DO, 

University of Utah School of Medicine 

“Polymer-Chelate Conjugates for Diagnostic 

Cancer Imaging” 

Christopher Lascola, MD, PhD, 


“Magnetic Resonance Imaging of Spreading 

Depression-Induced Reactive Gliosis in Mice” 

2003-04 

Dheeraj Gandhi, MD, 

University of Michigan Health System 

“Can the Choline/Creatine Ratio Predict Early 

Treatment Response of Head and Neck Squamous 

Cell Carcinma Treated with Radiation Therapy in an 

Animal Model: A Prospective Study” 

Susan M. Kealey, MD, 


“Correlation of MR Permeability Measurements 

with Histologic Markers of Angiogenesis in Rodent 

High-Grade Brain Tumors Before and After Treatment 

with Antiangiogenesis Agent PTK 787” 

2004-05 

Tuong Huu Le, MD, PhD 


“Structural and Functional Correlates of Axonal 

Shearing in Traumatic Brain Injury: A Combined 

DTI, fMRI and MSI Study” 

Whitney B. Pope, MD, PhD 

David Geffen School of Medicine at University 

of California, Los Angeles 

“Identification of Unstable Atheroscelerotic Plaque at 

the Carotid Bisfurcation Using High-Resolution CT- 

PET Imaging: Correlation to Histopathology and 

Patient Symptoms” 

XXXI 




XXXII 



Neuroradiology Education and Research (NER) Foundation Scholar Award in 

Neuroradiology Research* 

Since 1995, the NER Foundation has been in the 

process of raising funds to support neuroradiology 

research. This is one of the most important goals of 

the NER Foundation, and of the ASNR as the 

premier organization for neuroradiology. This award 

was created for young investigators in the early 

stages of their careers, to enhance their competency 

in areas important to the future of neuroradiology, 

including health services research, physiological 

imaging and interventional neuroradiology. It also 

affords the Foundation the opportunity to begin to 

develop leadership in these areas. 

1999 

L. Santiago Medina, MD, MPH 

Children’s Hospital Medical Center, Cincinnati, OH 

“The Role and Cost-Effectiveness of Imaging in 

Newborns with Suspected Occult Spinal Dysraphism” 

2000 

Melanie B. Fukui, MD 

University of Pittsburgh Medical Center, Pittsburgh, PA 

“Carotid Stenosis Evaluation: Cost-Effectiveness of 

Computed Tomographic Angiography vs. Magnetic 

Resonance Angiography” 

2001 

Soonmee Cha, MD 

New York University Medical Center, New York, NY 

“Dynamic Contrast Enhanced T2*-weighted MRI and 

Histopathological Assessment of Experimental Glioma” 

2002 

James D. Eastwood, MD 

Duke University Medical Center, Durham, NC 

“CT Perfusion Imaging in Subarachnoid 

Hemorrhage Related Vasospasm” 

2005 NER Foundation/Boston Scientific Fellowship in Cerebrovascular Disease Research 

Established in 2002, this fellowship expanded 

eligibility to allow both neuroradiology fellows and 

all faculty at the Assistant Professor level to apply. 

It was created to provide an opportunity for a young 

neuroradiologist to pursue research in a topic that 

will advance the diagnosis and treatment of 

cerebrovascular disease, and is supported by 

Boston Scientific. 

The recipients of the 2005 scholar award are: 

Donna R. Roberts, MD 


“The Assessment of Image-guided Transcranial 

Magnetic Stimulation as an Adjuvant to Extradural 

Cortical Stimulation for the Treatment of Chronic 

Facial Pain” 

Steven G. Imbesi, MD* 

University of California, San Diego Medical Center 

“Alteration of Intracranial Aneurysm Flow Dynamics: 

Development and Evaluation of Potential 

Neurointerventional Endovascular Treatment 

Regimens for Wide Necked Aneurysms” 

*recipient awarded second year of funding 

Past NER Foundation Scholar Award in Neuroradiology Research Recipients 

2003 

Steven G. Imbesi, MD 

University of California, San Diego Medical Center, 

San Diego, CA 

“Alteration of Intracranial Aneurysm Flow Dynamics: 

Development and Evaluation of Potential 

Neurointerventional Endovascular Treatment 

Regimens of Wide Necked Aneurysms” 

2004 

Pratik Mukherjee, MD, PhD 

University of California San Francisco, 

San Franciso, CA 

“Diffusion Tensor MR Imaging and Quantitative 

Tractography of Brain Development in 

Premature Newborns” 

The recipient of the 2005 fellowship is: 

Donna Hoghooghi, MD 


“Extent and Effectiveness of Embolization and 

Determination of Vascular Supply of Meningiomas 

Using a Combined Interventional X-ray/MR 

Fluoroscopy Suite”



Neuroradiology Education and Research (NER) Foundation Outcomes 

Research Grant Related to Neuroradiologic Imaging 

This newly created grant is targeted to the 

characterization of brain tumors and specifically, the 

differentiation of neoplastic from nonneoplastic 

condition, effect of MRS on need for biopsy or the 

election of a biopsy site, and evaluation of MRS in 

radiation necrosis. 

2005 ASNR Cornelius G. Dyke Memorial Award 

No award is being given in 2005. 

Past ASNR Cornelius G. Dyke Memorial Award Recipients 

1972 

George M. McCord, MD 

“The Venous Drainage to The Inferior Sagittal Sinus” 

1973 

Barton Lane, MD 

“Cerebrospinal Fluid Pulsations at Myelography: 

A Video-Densitometric Study” 

1974 

Jacques Theron, MD 

“Anatomical-Radiological Correlates of the Anterior 

Choroidal Artery” 

1975 

Thomas P. Naidich, MD 

“The Normal Anterior Inferior Cerebellar Artery” 

1976 

No Award 

1977 

Burton P. Drayer, MD 

“The Capacity for CT Diagnosis of Cerebral Infarction. 

An Experimental Study in the Non-Human Primate” 

The recipient of the 2005 grant is: 

William Hollingsworth, PhD 


“Systematic Literature Review of Magnetic Resonance 

Spectroscopy (MRS) of the Characterization of 

Brain Tumors” 

1978 

Joseph A. Horton, MD 

“The Grain in the Stone: A Computer Search 

for Hidden CT Patterns” 

1979 

Dieter R. Enzmann, MD 

“Experimental Brain Abscess Evolution Studied with 

the CT Scan and Neuropathological Correlation” 

1980 

No Award 

1981 

A. Ronald Cowley, MD 

“The Influence of Fiber Tracts on the CT 

Appearance of Cerebral Edema: An 

Anatomical Pathological Correlation” 

1982 

B. Ludwig, MD 

“Postmortem CT and Autopsy in Perinatal 

Intracranial Hemorrhage” 

XXXIII 




XXXIV 


Past ASNR Cornelius G. Dyke Memorial Award Recipients (continued) 

1983 

No Award 

1984 

Val M. Runge, MD 

“Contrast Enhanced Magnetic Resonance 

Evaluation of a Brain Abscess Model” 

1985 

No Award 

1986 

Jeremy B. Rubin, MD 

“Part 1 Imaging Spinal CSF Pulsation by 

2DFT Magnetic Resonance: Significance 

During Clinical Imaging” 

“Part 2 Harmonic Modulation of Proton MR 

Precessional Phase by Pulsatile Motion Origin 

of Spinal CSF Flow Phenomenon” 

1987 

No Award 

1988 

Vincent P. Mathews, MD 

“Gadolinium Enhanced MR Imaging of 

Experimental Bacterial Meningitis: Evaluation 

and Comparison of CT” 

1989 

Allen D. Elster, MD 

“Europium-DTPA: Development and Testing 

of a Gadolinium Analogue Traceable by 

Fluorescence Microscopy” 

1990 

Marvin D. Nelson, Jr., MD 

“The Search for Human Telencephalic 

Ventriculofungal Arteries” 

1991 

Udo P. Schmiedl, MD 

“Quantitation of Pathological Blood-Brain Barrier 

Permeability in an Astrocytic Glioma using Contrast 

Enhanced MR” 

1992 

R. Gilberto Gonzalez, MD 

“Quantitative In Vivo Human Brain Lithium Magnetic 

Resonance Spectroscopy” 

Frank J. Lexa, VII, MD 

“Wallerian Degeneration in the Feline Visual System: 

Characterization by Magnetization Transfer Rate 

with Histopathologic Correlation” 

1993 

Marc Jouandet, MD 

“Mapping the Human Cerebral Cortex 

with Brainprints” 

1994 


“Functional Magnetic Resonance Imaging of Brain 

Activity and Perfusion in Patients with Chronic 

Cortical Stroke A” 

1995 


“Magnetization Transfer or Spin-Lock? An 

Investigation of Off-Resonance Saturation Pulse 

Imaging Using Varying Frequency Offsets” 

1996 

John C. Strainer, MD 

“fMRI of Primary Auditory Cortex: An Analysis of 

Pure Tone Activation and Tone Discrimination” 

1997 

Stephen G. Imbesi, MD 

“Why Do Ulcerated Atherosclerotic Caroid Artery 

Plaques Embolize? A Flow Dynamics Study” 

David F. Kallmes, MD 

“Guglielmi Detachable Coil Embolization for 

Unruptured Aneurysms in Neurosurgical 

Candidates: A Cost Effectiveness Exploration” 

1998 

No Award 

1999 

Aquilla S. Turk, DO 

“Definition of Aneurysm Ostium (Neck) and 

Morphology Using Intravascular Ultrasound: 

An Experimental Study in Canines” 

2000 

William F. Marx, MD 

“Endovascular Treatment of Experimental Aneurysms 

Using Biologically Modified Embolic Devices: Coil- 

Mediated Intra-Aneurysmal Delivery of Fibroblast 

Tissue Allografts” 

2001 

No Award 

2002 

Mehmet Kocak, MD 

“Functional MR Imaging of the Motor Homunculus: 

Towards Optimizing Paradigms for Clinical Scenarios” 

2003 

No Award 

2004 


“Apparent Diffusion Coefficients Within Spinal Cord 

Transplants and Surrounding White Matter Correlate 

With Degree of Axonal Dieback Following Injury”

CME 

SCIENTIFIC PROGRAM AND MEETING EVALUATION 

Continuing Medical Education (CME) 

The 2005 Continuing Medical Education (CME) 

Pavilion allows online recording of CME credits via the 

Internet. The improvements have created a faster and 

more user-friendly system for evaluating sessions and 

speakers and recording CME hours electronically. 

The CME Pavilion is easily accessible in Room 104A 

(Level 100). Please complete the evaluations for each 

session to assist in planning future meetings and to help 

us maintain accreditation of future programs. 

CME PAVILION 

To access the CME evaluation program, run the 

“ExpoCard” included in your registration packet 

through the card reader at one of the terminals and 

follow the simple directions for selecting and 

evaluating the sessions you have attended. The 

CME credit hours awarded to a session will 

automatically be recorded in your record when the 

evaluation for a session is completed. Evaluations 

can be completed at the end of a session, during 

breaks, at the end of the day or the end of the week. 

You will be able to view a record of the sessions you 

have evaluated and the number of CME credit hours 

earned throughout the program. It will also be 

possible to print your certificate and transcript to 

take home with you. 

Please Note: To receive CME credit for sessions 

attended at the NER Foundation Symposium 2005 

and ASNR 43rd Annual Meeting, all evaluations must 

be entered by the end of the meeting. The CME 

Pavilion replaces the CME booklet of previous years 

and is the only method available for receiving your 

CME credit. 

ACCREDITATION 

Accreditation Statement 

AUDIENCE 

TAKE YOUR OFFICIAL CONTINUING MEDICAL EDUCATION 

(CME) CERTIFICATE HOME WITH YOU! 

An enhancement of the Continuing Medical Education 

online evaluation system allows for attendees to print 

out their official CME certificate for the number of 

hours claimed during the NER Foundation Symposium 

and ASNR 43rd Annual Meeting and take it with them 

when they leave. Go to any terminal in the CME 

Pavilion and follow the simple directions for printing 

out an official NER Foundation Symposium 2005 and 

ASNR 43rd Annual Meeting CME Certificate. 

Following the meeting, the ASNR 2005 CME 

Certificate site will be available online for 90 days for 

attendees to print out their CME certificates. 

Please Note: Due to the availability of CME Certificates 

online, certificates will not be mailed to attendees. 

LETTER OF ATTENDANCE 

If you wish to obtain a Letter of Attendance, 

please request one at the Registration Desk 

located in the North Building Lobby (Level 200) 

of the Metro Toronto Convention Centre. 

The American Society of Neuroradiology is accredited by the Accreditation Council for Continuing Medical 

Education (ACCME) to provide continuing medical education for physicians. 

The American Society of Neuroradiology designates this educational activity for a maximum of 34 category 

1 credits towards the American Medical Association Physician’s Recognition Award. Each physician should 

claim only those hours of credits that he/she spent in the educational activity. 

Target Audience 

The ASNR 43rd Annual Meeting is designed specifically for the practicing general neuroradiologist who wishes 

to integrate advanced imaging such as magnetic resonance spectroscopy, diffusion and perfusion imaging, and 

functional magnetic resonance imaging into his/her daily practice. Programming is also focused toward the 

neuroradiologist who seeks to better understand modern imaging techniques applied to a practice which includes 

adults or children, disorders of the spine, head and neck disease, and neurovascular intervention. 

XXXV 




XXXVI 

EDUCATIONAL OBJECTIVES 

ASNR 43rd Annual Meeting Educational Objectives 

At the conclusion of this meeting, participants will be able to: 

General and Advanced Imaging Programming 

• Describe the current status of neuroradiological imaging of 

the head, neck, spine and interventional imaging in both 

adult and pediatric populations. 

• Apply state of the art techniques in diffusion imaging, 

perfusion imaging, MRS, parallel and high field MR in 

the treatment of both adult and pediatric populations. 

• Identify a rational imaging approach to stroke, central 

nervous system infection, and demyelinating disease. 

• Describe the differences between 1.5T and 3T MR 

imaging, with emphasis on image artifacts and how 

to improve signal-to-noise ratios. 

• Discuss MR safety issues, and re-evaluate the relative 

contraindications to MR imaging. 

Head and Neck Programming 

• Identify common imaging presentations of temporal bone 

lesions, including inflammatory processes involving the 

tympanic cavity, facial palsy, and tinnitus. 

• Identify on sectional imaging studies the major 

anatomic landmarks separating the superficial and 

deeper spaces of the face, and the course of major 

neural structures as they traverse the soft tissues 

of the neck and cervicothoracic junction. 

• Differentiate the imaging appearance of motor denervation 

from other pathologic processes involving the head and neck. 

• Interpret, in a systematic fashion, sectional-imaging studies 

performed on patients with proptosis, and discuss the 

evaluation of orbital lesions such as neoplasms and trauma. 

• List criteria by which a differential diagnosis of mandibular 

and major salivary gland lesions can be created. 

• Discuss the prognostic implications of differentiating highgrade 

from low-grade sarcomas, and list imaging findings 

that significantly impact the surgical and nonsurgical 

management of patients with sarcomas of head and neck. 

Interventional Programming 

• Describe the management strategies for treatment 

of cervical carotid artery stenosis. 

• Review characteristics and pathology associated with 

vascular and lymphatic malformations, and treatment 

options for both low-flow and hi-flow vascular malformations. 

• Describe the management strategies for treatment of 

patients with intracranial atherosclerotic stenosis. 

• Identify the processes which underlie reimbursement for 

both hospitals and physicians and the method and rationale 

of current reimbursement initiatives. 

• Examine the RUC survey process and its importance. 

• Review endovascular treatment, appropriate patient 

selection, relative benefits of each therapy and expected 

clinical outcomes for head and neck neoplasms. 

Functional Programming 

• Describe the use of pre-operative fMRI data in neurosurgical 

treatment planning. 

• Identify the common neurodegenerative disorders in the 

adult patient and review neuroradiologic imaging findings. 

• Review the role of angiogenesis in brain tumors. 

• Describe the advances in molecular imaging as they relate 

to neuroradiology. 

Pediatric Programming 

• Demonstrate how children with intractable seizure 

disorders are evaluated using advanced imaging 

techniques and monitoring. 

• Review and illustrate the various surgical methods currently 

being used to treat children with intractable seizures. 

• Discuss outcomes for the various procedures currently used 

to treat children with seizure disorders. 

• Using case-based methods, present imaging studies of both 

common and rare pediatric neurological abnormalities. 

• Review the history and pathology of focal white matter 

necrosis in children. 

• Review the studies of causation of pediatric lesions. 

• Describe the epidemiology of periventricular leukomalacia 

in premature and term infants. 

• Illustrate the appearance and distribution of pediatric lesions 

using advanced imaging techniques. 

Spine Programming 

• Evaluate the use of higher field strength and other advanced 

imaging techniques in the diagnosis of spinal problems 

in the adult population. 

• Discuss recent advances in multi-detector CT imaging 

of the spine. 

• Evaluate interventional procedures for use in management 

of the disorders and diseases in the adult spine, especially 

as they relate to pain management. 

• Review the imaging findings of common spinal diseases 

in an orderly concise manner. 

Electronic Learning Center (ELC) and National Library 

of Medicine (NLM) Programming 

• Demonstrate the basic knowledge of the use of the computer 

in the practice of neuroradiology with reference to hardware, 

operating system and peripherals for both the Macintosh 

and PC platforms. 

• Identify and demonstrate use of Internet resources 

including PubMed ® , MedlinePlus, other search engines, 

and literature search sites.

Electronic ELC Learning WORKSHOPS 

Center 2005 Workshops 

Gregory L. Katzman, MD, Chair, Program 

Hervey D. Segall, MD, Chair Emeritus 

ELC Workshops 

Electronic Learning Center (ELC) workshops provide the 

opportunity for participants to learn new electronic methods in 

an interactive small group environment. The workshop format 

allows for hands-on and experiential learning with computers, 

software, and knowledgeable assistants. Attendance is limited 

to 50 participants. 

NOTE: 2 participants per computer 

The 2005 ELC workshops and lectures will update 

attendees on the use of electronic methods useful for the 

practicing neuroradiologist and neuroradiologist-educator. 

This year’s program will build on the sessions offered at the 

2004 meeting. 

Workshop registrants will receive a copy of the ELC 2005 

Syllabus, an invaluable CD-ROM designed to complement the 

ELC program. 

The faculty and assisting moderators at the workshops include 

both PC and Mac users. 

PLEASE NOTE: There is an additional registration fee per 

workshop of $50 Member/Non member/Other Professional*; 

$10 Fellow/Trainee*. 

* Letter from Neuroradiology Fellowship Program Director 

confirming status is required. A letter from place of employment 

to confirm position is required for Other Professionals. 

NOTE: All ELC Workshops will be held in Room 103 (Level 100). 

All ELC ticket holders must be present in the room 5 minutes 

prior to the start of the session, or your ticket will be resold. 

ELC Workshop A: Introductory Powerpoint 

Monday, May 23 ■ 8:05am – 9:35am 

Tuesday, May 24 ■ 8:00am – 9:30am 

Wednesday, May 25 ■ 3:00pm – 4:30pm 

David S. Martin, MD (Tues/Wed) 

John L. Go, MD (Mon/Tues) 

The goal of this workshop is to instruct registrants in the creation 

of educational presentations by learning the core concepts of 

Microsoft PowerPoint software. Learn how lecture material can be 

developed for display using LCD projectors. Learning Objectives of 

this workshop include: creating a new presentation from scratch; 

using the Office and Presentation Assistants; copying, deleting, 

and modifying the sequence of slides; formatting and editing the 

text in slides; working with Clip Art, pictures, and other objects; 

preparing an entire presentation; saving a presentation in normal 

and HTML formats; and printing audience and speaker notes. 

ELC Workshop B: Advanced PowerPoint 






This workshop will address the more advanced techniques of 

PowerPoint presentation construction, including insertion of 

graphs or tables, linking of objects, transitions, animations, 

Richard H. Wiggins, III, MD, Vice Chair, Syllabus Editor 

Richard M. Berger, MD, Vice Chair, Moderator & Technical Coordinator 

and adding sound or video clips. Learn how to create more 

dynamic presentations by making items appear and disappear 

on slides and inserting "hidden" controls in the slides. The 

lecturers will demonstrate how to link one PowerPoint 

presentation to another to control the flow of information and 

enable toggling between presentations. Lastly, comments will 

be made on the appropriate use of multimedia components for 

keeping a talk interesting yet avoiding audience distraction by 

too much flash. 

ELC Workshop C: Website Creation for the Novice 

Monday, May 23 ■ 1:00pm – 2:30pm 

Tuesday, May 24 ■ 3:00pm – 4:30pm 


This course is geared toward anyone who is beginning to think 

about creating a website for work or personal use. Through a 

series of practical demonstrations and hands-on exercises, this 

workshop will help you acquire the skills necessary to build and 

maintain a basic website. Topics to be covered include: how a 

standard web editor works, how to create a basic web page 

using images and text, creating links between web pages, 

formatting text and paragraphs, creating and changing the 

layout of tables, and putting your pages on the web. 

ELC Workshop D: Adobe Photoshop and Elements 


Thursday, May 26 ■ 8:00am – 9:30am 


This workshop will enable the attendee to become familiar 

with the more advanced graphics editing techniques and 

options available in Adobe "Photoshop" and its newer 

younger brother "Elements". This hands-on, interactive 

workshop will provide participants with the opportunity to 

learn how to edit images from any origin. Topics covered will 

include determining optimal image size and resolution for 

print, PowerPoint, and email graphics; adding text and arrow 

annotations; re-windowing and leveling; cropping and 

removing extraneous text and markings; and converting to 

gray-scale. Attendees will also use the more common 

graphics tools such as airbrush, blur, rubber stamp, 

eyedropper, magic wand, paint bucket, and others. 

ELC Workshop E: Advanced Website Production 


Dale A. Charletta, MD 

This hands-on expert experience will cover the effective use of, 

and creating, advanced content for the internet. The session will 

cover various editors available for creating web pages in HTML 

(HyperText Markup Language) as well as how to insert XML 

(eXtensible Markup Language). Familiar tools, such as Microsoft 

Word and Netscape Communicator, will be discussed with a 

focus on the use of images, links, tables, and uploading content 

via FTP (file transfer protocol) to a web server. Advanced topics 

such as DHTML (dynamic HTML), CSS (cascading style sheets), 

Javascript, Java, Perl, and other languages used for web page 

creation will also be discussed. 

XXXVII 




XXXVIII 

ELC LECTURES 

Electronic Learning Center 2005 Lectures 

The ELC Lectures were introduced in 2001 in response to the 

overwhelming demand for computer education beyond what 

could be accommodated in small workshops. The ELC Lectures 

focus on topics that are of general interest and in which the 

hands-on experience is not as crucial to the learning process. 

• ELC lectures are held throughout the Annual Meeting 

in a large session room 

• ELC lectures are included in the Annual Meeting 

registration fee 

• No preregistration or ticket is required for admission 

ELC Lecture A: The Radiologist’s Computer: 

PC and Mac Perspectives 


Hervey D. Segall, MD | Gregory L. Katzman, MD 

Up-to-date basics of computer hardware and operating systems 

will be covered, including both the Macintosh and PC 

platforms. Components inside the computer (e.g. motherboard, 

processor, chipsets, drives, cards, etc.), peripherals (e.g. 

monitors, speakers, keyboards, etc.), and lastly connections 

will be discussed. 

ELC Lecture B: High Speed Connectivity Update 


Gerard J. Muro, MD 

Attendees will learn about the different types of highbandwidth 

internet access as well as the advantages and 

limitations of each. Basic principles in connecting a highbandwidth 

internet connection to a home LAN (Local Area 

Network) will also be presented, with a discussion of 

networking for both work and home. 

ELC Lecture C: Capturing and Using Image Files 


Barton F. Branstetter, VI, MD 

The audience will get an appreciation for some of the 

technical details of digital images, requirements for digital 

presentations, methods for obtaining digital images, and 

tricks for making PowerPoint lectures. Recommendations 

will also be made for image storage and organization. 

ELC Lecture D: Multimedia Conference Room 2005 


Venkata Natarajan, PhD 

This lecture will outline the creation of a modern digital 

multimedia conference room to facilitate more efficient 

presentation of radiology learning materials. Topics will 

include video requirements (projection devices, cameras, 

screens, and monitors) audio systems, teaching aids (laser 

pointers, lecterns, and interactive screens), presentation 

devices (computers, VCR/DVD players, slide projectors, 

overhead projectors, PACS and Workstations), network 

considerations (LAN and WAN), and overall ergonomics 

(lighting, soundproofing, seating, and room architecture). 

ELC Lecture E: Advanced Image Processing in MRI 


Todd B. Parrish, PhD 

Neuroradiology relies on sophisticated imaging equipment 

and software to generate images of the patient. The 

purpose of this talk is to start with the basics of imageprocessing 

and build an understanding as it relates to 

neuroimaging methods. The talk will cover issues such as 

image dimensionality, filtering, 3D imaging including surface 

rendering and maximum intensity projections, segmentation, 

normalization, and applications of image processing. 

Advanced topics such as diffusion tensor imaging and 

perfusion (DSC and ASL) processing will be covered. 

ELC Lecture F: fMRI Post-Processing 



This lecture will begin with an overview of the key processes of 

analysis for fMRI data. This includes re-alignment, modeling of 

hemodynamic response, hypothesis testing and statistical 

parametric mapping. We will discuss the characteristics of block 

vs. event-related designs. Lastly, comparison will be made of the 

commonly used data processing packages, including Stimulate, 

AFNI and SPM. 

ELC Lecture G: PDA’s and the Radiologist 



Attendees will obtain the current understanding of PDA 

hardware features, expansion interfaces, and the differences 

between the various operating systems. The technological 

future for PDA’s will be discussed in light of the rapid evolution 

of these types of products. 

ELC Lecture H: Digital Teaching Files: 

An Overview of Techniques 



This lecture is directed at the novice, who may be considering 

constructing electronic digital teaching files on his or her 

desktop computer, either at work, on a laptop, or at home. 

Several simple methodologies will be presented, all of which 

are easily learned and inexpensive. Department-wide solutions 

and networks will not be discussed.

NLM • SEMINARS 

Business Center • Grant Writing Seminar 

NLM Workshops 

A showcase by the National Library of 

Medicine (NLM) will be offering handson-workshops 

focusing on PubMed ® 

and MedlinePlus throughout the ASNR 

Annual Meeting. PubMed ® is the 

National Library of Medicine's webbased 

portal to the MEDLINE database. 

Searchable without charge, PubMed ® provides 

bibliographic access to the health literature, currently 

containing over 15 million citations to articles written over 

the past 50 years. NLM also produces MedlinePlus, a free 

consumer friendly source of up-to-date health information 

for patients. Visit and consult with expert librarians or testdrive 

these and other NLM resources. 

There is no registration fee for these workshops. To register 

for a workshop, please sign up at the ELC/NLM desk 

located outside Room 103 (Level 100). 

WORKSHOPS 

Room 103 (in cooperation with the Electric Learning Center 

(ELC) Committee). 

NLM Workshop: PubMed/MedlinePlus Advanced 

Tips and Tricks 



Linda Milgrom 

Learn to fine tune your PubMed searches and limit your 

retrieval to the most relevant articles. Have you used Cubby, an 

easy way to ask PubMed to send you periodic updates on topics 

of continuing interest? Try special features such as Clipboard, 

Clinical Queries, and History. Come to the demo and play 

Stump the Librarian. Any PubMed question is fair game. 

NLM Workshop: When PubMed Is Not the Answer… 


Thursday, May 26 ■ 1:00pm – 2:00pm 

Maryanne Blake 

PubMed may not always be the right tool for your information 

needs. At this session you will find out how to maximize the 

utility of your time on the Web and make your searches more 

effective and efficient. You’ll get helpful hints about using 

Web resources from a “search all the Web” site like Google to 

health-specific sites like Clinicaltrials.gov and MedlinePlus. 

ASNR GRANT WRITING SEMINAR (Possible Program Cancellation, if registration minimums are not met) 

Practical Tips on Getting Your Grant Funded, a two-part seminar on grant writing, will be presented by Janet S. Rasey, 

PhD., Director of the Research Funding Service at the University of Washington Medical Center in Seattle, Washington and 

a retired Professor of Radiation Oncology. This seminar is specifically designed for young investigators with an interest in 

writing funded research proposals. 

The workshop, presented over 2 days (4 hours each day), will focus on three major areas in the grant writing process, 

Before You Write, Writing the Grant, and The Review Process. Topics covered will include the elements of writing a 

competitive research grant with an emphasis on NIH R01 proposals and an explanation of grant review procedures and 

psychology. The seminar will take place on Monday, May 23 and Tuesday, May 24. Enrollment is limited to 25 

participants. Check at the ASNR registration desk, North Building Lobby (Level 100), for availability on-site. 

NEW 

ASNR BUSINESS CENTER 

Synopsis: The ASNR Business Center will offer executive lectures 

encompassing topics pertinent to management and 

administration for both private practice and academic 

departments. Business Center lectures will be given in two 

blocks, each of which is 90 minutes in duration. Within each 

block, three speakers are scheduled for 20-25 minute didactic 

lectures with 5-10 minutes of question and answer time. 

Goals: Topics will be presented spanning ability levels from 

basic review to specific financial principles. Experienced 

luminary neuroradiologist speakers with significant business 

experience will be recruited from within our own membership. 

Target Audience: Radiologists who make business decisions for 

their practices or any radiologist interested in learning more 

about the mechanisms by which a practice functions. 

Monday, May 23 1:00pm - 2:30pm 

“Welcome and Introduction to Business Issues in Radiology” 


Chief, Department of Radiology 

VA Salt Lake City Health Care System 

“Net Present Value and Finance for Practicing Radiologists” 

Jonathan Breslau, MD 

Radiological Associates of Sacramento 

“Managerial Accounting Applications in Radiology” 

Frank J. Lexa, MD, MBA 

Professor, Wharton Graduate School of Business 

Tuesday, May 24 1:00pm - 2:30pm 

“Financial Perspectives and Concepts Useful in Radiology” 

Dieter Enzmann, MD, MBA 

Chairman, Radiology Department 

University of California, Los Angeles 

“Private Practice vs. Academic Management” 

Steven Stevens, MD 



“Offshore Teleradiology: Bane or Boon?” 



University of California San Diego Healthcare 

Toronto, Canada Canada 

XXXIX



XL 

HOW-TO SESSIONS 

How-To Sessions (As of 3/24/05) 

In addition to the Technical Exhibition, 

the leadership of the ASNR is pleased to 

announce the ninth annual slate of 

instructional How-To forums. These 

sessions, presented in conjunction with 

major corporate contributors, deal with advances in imaging 

and procedures as well as principles in neuroradiology and 

image information management. How-To Luncheon Sessions 

are scheduled Sunday, May 21 through Wednesday, May 25. 

How-To Breakfast Sessions are scheduled Monday, May 23 

through Wednesday, May 25. Again in 2005, How-To Session 

programming will be offered during specific breakfast, lunch 

and reception sessions. 

Monday, May 23..................6:50 am - 7:50 am 

“The New Breakthrough Data on Aneurysm Healing” 

John Perl, MD, Jacques Moret, MD, 

John C. Chaloupka 

Monday, May 23..............11:50 am - 12:50 pm 

“Theory and New Technology for Improving Interventional 

Therapy for Intracranial Atherosclerotic Disease: 

Full Results from Wingspan HDE Safety Study” 

Arani Bose, MD 

Wednesday, May 25 ............6:50 am - 7:50 am 

“CT Assessment of Perfusion in Stroke Patients” 

Lawrence N. Tanenbaum, MD 

“Multihance: How is it the Same, How is it Different?” 


Tuesday, May 24 ................6:00 pm - 7:30 pm 

Thursday, May 26 ................6:50 am - 7:50 am 

Sunday, May 22 ..................5:30 pm - 7:00 pm 

“HydroCoil Clinical Experience: 

Aneurysm Follow-Up and New Evaluation” 

John D. Barr, MD, John Thornton, MD, 

Karel G. TerBrugge, MD, Phil White, MD, 

Yasunari Niimi, MD 

The How-To Sessions offer a unique opportunity for 

neuroradiologists to discuss techniques, procedures, and 

products with their colleagues as well as with technical 

specialists from the imaging industry. Comments and 

suggestions from meeting registrants over the last eight years 

were integrated into this year’s format. 

The sessions vary and include both didactic presentations 

and demonstrations, all with a strong practical emphasis. 

A significant portion of each session is devoted to questions 

and answers. Indications, problems, and solutions relating 

to imaging techniques will be addressed including advances 

in helical CTA, CT perfusion, MDCT applications and MRI of 

the brain. 

Tuesday, May 24..................6:50 am - 7:50 am 

“Intraoperative MR-Guided Neurosurgery at 3T” 

Charles L. Truwit, MD 

Wednesday, May 25 ........11:50 am - 12:50 pm 

“MR Diffusion” 

Christopher G. Filippi, MD 

“Brain Perfusion” 

Max Wintermark, MD 

Sunday, May 22 ..............11:20 am - 12:20 pm 

“Advanced Interventional Neuro Imaging in the 

Angio Suite” 

Richard P. Klucznik, MD 

Tuesday, May 24 ............11:50 am - 12:50 pm 

“Advanced Neuro MRI with Tim” 

Edmond A. Knopp, MD 

Sunday, May 22 ..................6:50 am - 7:50 am 

PLEASE NOTE: Due to the direct financial support 

from these companies and the commercial content, 

CME credit will not be granted for these sessions.

Scientific PROGRAM Program Overview OVERVIEW 

(As of 3/24/05) 

Meals and Breaks: Breakfasts, Morning and Afternoon Coffee Service, and Box Lunches will be provided 

throughout the week. PLEASE NOTE: Annual Meeting food service locations vary throughout week based on 

Technical Exhibit hours and How-to Session programming. 

NOTE: Page numbers referenced throughout the program correspond to the page number within 

the Proceeding Book. 

ASNR 43RD ANNUAL MEETING 


10:00am - 11:45am 

(7) PARALLEL SCIENTIFIC PAPER SESSIONS 

6:30am - 7:55am 

(A) INTERVENTIONAL: ENT Intervention 

BREAKFAST 

and Miscellaneous (Page 4) 

(B) HEAD AND NECK: Temporal Bone, 

6:50am - 7:50am 

Sinonasal (Page 11) 

HOW-TO SESSION BREAKFAST 

(C) INTERVENTIONAL: Aneurysms (Page 17) 

Sponsor: Boston Scientific 

(D) ADULT BRAIN: Vascular, Extracranial 

8:00am - 8:05am 

(Page 25) 

(1) OPENING REMARKS 

10:15am - 11:45am 

Victor M. Haughton, MD, ASNR President 

(8) ELC WORKSHOP B: ADVANCED POWERPOINT 

8:00am - 12:00pm 

Page 31 

(2) ASNR RESEARCH GRANT WRITING SEMINAR: 

PRACTICAL TIPS ON GETTING YOUR GRANT 

FUNDED - PART I 

Page 1 

8:05am- 9:05am 

(3) ELC LECTURE A: THE RADIOLOGIST'S COMPUTER: 

PC & MACINTOSH PERSPECTIVES 

Page 1 

8:05am - 9:35am 

(4) CERVICAL CAROTID ARTERY 

ATHEROSCLEROSIS (ASITN) 

Page 1 

8:05am- 9:35am 

(5) NON-NEOPLASTIC TEMPORAL 

BONE IMAGING (ASHNR) 

Page 3 

8:05am- 9:35am 

(6) ELC WORKSHOP A: INTRODUCTORY POWERPOINT 

Page 4 

9:35am - 9:55am 

MORNING BREAK 

11:45am - 12:50pm 

LUNCH BREAK 

11:50am - 12:50pm 

HOW-TO SESSION LUNCH 

Sponsor: Boston Scientific 

1:00pm - 1:30pm 

(9) ELC LECTURE B: HIGH SPEED 

CONNECTIVITY UPDATE 

Page 32 

1:00pm - 2:30pm 

(10) VASCULAR MALFORMATIONS OF THE 

HEAD AND NECK (ASITN) 

Page 32 

1:00pm - 2:30pm 

(11) NEUROPATHIES OF THE HEAD 

AND NECK (ASHNR) 

Page 34 

1:00pm - 2:30pm 

(12) NEURONEWS PAPER SESSION 

Page 36 

1:00pm - 2:30pm 

(13) ASNR BUSINESS CENTER - PART I 

Page 36 

XLI 




XLII 

PROGRAM OVERVIEW 

Scientific Program Overview (continued) (As of 4/21/05) 

(Monday, May 23 continued) 

1:00pm - 2:30pm 

(14) ELC WORKSHOP C: WEBSITE 

CREATION FOR THE NOVICE 

Page 36 

1:30pm - 2:00pm 

(15) ELC LECTURE C: CAPTURING 

& USING IMAGE FILES 

Page 36 

2:30pm - 2:55pm 

AFTERNOON BREAK 

3:00pm - 4:00pm 

(16) NATIONAL LIBRARY OF MEDICINE (NLM) 

WORKSHOP: PUBMED/MEDLINEPLUS: 

ADVANCED TIPS AND TRICKS 

Page 36 

3:00pm - 4:30pm 


(A) HEAD AND NECK: Neck and Extracranial 

Carotid (Page 36) 

(B) INTERVENTIONAL: Arteriovenous 

Malformations/Fistulae (Page 43) 

(C) HEAD AND NECK: Neck, New 

Techniques, and Excerptas (Page 49) 

(D) ADULT BRAIN: Infarction and 

Vasospasm (Page 55) 

4:40pm - 6:10pm 

(18) ORBITAL IMAGING (ASHNR) 

Page 62 

4:45pm - 6:15pm 

(19) ADVANCED IMAGING SEMINAR- DIFFUSION 

Page 63 

6:00pm - 7:30pm 

WELCOME RECEPTION WITH TECHNICAL EXHIBITORS 


6:30am - 7:55am 

BREAKFAST 

6:50am - 7:50am 


Sponsor: Philips Medical Systems 

8:00am - 12:00pm 

(20) ASNR GRANT WRITING SEMINAR: PRACTICAL 

TIPS ON GETTING YOUR GRANT FUNDED - PART II 

Page 67 

8:00am - 9:00am 

(21) ELC LECTURE D: MULTIMEDIA 

CONFERENCE ROOM 2005 

Page 67 

8:00am - 9:30am 

(22) INTRACRANIAL ATHEROSCLEROTIC 

DISEASE (ASITN) 

Page 67 

8:00am - 9:30am 

(23) FACIAL IMAGING (ASHNR) 

Page 69 

8:00am - 9:30am 

(24) UPDATES AND CONTROVERSIES ON MR SAFETY: 

CAN PATIENTS WITH IMPLANTABLE CARDIAC 

RHYTHM DEVICES SAFELY UNDERGO MAGNETIC 

RESONANCE IMAGING? (General) (ARS) 

Page 72 

8:00am - 9:30am 

(25) ELC WORKSHOP A: INTRODUCTORY POWERPOINT 

Page 72 

9:30am - 9:55am 

MORNING BREAK 

10:00am - 11:45am 


(A) INTERVENTIONAL: Intracranial 

Aneurysms (Page 73) 

(B) INTERVENTIONAL: Stroke Therapy 

and Stenting (Page 79) 

(C) INTERVENTIONAL: Angioplasty, 

Stenting, and New Techniques (Page 87) 

(D) ADULT BRAIN: Neoplasms and 

New Techniques (Page 94)

Scientific PROGRAM Program Overview OVERVIEW 

(continued) (As of 3/24/05) 

(Tuesday, May 24 continued) 

10:15am - 11:45am 


Page 101 

11:45am - 12:55pm 

LUNCH BREAK 

11:50am - 12:50pm 


Sponsor: Siemens Medical Systems 

1:00pm - 2:30pm 

(28) FUNCTIONAL MRI (ASFNR) 

Page 101 

1:00pm - 2:30pm 

(29) FUNDAMENTALS OF MEDICAL REIMBURSEMENT 

FOR DIAGNOSTIC AND THERAPEUTIC 

NEURORADIOLOGY (ASITN) 

Page 102 

1:00pm - 2:45pm 

(30) HEAD AND NECK SARCOMA (ASHNR) 

Page 103 

1:00pm - 2:30pm 

(31) ASNR BUSINESS CENTER - PART II 

Page 106 

1:00pm - 2:30pm 

(32) ELC WORKSHOP D: ADOBE PHOTOSHOP 

AND ELEMENTS 

Page 106 

2:30pm - 2:55pm 


3:00pm - 4:30pm 

(33) PARALLEL SCIENTIFIC PAPER SESSION 

(A) INTERVENTIONAL: Aneurysms and 

Spinal Vascular Malformations (Page 106) 

(B) ADULT BRAIN: Functional Imaging and 

Advanced Techniques (Page 112) 

(C) ADULT BRAIN: Functional and Advanced 

Imaging of Brain Neoplasms (Page 120) 

(D) ADULT BRAIN: Degenerative, Dementias 

and Destructive Lesions (Page 127) 

ARS = Audience Responce System 

3:00pm - 4:30pm 

(34) ELC WORKSHOP C: WEBSITE 

CREATION FOR THE NOVICE 

Page 133 

4:40pm - 5:40pm 

(35) ELC LECTURE E: ADVANCED IMAGING 

PROCESSING IN MRI 

Page 134 

4:40pm - 6:10pm 

(36) INTERVENTIONS FOR HEAD AND 

NECK NEOPLASMS (ASITN) 

Page 134 

4:40pm - 6:10pm 

(37) GENERAL NEURORADIOLOGY SESSION (ARS) 

Page 135 

4:45pm - 6:15pm 

(38) ADVANCED IMAGING SEMINAR - PERFUSION 

Page 135 

5:00pm - 6:00pm 


WORKSHOP: WHEN PUBMED IS NOT THE ANSWER 

Page 137 

6:00pm - 7:30pm 

HOW-TO SESSION RECEPTION 

Sponsor: GE Healthcare 


6:30am - 7:55am 

BREAKFAST 

6:50am - 7:50am 


Sponsor: Bracco Diagnostics Inc. 

8:00am - 9:30am 

(40) SPECIAL SESSION: SURVIVING CHANGE: 

NEURORADIOLOGY PRACTICE 

Page 139 

8:00am - 9:30am 

(41) PERSPECTIVES ON SUBMISSION 

OF GRANT APPLICATIONS 

Page 139 

9:30am - 9:55am 

MORNING BREAK 

XLIII 




XLIV 



(Wednesday, May 25 continued) 

10:00am - 10:15am 

(42) ASNR PRESIDENTIAL ADDRESS 

Victor M. Haughton, MD, ASNR President 

10:15am - 11:05am 

(43) ASNR AWARDS PRESENTATION 

Presentation of 2005 ASNR Gold Medal Award 

Moderators: Victor M. Haughton, MD, 

ASNR President 

Michael Deck, MD, FACR, 

Chair, Gold Medal Award Committee 

Presentation of 2005 ASNR Honorary Member 

Moderator: Norman E. Leeds, MD, 

Chair, Honorary Member Committee 

Announcement of 2004 Outstanding 

Presentation Awards 

Moderator: James G. Smirniotopoulos, MD, 

Chair, Education Committee 

Presentation of the Neuroradiology Education and 

Research (NER) Foundation Award for Outstanding 

Contributions in Research 

Moderator: A. James Barkovich MD, 

Chair, Neuroradiology Education and 

Research (NER) Foundation 

Announcement of 2005 NER Foundation Scholar 

Award in Neuroradiology Research 

Moderator: A. James Barkovich MD, 

Chair Neuroradiology Education and 


Announcement of 2005 NER Foundation Outcome 

Research Grant in Neuroradiologic Imaging 

Announcement of 2005 NER Foundation/Boston 

Scientific Fellowship in Cerebrovascular 

Disease Research 

Announcement of 2005 Berlex/NER Foundation 

Fellowship in Basic Science Research Awards 

Moderator: Howard A. Rowley, MD, 

Chair, Research Committee 

11:05am - 11:10am 

ANNOUNCEMENT OF SYMPOSIUM 

NEURORADIOLOGICUM (SNR) XVII/WORLD 

FEDERATION OF NEURORADIOLOGICAL SOCIETIES 

Michael S. Huckman, MD, WFNRS President 

11:10am - 11:40am 

(44) AMERICAN SOCIETY OF NEURORADIOLOGY (ASNR) 

ANNUAL BUSINESS MEETING (Members only) 

11:40am - 12:55pm 

LUNCH BREAK 

11:50am - 12:50pm 


Sponsor: Philips Medical Systems 

12:00pm - 12:55pm 

(45) AMERICAN SOCIETY OF FUNCTIONAL 

NEURORADIOLOGY (ASFNR) ANNUAL 

BUSINESS MEETING (Members only) 

1:00pm - 2:30pm 

(46) NEURODEGENERATIVE DISORDERS (ASFNR) 

Page 142 

1:00pm - 2:30pm 

(47) EVALUATION AND MANAGEMENT OF 

INTRACTABLE SEIZURES IN CHILDREN (ASPNR) 

Page 143 

1:00pm - 2:30pm 

(48) SPINE CASE-BASED REVIEW (ASSR) (ARS) 

Page 145 

1:00pm - 1:45pm 

(49) SOCIOECONOMIC PAPER SESSION 

Page 145 

1:00pm - 2:30pm 


Page 148 

2:30pm - 2:55pm 


3:00pm - 4:40pm 


(A) PEDIATRICS: Epilepsy and Non-Neoplastic 

Miscellaneous (Page 148) 

(B) ADULT BRAIN: Demyelinating and Miscellaneous 

Non-Neoplastic Lesions (Page 155) 

(C) SPINE: Functional and Degenerative 

(Page 163) 

(D) ADULT BRAIN: New Imaging and 

Post-Processing Techniques (Page 169) 

3:00pm - 4:30pm 

(52) ELC WORKSHOP A: 

INTRODUCTORY POWERPOINT 

Page 176



(Wednesday, May 25 continued) 

4:40pm - 5:40pm 

(53) ELC LECTURE F: fMRI POST-PROCESSING 

Page 176 

4:40pm - 6:10pm 

(54) PEDIATRIC CASE-BASED STUDIES 

(ASPNR) (ARS) 

Page 176 

4:40pm - 6:10pm 

(55) NEURORADIOLOGY CORE 

CURRICULUM (GENERAL) 

Page 177 

4:45pm - 6:15pm 

(56) ADVANCED IMAGING SEMINAR - SPECTROSCOPY 

Page 178 

5:00pm - 6:00pm 


WORKSHOP: PUBMED/MEDLINEPLUS: 

ADVANCED TIPS AND TRICKS 

Page 180 

5:40pm - 6:10pm 

(58) ELC LECTURE G: PDA'S AND THE RADIOLOGIST 

Page 181 

6:15pm - 9:00pm 

EVENING OF COMEDY AND CUISINE RECEPTION 

AND PERFORMANCE 

Thursday, May 26 

6:30am - 7:55am 

BREAKFAST 

Sponsor: GE Healthcare 

6:50am - 7:50am 


8:00am - 9:00am 

(59) ELC LECTURE H: DIGITAL TEACHING FILES: 

OVERVIEW FOR BEGINNERS 

Page 183 

8:00am - 9:30am 

(60) UPDATES ON "PVL" (ASPNR) 

Page 183 

8:00am - 9:30am 

(61) CT VS. MR IN THE SPINE (ASSR) 

Page 185 

8:00am - 9:30am 

(62) BRAIN TUMORS (ASFNR) 

Page 187 

8:00am - 9:30am 

(63) ELC WORKSHOP D: ADOBE PHOTOSHOP 

AND ELEMENTS 

Page 188 

9:30am - 9:55am 

MORNING BREAK 

10:00am - 11:30am 


(A) SPINAL CORD AND PERIPHERAL NERVES: 

Functional and Advanced Imaging Techniques 

(Page 188) 

(B) ADULT BRAIN: Metabolic 


(C) ADULT BRAIN: Trauma and Vascular 

Lesions (Page 202) 

(D) Great Cases and Excerptas (Page 209) 

10:15am - 11:45am 

(65) ELC WORKSHOP E: ADVANCED 

WEBSITE PRODUCTION 

Page 220 

11:45am - 12:55pm 

LUNCH BREAK 

12:00pm - 12:55pm 

(66) AMERICAN SOCIETY OF SPINE RADIOLOGY 

(ASSR) ANNUAL BUSINESS MEETING 

(Members only) 

12:00pm - 12:55pm 

(67) AMERICAN SOCIETY OF PEDIATRIC 

NEURORADIOLOGY (ASPNR) ANNUAL BUSINESS 

MEETING (Members only) 

1:00pm - 2:30pm 

(68) REVIEW OF PEDIATRIC BRAIN IMAGING (ASPNR) 

Page 220 

XLV 




XLVI 

Scientific PROGRAM Program Overview (continued) OVERVIEW 

(As of 3/24/05) 

(Thursday, May 25 continued) 

Friday, May 27 

1:00pm - 2:30pm 

(69) ADVANCED IMAGING IN THE SPINE (ASSR) 

Page 221 

1:00pm - 2:30pm 

(70) EMERGING FUNCTIONAL MODALITIES (ASFNR) 

Page 223 

1:00pm - 2:00pm 


WORKSHOP: WHEN PUBMED IS NOT THE ANSWER 

Page 224 

2:30pm - 2:55pm 


3:00pm - 4:30pm 


(A) ADULT BRAIN: Innovative Techniques 


(B) SPINE: Interventional and Innovative 

Techniques (Page 230) 

(C) PEDIATRICS: Vascular, Trauma and 

Tumors (Page 236) 

(D) PEDIATRICS: Vascular and 

Miscellaneous (Page 244) 

4:40pm - 6:10pm 

(73) DIFFUSION AND OTHER TECHNIQUES IN 

SPINAL MR IMAGING (ASSR) 

Page 251 

4:45pm - 6:15pm 

(74) ADVANCED IMAGING SEMINAR-PARALLEL 

& HIGH FIELD 

Page 252 

6:30am - 7:55am 

BREAKFAST 

8:00am - 9:30am 


(A) SPINE: Trauma, Degenerative and 

Interventional (Page 255) 

(B) PEDIATRICS: Miscellaneous (Page 262) 

(C) ADULT and PEDIATRIC BRAIN: 

Neoplasms (Page 269) 

(D) ADULT BRAIN: Miscellaneous (Page 275) 

9:30am - 9:55am 

MORNING BREAK 

10:00am - 11:30am 

(76) REVIEW OF PEDIATRIC HEAD, NECK, 

AND SPINE (ASPNR) 

Page 282 

10:00am - 11:30am 

(77) INTERVENTIONAL SPINE (ASSR) 

Page 283 

11:30am - 11:45am 

CLOSING REMARKS 

Patricia A. Hudgins, MD, ASNR President

Notes:

Notes:

Monday Morning 

8:00 AM - 8:05 AM 

Theatre 

(1) Opening Remarks 

— Victor M. Haughton, MD, ASNR President 


8:00 AM - 12:00 PM 

Room 201 

(2) ASNR Research Grant Writing 

Seminar: Practical Tips on Getting 

Your Grant Funded - Part I 


8:05 AM - 9:05 AM 

Room 205 

— Janet S. Rasey, PhD 

(3) ELC Lecture A: The Radiologist's 

Computer: PC & Macintosh 

Perspectives 

— Hervey D. Segall, MD 

— Gregory L. Katzman, MD 

1 

NOTE ABOUT SCANNED IMAGES: Scanned images are included in the 

proceedings book. Some submitted images were reduced during the printing process, thereby 

decreasing clarity. The images as originally submitted can be viewed within the abstract on the 

ASNR website at www.asnr.org/2005. 


8:05 AM - 9:35 AM 

Theatre 

(4) Cervical Carotid Artery 

Atherosclerosis (ASITN) 

(5) Imaging the Cervical Carotid Artery 

— Jeffrey L. Sunshine, MD, PhD 

(6) Stroke Risk and Medical Treatment of Cervical 

Carotid Artery Stenosis 

— Helmi Lutsep, MD 

(7) Current Status of Carotid Artery Stenting 

— Gary R. Duckwiler, MD 

(8) Current Status of Endarterectomy 

— Philip E. Stieg, MD, PhD 

Case Management Discussion 

Moderator: Gary R. Duckwiler, MD 

Imaging the Cervical Carotid Artery 


LEARNING OBJECTIVES 

Upon completion of this presentation, participants will be 

able to: 

1) Learn and review current tools to acquire and process 

images of the carotid arteries in the setting of potential 

neuro-intervention. 

PRESENTATION SUMMARY 

We will discuss the conventional tools available to neuroradiologists 

with particular focus on CT and MR modalities. 

We will touch on any current issues in our gold standard of 

catheter angiography. In MR imaging we will review timeof-flight 

acquisitions with some attention to coil choices. 

Then we will look at the addition of contrast in MR imaging 

or CT to the acquisition. Postprocessing techniques will look 

at traditional newer methods of multiplanar interpretations. 

Monday

Monday 

Last we will look at the developing areas to evaluate plaque 

morphology and consequent risks for thromboembolic 

events. Endovascular imaging will be described at least in 

the model systems and trials now available. 

Stroke Risk and Medical Treatment of Cervical Carotid 

Artery Stenosis 


Dr. Lutsep is an Associate Professor in the Department of 

Neurology at Oregon Health & Science University. Dr. 

Lutsep completed her medical school and residency at the 

Mayo Clinic and completed a behavioral neurology fellowship 

at the University of California at Davis as well as a fellowship 

in cerebrovascular disease at Stanford University. 

She is a member of the stroke team and is the Associate 

Director of the Oregon Stroke Center. Dr. Lutsep’s research 

interests involve the development of therapies for stroke prevention, 

acute treatment and stroke recovery, with a focus on 

the use of devices in clinical stroke treatment. She also studies 

stroke imaging and the cognitive effects of focal brain 

lesions. Dr. Lutsep is the author of numerous publications 

and presentations. She serves on the executive board of the 

Western States Stroke Consortium and is a chief editor of the 

eMedicine online neurology textbook. 



able to: 

1) Review the stroke risks in the medical compared to surgical 

groups of trials involving patients with symptomatic 

carotid artery stenosis. 

2) Compare the stroke risks in the medical compared to surgical 

groups of trials involving patients with asymptomatic 

carotid artery stenosis. 

3) Identify characteristics that increase stroke risk in patients 

with carotid artery stenosis. 


The risk of stroke due to cervical carotid artery stenosis is 

well elucidated from randomized trials of medical treatment 

versus endarterectomy in patients with either symptomatic 

or asymptomatic carotid stenosis. The North American 

Symptomatic Carotid Endarterectomy (NASCET) trial 

assessed risk in patients who had had symptoms ipsilateral to 

a carotid artery stenosis. In medically treated patients, a 70% 

or greater stenosis carried a stroke risk of 26% at 2 years 

(only 9% in the surgical group). In medically treated patients 

with near-occlusion and collateral formation, however, risks 

were lower. A 50-69% stenosis carried a stroke risk of 22% 

at 5 years (15.7% in the surgical group). In the 50-69% 

stenosis group, stroke risks in the medical treatment group 

were sufficiently lower in patients who presented with amaurosis 

instead of hemispheric symptoms, and in women compared 

to men, that there was no benefit to surgery compared 

to medical treatment. In all stenosis groups, stroke risks were 

higher in older patients compared to younger ones and benefits 

of surgery compared to medical treatment were greater 

in older patients. Two large trials, the North American 

Asymptomatic Carotid Atherosclerosis Study (ACAS) and 

the European Asymptomatic Carotid Surgery Trial (ACST), 

assessed stroke risks in asymptomatic vessels with approxi- 

2 

mately 60% or greater stenosis. In medically treated patients 

at 5 years, ACAS revealed an 11.0% rate of ipsilateral 

strokes and perioperative period events (5.1% with surgery), 

and ACST showed an 11.8% rate of strokes and perioperative 

period events (6.4% with surgery). Neither ACAS nor 

the larger ACST study suggested that stroke risk increased 

with increasing stenosis in asymptomatic patients. Since 

absolute risk reductions with surgery were lower than in 

symptomatic vessels, the 30-day surgical morbidity needed 

to be less than 3% for intervention to show benefit. The 

ACST trial did not show benefit of surgery for patients older 

than 74 years of age, but did show a modest benefit for 

women as well as stronger benefit for men. In these trials, 

medical treatment consisted of aspirin and risk factor control. 

Other antithrombotics have not been assessed in this 

population, although anticoagulation has been considered for 

symptomatic high-grade stenosis because of its high stroke 

risk. 

Current Status of Carotid Artery Stenting 




able to: 

1) Know the current clinical indications for carotid artery 

stenting (CAS). 

2) Know the current approved devices for carotid artery 

stenting (CAS). 

3) Know the current or anticipated Medicare reimbursement 

guidelines for carotid artery stenting (CAS). 

4) Know the current clinical trials for carotid artery stenting 

(CAS). 


Carotid artery stenting (CAS) is now entering the mainstream 

of treatments for carotid atherosclerosis. With the 

FDA approval of devices, the demand for this procedure will 

increase. At the time of the conference, at least two devices 

will be on the market, with several more either approved or 

in process of approval. With the approval of the protection 

devices, they will likely be used as the standard of care. 

However, the payment guidelines are as of now uncertain. 

Although the Stenting and Angioplasty with Protection in 

Patients at HIgh Risk for Endarterectomy (Sapphire) study 

did indicate a better outcome in “high surgical risk” patients 

having a CAS versus a carotid endarterectomy (CEA), this 

pooled symptomatic and asymptomatic patients. (1) At the 

Food and Drug Administration (FDA) panel meeting regarding 

the Sapphire study, the issue of procedural risk in asymptomatic 

patients was a sore point of discussion. This issue is 

reflected in the recent Centers for Medicare & Medicaid 

Services (CMS) statement dated December 17th, 2004. (2) 

At this time, pending the comment period, CMS does not 

intend pay for CAS in asymptomatic patients outside of an 

ongoing study. Of course, the majority of patients in the 

United States having CEA or CAS are indeed asymptomatic. 

The CMS also has mandated a nationwide training and outcome 

analysis process, which is unprecedented and unspecified 

at present. It remains to be seen how this will affect the 

CAS practice. This issue will likely result in a study to 

resolve the status of CAS in asymptomatic patients. The

Carotid Revascularization Endarterectomy vs Stent Trial 

(CREST) study has applied and likely will begin to enroll 

asymptomatic patients in the randomization phase and help 

in clarifying this issue. At the conference, the latest detail of 

these issues will be discussed. 

REFERENCES 

1. Yadav JS, Wholey MH, Kuntz RE, et al. Stenting and angioplasty 

with protection in patients at high risk for 

endarterectomy investigators. N Engl J Med 

2004;351(15):1493-1501 

Current Status of Endarterectomy 

Philip E. Stieg, MD, PhD 

Dr. Philip E. Stieg is the Professor and Chairman of the 

Department of Neurological Surgery at Weill Medical 

College and Neurosurgeon-in-Chief at New York- 

Presbyterian Hospital. He received his BS degree in 1974 

from the University of Wisconsin at Madison, his PhD in 

Anatomy and Neuroscience from Albany Medical College of 

Union University in 1980, and his MD from the Medical 

College of Wisconsin in 1983. He completed his internship 

and residency at the University of Texas Southwestern 

Medical School and a fellowship in cell transplantation for 

restorative neurologic function at Karolinska Institute in 

Stockholm, Sweden. Dr. Stieg has developed an international 

reputation in the area of cerebrovascular disorders. He 

has been active in many international courses and been published 

broadly. He has contributed to groups such as the 

Joint Sections of Cerebrovascular Surgery of the American 

Association of Neurological Surgeons and Congress of 

Neurological Surgeons (AANS/CNS) where he now assists in 

the capacity as Secretary. In addition, he is President-Elect 

of the Society of University Neurosurgeons. A recipient of 

several awards and honors, including citations in “Who’s 

Who in Health and Medical Services” and “The Best 

Doctors in America,” Dr. Stieg has played a major role in 

developing and enhancing the neurosurgical programs at 

Harvard, Brigham and Women’s Hospital, and Children’s 

Hospital of Boston. 


3 


8:05 AM - 9:35 AM 

Room 105/106 

(5) Nonneoplastic Temporal Bone 

Imaging (ASHNR) 

(9) Middle Ear Disease 

Middle Ear Disease 

H. Ric Harnsberger, MD 

— H. Ric Harnsberger. MD 

(10) Tinnitus-Pulsatile and Not 

— Michele H Johnson, MD 

(11) Facial Palsy 

— Richard H. Wiggins, III, MD 

Moderators: H. Ric Harnsberger, MD 




able to: 

1) Review T-bone CT imaging anatomy of the middle ear 

and mastoid relative to congenital, inflammatory, and infectious 

diseases that occur in this area. 

2) Identify a rational imaging approach to non-neoplastic 

diseases of the middle ear and mastoid. 

3) Become familiar with the key facts, imaging findings, and 

differential diagnoses associated with congenital, inflammatory, 

and infectious diseases affecting the middle ear and 

mastoid. 


In this presentation middle ear-mastoid anatomy, imaging 

strategies and disease states will be reviewed. Complex temporal 

bone CT anatomy is simplified by focusing on anatomical 

structures of the middle ear-mastoid commonly affected 

by diseases of this area. Focusing on the normal structures 

that may be affected by cholesteatoma and otomastoiditis 

allows a practical anatomical review to be completed. 

Nonneoplastic diseases of the middle ear and mastoid fall 

into three major pathologic categories, congenital, inflammatory, 

and infectious. Chief among the lesions that may be 

encountered by the radiologist are congenital and acquired 

cholesteatoma, cholesterol granuloma, and otomastoiditis 

with extracranial or intracranial complications. Acquired 

cholesteatoma can be divided further into pars tensa, pars 

flaccida, and mural subtypes. Acute otomastoiditis causes a 

Monday

Monday 

range of extracranial and intracranial complications depending 

on direction of spread. In the case of lateral spread, 

postauricular abscess occurs. Cephalad spread through the 

tegmen tympani results in temporal lobe cerebritis or 

abscess. Medial spread across the sigmoid plate of the medial 

mastoid wall can cause sigmoid sinus thrombosis with or 

without vein of Labbe thrombosis with temporal lobe infarction. 

If the infection accesses the inner ear fluid spaces 

through the oval or round window, suppurative labyrinthitis 

may lead to meningitis of the IAC-CPA region. Finally, 

dehiscence of the inferior mastoid cortex may present as 

acute suppuration of the sternocleidomastoid muscle. Each 

of these infection-spread patterns with their associated complications 

will be highlighted in this talk. 

REFERENCES 

1. Penido Nde O, Borin A, Iha LC, et al. Intracranial complications 

of otitis media: 15 years of experience in 33 patients. 

Otolaryngol Head Neck Surg 2005;132:37-42 

2. Vazquez E, Castellote A, Piqueras J, et al. Imaging of complications 

of acute mastoiditis in children. Radiographics 

2003;23:359-372 

3. Dobben GD, Raofi B, Mafee MF, et al. Otogenic intracranial 

inflammations: role of magnetic resonance imaging. Top 

Magn Reson Imaging 2000;11:76-86 

4. Swartz JD, Harnsberger HR, Mukherji SK. The temporal 

bone. Contemporary diagnostic dilemmas. Radiol Clin North 

Am 1998; 36:819-853 

Tinnitus-Pulsatile and Not 

Michele H Johnson, MD 

Facial Palsy 




able to: 

1) Understand the complicated anatomy of the facial nerve. 

2) Recognize the importance of the surrounding anatomical 

landscape of the intratemporal facial nerve course. 

3) Illustrate facial nerve pathology. 


It is imperative that the tortuous course of the facial nerve be 

taken into account when imaging the temporal bone. This is 

a mixed nerve, with branchial motor, visceral motor, general 

sensory, and special sensory functions. The accurate evaluation 

of these individual functions can greatly assist with the 

accurate localization of pathology along its entire course, 

from the nucleus of origin in the brain stem to the end organ 

(the tongue, facial expression muscles, and the lacrimal and 

salivary glands). Imaging of the facial nerve consists primarily 

of MR imaging and CT examinations. High resolution 

MR imaging is most useful for imaging the cisternal and 

canalicular segments of the nerve, while CT is most useful 

for evaluation of the osseous intratemporal nerve course. 

The pathology of the facial nerve is similar to other nervous 

structures of the head and neck, but the unique surrounding 

4 

anatomical landscape predicts and generates the imaging 

appearances seen. 

REFERENCES 

1. Harnsberger HR, Wiggins RH, et al. Diagnostic imaging: 

Head and neck. Amirsys 2004 pp I - 2 - 176-199 

2. Harnsberger HR, Wiggins RH, et al. Pocket radiologist top 

100 diagnoses: Temporal bone. Amirsys 2003 


8:00 AM - 9:35 AM 

Room 103 

(6) ELC Workshop A: Introductory 

PowerPoint 


10:00 AM - 11:30 AM 

Room 105/106 

(7a) INTERVENTIONAL: ENT 

Intervention and Miscellaneous 

(Scientific Papers 13 - 23) 

See also Parallel Sessions 

(7b) HEAD AND NECK: Temporal Bone, Sinonasal 

(7c) INTERVENTIONAL: Aneurysms 

(7d) ADULT BRAIN: Vascular, Extracranial 

Moderators: Gary M. Nesbit, MD 


— John L. Go, MD 

Paper 13 Starting at 10:00 AM, Ending at 10:08 AM 

Intraarterial 

Carcinoma 

Tantivatana, J. 

Chemoinfusion of Nasopharyngeal 

Chulalongkorn University 

Bangkok, THAILAND 

PURPOSE 

Patients with advanced nasopharyngeal carcinoma who 

failed irradiation, surgery, or chemotherapy treatment suffer 

from locally recurrent symptoms. Few treatment options are 

available for palliation. We evaluate possible palliative inter-

vention with concept of transcatheter intraarterial delivery of 

chemotherapeutic agent directly to the involved vascular territory. 

MATERIALS & METHODS 

Fifteen patients, 10 males and 5 females with locally recurrent 

nasopharyngeal carcinoma (NPC), including inoperable 

tumor, skull base invasion, failed irradiation or chemotherapy 

were treated with intraarterial chemoinfusion. Ninety mg 

of Paclitaxel was infused over a period of 2 hours via a 

microcatheter placed in corresponding branch of external 

carotid artery which supplied the tumor every 4-6 weeks for 

3 courses. Result of treatment was reviewed with follow-up 

period of 3 to 6 months. 

RESULTS 

Nine patients had significantly decreased tumor volume. 

Three patients show no significant change of tumor volume. 

Two patients had progression of disease. One patient died of 

distant metastasis. One of five patients who had cranial 

nerve palsy improved. Three patients had transient alopecia. 

Four patients had transient hemifacial numbness. No hematologic 

or systemic side effect was found. 

CONCLUSION 

Intraarterial chemoinfusion is a potential choice of local palliation 

in cases of recurrent or refractory tumor with poor 

response to radiation, iv chemotherapy, or surgery. 

KEY WORDS: Intraarterial chemotherapy, nasopharyngeal 

cancer, interventional radiology 


Endovascular Treatment of Idiopathic Epistaxsis: 

Comparative Study of Unilateral versus Bilateral 

External Carotid Artery Embolization 

Myers, M. E. · Madison, M. T. · Goddard, J. · Myers, T. 

St. Paul Radiology 

St. Paul, MN 

PURPOSE 

Endovascular techniques for treating persistent idiopathic 

epistaxsis have become accepted medical practice. However, 

many varied approaches to treating this disease exist. We 

sought to determine whether embolizing branches of both 

external carotid arteries had any clinical advantage over unilateral 

embolization. 


All patients suffered from persistent unilateral epistaxsis 

despite agressive packing by ENT. A total of 150 patients 

were treated from 2000 to 2004 using endovascular techniques. 

Patients were randomized and treated either with 

bilateral internal maxillary and unilateral facial artery 

embolization (250 micron particles + gelfoam) or with unilateral 

internal maxillary artery embolization alone. Mean 

follow up was 24 months. Outcomes measured were recurrent 

epistaxsis, procedural complications, and facial pain. 

RESULTS 

Seventy-five patients were entered into each treatment 

group. In the bilateral embolization group, there was 1 

rebleed requiring further treatment, 1 minor stroke, and 1 

5 

TIA. Facial pain requiring medication occurred in 15 

patients. In the unilateral embolization group, there were 2 

rebleeds requiring further treatment, no strokes or TIAs and 

2 patients with facial pain requiring medication. 

CONCLUSION 

Endovascular embolization for treatment of idiopathic epistaxsis 

was safe and effective in both groups. The unilateral 

less invasive approach did not result an any significant 

increase in recurrent nosebleeds and had the added benefits 

of being less time consuming for the operator and considerably 

less postembolization facial pain for the patient. 

KEY WORDS: Epistaxsis 


Ethanol Endovascular Repair of Ear Arteriovenous 

Malformations 

Yakes, W. F. 

Vascular Malformation Center 

Englewood, CO 

PURPOSE 

To determine the efficacy of ethanol endovascular repair of 

ear arteriovenous malformations (AVMs). 


Six patients (5 female, 1 male; age range 6-39 years; mean 

age: 22 years) with ear AVMs presented for therapy. Two 

patients had failed prior embolizations 

(PVA/coils/nBCA/steroids) and 2 patients had other therapies 

(laser/excisions/grafting). All presented with grossly 

enlarged painful ear with intermittent bleeding. All patients 

underwent transcatheter and direct puncture ethanol Rx (77 

procedures). 

RESULTS 

All 6 patients were cured of the AVM at long-term follow up 

(mean follow up: 39 months). One patient had transient partial 

VII nerve palsy. Two patients had minor blisters. 

CONCLUSION 

Ethanol endovascular repair of ear AVMs can effect cures in 

this vexing lesion. 

KEY WORDS: AVM, embolization, ethanol 


Safety and Efficacy of Definitive Embolization of 

Meningiomas Trial: Rationale for Study and Study 

Design 

Bateman, B. T. · Berman, M. F. · Pile-Spellman, J. 


New York, NY 

PURPOSE 

The purpose of this presentation is to discuss a new multicenter 

pilot study that is being initiated to determine the safety 

and efficacy of definitive embolization of meningiomas 

(SEDEM trial). In part one, outcomes from surgery and 

Monday

Monday 

radiosurgery will be discussed, providing a background on 

the currently accepted treatments for this disease. In part 

two, the literature on definitive meningioma embolization 

will be reviewed. In part three, the design of the SEDEM 

trial will be presented, as will any preliminary results from 

the trial that may be available. 


A systematic literature search was performed to identify all 

meningioma surgery series involving >100 patients published 

since 1980. Both short- and long-term outcomes were 

recorded. The radiosurgery literature also was reviewed, and 

the results of studies enrolling > 50 patients were recorded. 

The Nationwide Inpatient Sample, the largest discharge database 

available in the U.S. with approximately 8 million discharge 

records reported annually, was searched for all 

meningioma resections performed between1998-2002. 

Outcomes following resection were determined using this 

database. The literature available on definitive meningioma 

embolization also was reviewed and summarized. A multicenter 

study was designed to assess the safety and efficacy of 

treating meningiomas by definitive embolization. 

RESULTS 

The review of the surgical literature showed a relatively high 

rate of morbidity and mortality following surgery, with mortality 

rates ranging from 1-9% and complication rates ranging 

from 10-29%. The search of the NIS database showed an 

overall in-hospital mortality rate of 1.9% following resection 

(but this rate climbed to 3% for patients 70-79 years old and 

7% for patients > 80). There is also a substantial risk of 

recurrence following surgery, with recurrence rates reported 

as high as 32% following gross total resection and 91% following 

subtotal resection. The review of the radiosurgery literature 

showed similar tumor control rates to surgery, with a 

slightly lower rate of complications (5-13%). But this 

modality is limited to tumors smaller than 3.5 cm. There is 

one case report of long-term control of a meningioma using 

embolization alone. There is also a series published of seven 

patients treated with embolization alone, with tumor control 

achieved in six of the patients. The trial which is being initiated, 

the SEDEM trial, is a multicenter study designed to 

provide estimates of the safety and efficacy of definitive 

embolization. Forty patients are to be enrolled, with each followed 

for at least 5 years after embolization with imaging 

and neurologic exams. Preliminary results from the SEDEM 

trial will be presented, if available. 

CONCLUSION 

The review of outcomes from surgery and radiosurgery suggest 

that novel therapies that are safer and more effective 

need to be developed to treat meningiomas. The limited data 

on definitive embolization available from the literature suggest 

this is a promising approach, meriting further investigation. 

The SEDEM trial should do much to clarify the utility 

of definitive embolization as a therapeutic modality for 

meningiomas. 

KEY WORDS: Meningioma, embolization 

6 


Quantification of Angiogenesis Using Angiography 

Gounis, M. J. 1 · Lieber, B. B. 1 · Webster, K. A. 2 · Bishopric, 

N. H. 2 · Spiga, M. G. 2 · Wakhloo, A. K. 2 

1 University of Miami, Coral Gables, FL, 2 University of 

Miami, Miami, FL 

PURPOSE 

Numerous methods of therapeutic angiogenesis, including 

different gene delivery vehicles as well as the route of 

administration of these vectors, have been studied using 

angiographic time-intensity data. Mathematical modeling of 

these data provides indices of the formation and function of 

new vessel growth in response to the treatment modality. 


The rabbit hindlimb model of ischemia was used for the in 

vivo study. Briefly, the external iliac artery was ligated and a 

segment of the left femoral artery was resected. High-speed 

angiographic images of the hindlimb vasculature were 

obtained prior to and after surgically induced ischemia and at 

various follow-up times (Siemens Angiostar Plus, Forchheim, 

Germany). Various treatments were studied, including intramuscular 

injections of adenoviral (Ad) and regulated adenoassociated 

(AAV) viral vectors expressing vascular endothelial 

growth factor (VEGF), as well as autologous bone marrow- 

derived stem cells that were transduced with these vectors 

ex vivo. Super selective intraarterial injections of these 

treatments under flow isolation were explored. All treatment 

groups were compared to control groups that had PBS injected. 

Using angiography, changes occurring within the visible 

arteries having a diameter greater than 250 µm and the 

microvasculature where arteries are too small to be resolved as 

individual vessels were quantified. The normalized, average 

inverted gray values (NAIGV) within were calculated. Timeintensity 

curves of the NAIGV throughout the observation 

period then were constructed and modeled. The model fit was 

optimized and the resulting parameters provide quantification 

new vessel formation as well as contrast trasport. 

RESULTS 

In comparison with the control group, the group that 

received intramuscular injections of the Ad had significantly 

increased vessel growth and fluid transport through these 

vessels 1 week following treatment (p < 0.01). However, the 

functional angiogenesis dissipated at subsequent follow-up 

time points. Similar results were obtained in the angiographic 

quantification of the microvascular density, namely the 1 

week significant increase (p < 0.01) disappeared with time as 

compared to the control group. Histologic analyses indicated 

strong endothelial cell proliferation and angiogenesis during 

the first week of VEGF delivery that coincident closely with 

the transient surge of VEGF expression. Capillary density 

was enhanced during 1-3 weeks but returned to baseline after 

6 weeks. The latter decline of capillary density coincided 

with evidence of increased apoptosis. These results may help 

explain the absence of therapeutic benefit of single dose 

VEGF delivered by adenovirus. Current work using the regulated 

AAV indicates that sustained angiogenic benefit may 

be achieved. Superselective regulated AAV administration 

under flow arrest is being examined. The initial results indicate 

controlled neovascular formation, with organized vessel 

growth into the ischemic territory.

CONCLUSION 

The described method of modeling dynamic angiographic 

data gives not only information on the geometry of new vessel 

growth, but has the ability to quantify the function of vessels 

that form in response to therapeutic angiogenesis. Our 

analysis suggests that a single dose of Ad to express VEGF 

does not produce sustained angiogenic vessels. New methods 

of therapeutic angiogenesis including the vector as well 

as the mode of administration are being studied. 

KEY WORDS: Angiogenesis, mathematical modeling, 

ischemia 


Intraarterial Nicardipine for the Treatment of 

Subarachnoid Hemorrhage-Associated Vasospasm: 

Initial Clinical Experience with High-Dose Infusions 

Tejada, J. G. 1,2 · Chaloupka, J. C. 2 · Lee, S. K. 2,3 · Ugurel, M. 

S. 2 · Hayakawa, M. 2 · Taylor, R. A. 2 

1 2 Indiana University, Indianapolis, IN, University of 

Iowa, Iowa City, IA, 3University of Toronto, Toronto, 

ON, CANADA 

PURPOSE 

Delayed cerebral ischemia due to vasospasm is a major 

cause of morbidity and mortality following aneurysmal 

SAH. Although vasospasm refractory to medical therapy can 

be treated effectively by low-pressure balloon angioplasty, 

severe complications and limitations are associated with this 

technique. Alternatively, intraarterial infusion of papaverine 

has been shown to be at least transiently effective; however, 

several severe adverse effects also have been reported with 

this technique. Recently, alternative medications have been 

considered, such as amrinone and calcium-channel blockers. 

We report our initial experience with 11 patients treated with 

superselective IA nicardipine. 


Eleven consecutive patients with clinically significant and 

intractable vasospasm following aneurysmal SAH were 

treated with a high concentration (0.83 mg/mL) and high 

dose rate (0.415-0.83 mg/min) of nicardipine. Standard criteria 

for case selection/definition were used. After catheter 

angiographic confirmation of vasospasm, the affected arteries 

were targeted for superselective catheterization. Total 

dosages of 10 to 40 mg were infused. Control injections 

were obtained after 2-5 mg intervals to assess response. 

RESULTS 

Eleven patients underwent a total of 20 procedures for 

intractable vasospasm over the past year. Most were higher 

Hunt-Hess grades {4/11 [36%] grade IV, 3/11 [27.2%] grade 

III. Neurologic deficits included depressed LOC [n = 11 

(100%)], paresis [n = 6 (54.5%)], aphasia [n = 1 (1%)], and 

facial nerve palsy [n = 1 (9%)]}. The angiographic results 

were very good with significant improvement in the diameter 

of the affected vessel seen in 18/20 (90%) of the procedures. 

In one procedure (10%) the result was excellent with 

complete restoration of the diameter of the vessel. Clinical 

improvement with resolution of the focal symptoms or overall 

improvement in GCS was demonstrated in 10/11 

(90.9%). One of 11(9%) died from complications related to 

the initial intracranial hemorrhage. The procedure was well 

7 

tolerated for all the patients. No complications were 

observed in 16/20(80%), while minor complications without 

sequelae were observed in 4/20 (20%). Ten of 11 (90.9%) 

patients finally were discharged from the hospital. Clinical 

follow up was possible in 9/11 (81.8%) patients, 2-8 months 

after hospital discharge. 

CONCLUSION 

Rapid, high dose intraarterial nicardipine infusions for the 

treatment of SAH-associated vasospasm appear to be a safe 

and effective alternative intervention. Although retreatment 

was often necessary, such procedures seem to be well tolerated. 

Further clinical studies will be needed to confirm these 

results in a larger patient population. 

KEY WORDS: Vasospasm, nicardipine, intraarterial 


Effects of Carotid or Vertebrobasilar Stent Placement on 

Cerebral Perfusion and Cognition 

Turk, A. S. · Niemann, D. · Moftakhar, R. · Rowley, H. · 

Aagaard-Kientiz, B. · Turski, P. 

University of Wisconsin 

Madison, WI 

PURPOSE 

There are no well established physiologic or neuropsychological 

criteria for identifying which patients with stenosis of 

the cervicocerebral vessels are at high risk of stroke or cognitive 

impairment. Our purpose was to evaluate changes in 

cognitive performance and cerebral perfusion associated 

with endovascular stenting of the cervicocerebral vessels. 


A series of 28 patients, 30 to 88 years of age, who underwent 

31 stent procedures for arterial stenosis [14 extracranial 

carotid stents (ECS), 5 intracranial carotid stents (ICS) and 

12 extra or intracranial vertebrobasilar stents (VBS)] were 

reviewed retrospectively. All patients were evaluated with 

CT or MR perfusion studies before and after stent placement 

(Fig.). Cognitive response after stenting was evaluated with 

an informant questionnaire (informant questionnaire for cognitive 

decline in the elderly, IQ-CODE). 

RESULTS 

In cases with anterior circulation stenoses (ECS and ICS 

group), 15 of 19 (79%) had a baseline perfusion abnormality 

and all patients showed improved perfusion after stenting. 

Six of 12 (50%) cases with posterior circulation stenoses 

(VBS group) had a baseline perfusion abnormality and 4 of 

the 6 patients showed improved perfusion after stenting. 

Degree of stenosis was the strongest predictor of the presence 

of a baseline perfusion abnormality (p = 0.03). Twentyfour 

of 31 (77%) of the cases showed improved cognitive 

scores after stent placement. Among patients with improvement 

in perfusion after stent placement, 16 of 19 (84%) had 

improved cognitive scores. Improved perfusion after stent 

placement was a significant predictor of cognitive improvement 

(p = 0.04). Patients who were stented on an elective 

basis demonstrated greater improvement in cognition as 

compared to patients stented urgently (p = 0.01). 

Monday

Monday 

CONCLUSION 

Endovascular stenting of the cervicocerebral vessels can 

safely and effectively resolve cerebral perfusion abnormalities. 

Improvement in perfusion parameters is associated with 

cognitive improvement. Larger, blinded, prospective studies 

are warranted to confirm these preliminary observations. 

KEY WORDS: Stent, cerebral perfusion, cognition 


Real-Time 3D Endovascular Navigation 

Pujol, S. 1 · Frerichs, K. 2 · Westin, C. 1 · Vosburgh, K. 1 · 

Norbash, A. 3 

1 2 Harvard Medical School, Boston, MA, Brigham and 

Women’s Hospital, Boston, MA, 3Boston University School 

of Medicine, Boston, MA 

PURPOSE 

Endovascular therapy provides a minimally invasive solution 

for many vascular disorders. The procedure is monitored 

traditionally under fluoroscopic guidance, which intro- 

8 

duces a risk of radiation-induced morbidity. Multidetector 

CT angiography (CTA) can offer valuable anatomical information 

for the treatment of cerebral aneurysms. Our objective 

is to provide the neuroradiologist a computer-assisted 

navigation system to reduce the intraoperative X-ray exposure. 

The system displays — in real-time and without fluoroscopy 

— the position of the endovascular elements in a 3D 

model reconstructed from preoperative imaging. 


Our navigation system relies on the separation of the visualization 

of the patient’s anatomy, and the localization of the 

endovascular tools. During the preoperative phase, we 

reconstruct a 3D model of the brain vasculature from CTA 

exam (slice thickness of 2 mm - reconstruction interval of 

1.5 mm), using an implementation of the level-sets algorithm 

for contour detection. During the intraoperative phase, a 

magnetic localizer (MicroBird, Ascension Technology) 

tracks an endovascular guide whose tip is equipped with a 

miniaturized sensor. The navigation system registers in realtime 

the position of the tip of the guide inside the 3D CTA 

model of the vasculature. We have performed a first validation 

on a phantom of the head, made of a styrofoam skull and 

a plastic model of the cerebral vasculature. Four ECG electrodes 

were used as noninvasive fiducials, and seven control 

points were positioned along the vascular path. At the beginning 

of the procedure, the fiducials were palpated magnetically 

on the phantom, and the corresponding points were 

selected in CT images, in order to replace the CTA model in 

the reference frame of the magnetic localizer. The navigation 

system display was used to reach the control points with the 

endovascular guide. The accuracy was defined as the distance 

between the actual location of the guide, and its 

expected position measured for each control point in CT 

data. X-ray images of the guide validated the results of the 

navigation. 

RESULTS 

The navigation system provides a real-time display of the 

endovascular guide within the 3D CTA model, using magnetic 

tracking. In our initial experiments, the overall system 

accuracy was 2.5 mm. 

CONCLUSION 

We have developed a navigation system to assist neuroradiologists 

in the endovascular treatment of brain aneurysms. 

These phantom experiments demonstrated the potential to

provide real-time 3D navigation in cerebral arteries with 

clinically relevant accuracy, and thereby expect to reduce the 

use of fluoroscopy during neuroradiologic interventions. 

KEY WORDS: Computer-assisted navigation, interventional, 

aneurysm 


Comparison of Image Quality and Feasibility of Clinical 

Use of Various Commercial Endoscopic Devices for 

Intraspinal Navigation 

Purdy, P. 1 · Fujimoto, T. 2 · Replogle, R. 1 · Giles, B. 1 · 

Fujimoto, H. 3 · Miller, S. 1 

1 University of Texas Southwestern Medical Center, Dallas, 

TX, 2 Kobe University Medical School, Kobe, JAPAN, 

3 Okayama University Medical School, Okayama, JAPAN 

PURPOSE 

Percutaneous intraspinal navigation has been discussed as a 

potential technique for access to the CNS. The subarachnoid 

space is catheterized, usually via lumbar puncture, and the 

subarachnoid space is used as a conduit of transit. Fluoro and 

MR guidance has been described (1, 2). This purpose was to 

compare endoscopes for subarachnoid visualization. 


Using 3 unembalmed cadavers, devices were tested for visualization 

of the spinal cord, cranial nerves, and ventricles. 

These included a 1 mm diameter fiberscope introduced via 

an angiographic catheter, a 2.3 mm (7 F) fiberscope with a 1 

mm working lumen, a 2.8 mm fiberscope with a 1.2 mm 

working lumen, and a 3.8 mm videoscope (CCD chip on end 

of scope used for imaging). Each scope was introduced via 

an arterial sheath except for the videoscope, which required 

laminotomy. Images of the spinal cord, cranial nerves, and 

ventricular system were acquired. The 2.8 mm and 3.8 mm 

scopes were tested on the same cadaver. Comparisons of 

steerability, durability, field of view, and subjective image 

quality were performed with each device. 

RESULTS 

Each device was introduced successfully into the lumbar 

subarachnoid space, and successful catheterization of the 

spinal canal, posterior fossa, and ventricular system was 

accomplished easily with each device. Steerability and durability 

of the fiberscope were limited and field of view was 

narrow. Anatomical identification was limited. Steerability 

and durability of the 2.3 mm and the 2.8 mm fiberscopes 

were better. Both had deflecting tips, improving control. 

Ventricular visualization was better than the 1 mm scope 

(Fig.). Image quality was slightly better using the 2.8 mm 

fiberscope. The videoscope was superior in image quality 

and durability, as well as field of view. Its steerability was 

similar to the others. Size may restrict it. 

9 

CONCLUSION 

Prior studies have shown that spinal canal occlusion < 30% 

produces no spinal cord injury. In humans with a canal 1 cm 

diameter, either the 2.3 mm or 2.8 mm scope could be studied. 

Subarachnoid catheterization and visualization of spinal 

and cranial structures is feasible. 

REFERENCES 

1. Purdy PD, Replogle RE, Pride GL Jr, Adams C, Miller S, 

Samson D. Percutaneous intraspinal navigation (PIN): A 

feasibility study in cadavers of a new and minimally invasive 

approach to the spinal cord and brain. AJNR Am J 

Neuroradiol 2003;24:361-365 

2. Rappard G, Metzger GJ, Weatherall PT, Purdy PD. 

Interventional MR imaging with an endospinal imaging 

coil: Preliminary results with anatomic imaging of the 

canine and cadaver spinal cord. AJNR Am J Neuroradiol 

2004;25:835-839 

KEY WORDS: Endoscopes, percutaneous intraspinal navigation, 

subarachnoid 


Endoscopic Visualization of the Spinal Cord via 

Intraspinal Navigation: A Comparison Study 

Giles, B. 1 · Fujimoto, T. 2 · Robert, R. 1 · Fujimoto, H. 3 · Miller, 

S. 1 · Purdy, P. 1 




PURPOSE 

Percutaneous intraspinal navigation (PIN) is a percutaneous 

approach to the cerebral subarachnoid space from a lumbar 

puncture (1). Development of procedures using this technique 

requires adequate instrumentation. The purpose of this 

study is to compare fiberscopic (imaging via optical fiber in 

scope) versus videoscopic (imaging via CCD chip at tip of 

scope) visualization of spinal cord structures. This approach 

may provide an alternative to traditional surgical approaches 

and may minimize the morbidity and mortality associated 

with open surgical exposure. 


An unembalmed male cadaver was obtained and transported 

to our research angiography suite and placed in the prone 

position. Access to the lumbar subarachnoid space was 

Monday

Monday 

achieved via lumbar puncture and a 2.8 mm outer diameter 

(OD) tip-deflecting fiberscope with a 1.2 mm working channel 

was introduced into the subarachnoid space through a 9 

French sheath. The endoscope was advanced cranially under 

endoscopic and X-ray fluoroscopic guidance. The fiberscope 

was removed and a 3.8 mm OD tip-deflecting videoscope 

with a 1.2 mm working channel then was introduced into the 

subarachnoid space via a partial second lumbar laminectomy 

secondary to the inability to place an adequately sized sheath 

percutaneously and advanced cranially. Endoscopic images 

were stored using a direct digital capture apparatus. 

RESULTS 

Structures of the spinal cord were visualized readily with 

both scopes. Navigation from the lumbar entry to the foramen 

magnum was achieved easily and quickly with both 

scopes. The cauda equina was seen upon entering the subarachnoid 

space. As the endoscopes were advanced cranially 

the spinal cord was identified. Spinal nerve roots (Fig) 

were seen originating from the spinal cord and traversing laterally, 

exiting through the dura. 

CONCLUSION 

Obtaining access to spinal cord structures via this intraspinal 

approach is technically feasible under endoscopic and fluoroscopic 

guidance. The videoscope provides a greater degree 

of detail of the surface anatomy and a higher degree of spatial 

resolution when compared to traditional fiberscopes. 

This approach may provide an alternative to traditional surgical 

approaches to the spinal cord. Reduction of the outer 

diameter of the videoscope should allow it to be placed percutaneously. 

The safety and efficacy of PIN should be 

explored. 

REFERENCES 





Neuroradiol 2003;24:361-365 

KEY WORDS: Spinal cord, endoscopy, percutaneous 

intraspinal navigation 

10 


Endoscopic Visualization of Posterior Fossa Structures 

via Intraspinal Navigation: A Comparison Study 

Purdy, P. 1 · Giles, B. 1 · Fujimoto, T. 2 · Replogle, R. 1 · 

Fujimoto, H. 3 · Miller, S. 1 




PURPOSE 

Multiple disorders, such as trigeminal neuralgia, acoustic 

neuromas, and basilar arteries aneurysms, effect posterior 

fossa structures and require open surgery for treatment. In an 

effort to reduce the morbidity and mortality associated with 

surgery, minimally invasive approaches are being explored. 

Percutaneous intraspinal navigation (PIN) (1), provides 

access to the cranial subarachnoid space via lumbar puncture. 

This study compares the ability to visualize posterior 

fossa structures between a fiberscope and a videoscope 

(endoscope with CCD video chip on its tip) via an intraspinal 

route and may provide an alternative minimally invasive 

approach to intracranial structures. 


An unembalmed male cadaver underwent access to the lumbar 

subarachnoid space via lumbar puncture and a 2.8 mm 

outer diameter (OD) tip-deflecting fiberscope with a 1.2 mm 

working channel was introduced into the subarachnoid space 

through a 9 French sheath. The scope was advanced cranially 

under visual and fluoroscopic guidance. The fiberscope 

was removed, and a 3.8 mm OD tip-deflecting videoscope 

with a 1.2 mm working channel then was introduced via a 

partial second lumbar laminectomy. Images were stored 

using a direct digital capture apparatus. 

RESULTS 

Images of the posterior fossa structures were obtained using 

both scopes. Superior to the level of the cisterna magna the 

scope was directed anterolaterally. The seventh and eighth 

cranial nerves were identified and followed laterally. As the 

scopes were advanced cranially the fifth cranial nerve was 

identified. The scope was pulled back down to the level of 

the cisterna magna and directed cranially along the midline. 

The vertebral arteries were visualized and followed to the 

vertebrobasilar junction. The basilar artery was followed 

cranially and multiple pontine arteries were visualized. The 

third cranial nerve was identified and its relationship to the 

superior cerebellar and posterior cerebral arteries were readily 

appreciated (Fig).

CONCLUSION 

Accessing the posterior fossa via an intraspinal approach is 

technically feasible under endoscopic and fluoroscopic guidance. 

The videoscope provides a greater degree of detail of 

the surface anatomy and a higher degree of spatial resolution 

than compared to traditional fiberscopes. This approach may 

provide an alternative to traditional skull-base surgical procedures. 

The safety and efficacy of PIN require further study. 

REFERENCES 





Neuroradiol 2003;24:361-365 

KEY WORDS: Percutaneous intraspinal navigation, endoscopy, 

posterior fossa 


10:00 AM - 11:33 AM 

Theatre 

(7b) HEAD AND NECK: Temporal 

Bone, Sinonasal 



(7a) INTERVENTIONAL: ENT Intervention and 

Miscellaneous 



Moderators: Mahmood F. Mafee, MD, FACR 

Alexander S. Mark, MD 


Intralabyrinthine Schwannomas: Imaging Diagnosis and 

Classification 

Childs, A. M. 1 · Salzman, K. L. 1 · Harnsberger, H. R. 1 · Wiggins, 

R. H. 1 · Davidson, H. C. 1 · Kennedy, R. J. 2 · Shelton, C. 1 

1 2 University of Utah, Salt Lake City, UT, Private Practice 

Clinic, East Melbourne, AUSTRALIA 

PURPOSE 

Intralabyrinthine schwannomas are uncommon lesions seen 

both sporadically and in association with syndromes (NF2). 

They have characteristic features on MR imaging both in 

terms of location and signal characteristics. It is vital to identify 

these lesions in a timely manner, as prognosis and treatment 

hinge on their recognition. As a result, more standardized 

nomenclature would aid in the universal understanding 

of the anatomy of these lesions. The purpose of our study is 

11 

to characterize and name these rare lesions in a systematic 

and reproducible manner using anatomy as the basic tenet. 


A retrospective study of all patients with the diagnosis of 

vestibular schwannoma imaged between 1996 and 2004 was 

performed. Four hundred and two patients were reviewed, 

excluding those patients with known NF2. Patients with 

vestibular schwannoma whose tumors arose from the 

labyrinth (cochlea, vestibule, and semicircular canals) were 

included. Of these, 44 were included in the study. Patients 

were imaged with thin section, high-resolution T2 and contrast-enhanced 

imaging. 

RESULTS 

There were 44 patients with a diagnosis of intralabyrinthine 

schwannoma. Thirty-five patients had complete histories 

available. There were 12 male and 23 female patients with an 

age range of 21 to 78 years with a mean of 53 years. The 

most common presenting symptom was progressive sensorineural 

hearing loss. Using thin section, high-resolution 

MR imaging on a 1.5 T magnet, lesions were characterized 

based on their anatomical location. These included intravestibular, 

intracochlear, vestibulocochlear, transmodiolar, 

transmacular, and transotic. In addition, signal characteristics 

and enhancement were described. 

CONCLUSION 

Intralabyrinthine schwannomas are rare tumors which may 

represent a diagnostic dilemma. In this paper, we will characterize 

the MR appearance of these lesions using high-resolution 

techniques. In addition, we propose a classification 

system which will aid in identification as well as provide a 

reproducible method for diagnosis and a universal standard 

by which the surgical and radiologic communities can function. 

KEY WORDS: Intralabyrinthine, schwannoma 


Two Measurements on Temporal Bone CT Increase 

Recognition of Inner Ear Malformation and Predict 

Sensorineural Hearing Loss 

Purcell, D. D. 1 · Fischbein, N. J. 2 · Johnson, J. 1 · Patel, A. 3 · 

Lalwani, A. K. 4 

1University of California San Francisco Medical Center, San 

Francisco, CA, 2Stanford University Medical Center, Palo 

Alto, CA, 3University of California San Francisco School of 

Medicine, San Francisco, CA, 4New York University School 

of Medicine, New York, NY 

PURPOSE 

We previously have shown that two measurements of the 

cochlea and lateral semicircular canal (LSCC) on temporal 

bone CT scan can be used to detect malformations of the 

inner ear. In this prospective study, we sought to assess the 

reproducibility of the measurements of the cochlea and 

LSCC and determine if abnormal measurements predict sensorineural 

hearing loss (SNHL). 


Two readers independently measured the cochlear height on 

coronal section and the lateral semicircular canal (LSCC) 

Monday

Monday 

bony island width on axial section on 122 patients who 

underwent temporal bone CT. Each ear was interpreted and 

counted separately, and one reader read a subset of exams 

twice. Inter and intrareader variability was evaluated using 

intraclass correlation coefficients (ICC) based on random 

effects model. Hearing loss was determined by extracting 

pure tone audiometry data from medical records. We determined 

the positive and negative predictive values associated 

with each radiology finding. 

RESULTS 

There was substantial inter and intraobserver agreement in 

the identification of abnormality on either measurement 

(ICCs of > 90%.). A measurement was considered abnormal 

if it fell outside two standard deviations from the average 

measurement. Identification of an abnormality on either 

measurement strongly predicted the presence of SNHL. If 

either cochlear hypoplasia or LSCC bony island dysplasia 

was identified, then the patient had SNHL (PPV 100%). 

Additionally, no patients without SNHL had abnormal measurements. 

Furthermore, review of the formal radiology 

reports demonstrated that both cochlear hypoplasia and 

LSSC dysplasia were overlooked in > 50% of patients with 

both abnormal measurements and SNHL. 

CONCLUSION 

The measurement of coronal cochlear height and axial LSCC 

bony width on CT scan is both reproducible and highly predictive 

of SNHL. These measurements allow identification 

of cases of inner ear dysplasia missed by visual inspection 

alone. Therefore, the measurement of the cochlear height on 

coronal section and LSCC bony island width on axial section 

should be performed routinely on all temporal bone studies 

performed for hearing loss and/or suspected inner ear abnormality. 

KEY WORDS: Inner ear malformations, sensorineural hearing 

loss, standardized measurements 

12 


Diagnostic Accuracy of Virtual Otoscopy and 3D 

Reconstructions in the Preoperative Assessment of the 

Ossicular Chain in Atticoantral Otitis Media: Work in 

Progress 

Panday, A. K. · Bapuraj, J. R. · Gupta, A. K. · Khandelwal, 

N. · Suri, S. 

Postgraduate Institute of Medical Education and Research 

Chandigarh, INDIA 

PURPOSE 

To assess the diagnostic accuracy of virtual otoscopy (VO) 

and 3D ossicular reconstructions in the preoperative assessment 

of the ossicular chain in atticoantral otitis media: work 

in progress. 


Forty patients of chronic suppurrative otits media with conductive 

deafness of > 35dB are to be included in this 

prospective study. Of these results of 33 patients are available 

for evaluation and presentation at the time of submission 

of this abstract. All patients underwent baseline axial 

HRCT on a multislice helical CT scanner (GE Lightspeed 

QXi, GE Medical Systems, Milwaukee, WI). The axial CT 

data set was utilized for (a) coronal and sagittal 2D reformations 

and (b) virtual otoscopy and 3D reconstructions. These 

were read independently by two neuroradiologists who were 

blinded to the clinical findings and to the findings of each 

other. The results were recorded with respect to identification 

of the bones and the integrity of the ossicular chain and 

articulations. A 3-point scoring system was used to assist in 

an objective assessment. All patients then underwent surgery 

and the preoperative findings were recorded by the surgeon 

using the same scoring system. The surgeon was blinded to 

the findings on imaging. The scores for the 2D, VO/3D and 

preoperative finding were compared using the chi square 

test. Accuracy of the 2D and 3D images were calculated with 

respect to the operative findings. 

RESULTS 

Preliminary analyses indicate that there is significant correlation 

between 2D reformations, VO/3D reconstructions and 

operative findings for most parts of the ossicular chain. The 

accuracy of 2D reconstructions and VO for assessment of the 

larger ossicles (i.e., malleus and incus is comparable both 

being 100% accurate for evaluation of malleus, while the 

accuracy for the evaluation of the incus being 88.29% on 

VO/3D and 85.29% on 2D images). For evaluation of stapes 

superstructure, VO/3D images were more accurate (85.29%) 

than 2D images (76.47%). Evaluation for the type of ossicular 

destruction by both imaging modalities showed a high 

level of correlation (P < .001) with preoperative findings; 

however VO /3D images show a better correlation (R = 

0.8065) than HRCT /2D (R = .7273) images. For assessment 

of lenticular process of the incus both 2D and VO/3D image 

was unreliable (P > 0.1) Assessment of the incudostapedial 

joint by HRCT and 2D reformatted images is not reliable (P 

> 0.1); however significant correlation (P < 0.001) is present 

between VO/3D on preoperative findings. 

CONCLUSION 

Three-dimensional/VO reconstructions are helpful in assessing 

the ossicles in cases of CSOM, which are more difficult

to observed on direct otoscopy. It is especially useful to 

demonstrate smaller structures such as the stapes superstructure 

which may influence decision regarding planning 

of ossiculoplasties in this condition. 

KEY WORDS: Temporal bone, multidetector CT, 3D reconstructions 

Dr. Bapuraj will present the paper. 


Absent Semicircular Canals in CHARGE Syndrome: 

Radiologic Spectrum of Findings 

Morimoto, A. K. 1 · Wiggins, R. H. 1 · Mukherji, S. K. 2 · 

Telian, S. A. 3 · Hudgins, P. A. 4 · Hedlund, G. L. 5 · Hamilton, 

B. 1 · Harnsberger, H. R. 1 

1 University of Utah, Salt Lake City, UT, 2 University of 

Michigan Health System, Ann Arbor, MI, 3 University of 

Michigan, Ann Arbor, MI, 4 Emory University, Atlanta, GA, 

5 Primary Children’s Hospital, Salt Lake City, UT 

PURPOSE 

This paper will describe the combination of CT and MR 

findings that characterize the middle and inner ear anomalies 

in CHARGE (Coloboma, heart defects, choanal atresia, 

mental retardation, genitourinary and ear anomalies) syndrome. 

The combined findings of abnormalities of the incus 

and stapes, oval window atresia, and abnormalities of components 

of the inner ear (dysplastic cochlea, hypoplastic or 

dysplastic vestibule, and absent semicircular canals) are 

characteristic of the CHARGE syndrome. Relative frequencies 

of each finding are described as well as discussion of the 

proposed embryogenesis. 


CT and MR imaging studies from 19 patients were studied 

retrospectively to categorize the various middle and inner ear 

anomalies associated with CHARGE syndrome. This sample 

of patients was gathered over several years from multiple 

institutions. The diagnosis of CHARGE was made by the 

referring team based on diagnostic criteria. Investigational 

review board approval was obtained under HIPAA regulations. 

CT and MR imaging studies were reviewed by two 

CAQ neuroradiologists to establish middle and inner ear 

anomalies. Criteria for evaluation were as follows: aplasia of 

the semicircular canals was noted when the canals were 

completely absent and an isolated vestibule was present. 

Cochlear anomalies were related to size of the basal and apical 

turns. Cochlear aperture size also was recorded as normal 

or abnormal (small aperture and/or bony bar presence or 

absence) and whether the cochlear nerve was intact. The criterion 

for large vestibular aqueduct was based on a diameter 

greater than 2 mm in the intraosseous portion. Atresia of the 

oval or round window was identified when they were 

obscured partly or completely by an osseous septum. The 

middle ear ossicles were evaluated for dysplasia and fusion. 

Dysplasia was noted if the ossicles demonstrated abnormal 

morphology or size. The criteria for fusion were based on 

ankylosis to the wall of the mesotympanum or interossicular 

fusion. Abnormal facial nerve segment positions also were 

recorded. 

13 

RESULTS 

Each ear was evaluated separately with anomalies categorized 

as associated with right or left ears. Even with this 

association, no particular relevance to ear side was established. 

Thirty-two of 38 (84%) ears had some form of 

cochlear dysplasia. Sixteen of 38 (42%) ears had round window 

atresia or aplasia. Thirty-one of 38 (82%) ears had oval 

window atresia or aplasia. Nineteen of 38 (50%) ears had 

vestibular hypoplasia. Eleven of 38 (29%) ears had dysplasia 

of the vestibular aqueduct. All cases demonstrated absent 

semicircular canals. Regarding the facial nerve, all had normal 

labyrinthine segments. Thirty of 38 (79%) ears had posteriorly 

displaced first genu segments. Six of 38 (16%) ears 

had prolapsed or diagonal tympanic segments. Two of 38 

(0.05%) ears had diagonal mastoid segments. Other common 

findings included ossicular dysplasia and fusion, venous 

anomalies, and entrapped cochlea having a high association 

with cochlear nerve aplasia. 

CONCLUSION 

CHARGE syndrome patients should be screened for conductive 

as well as sensorineural hearing loss. All patients in 

this review had absent semicircular canals and a majority 

had an abnormally positioned facial nerve involving the first 

genu and/or oval window atresia. 

KEY WORDS: Head and neck, absent semicircular canal, 

CHARGE 


Cochlear-Carotid Interval: Anatomical Variation and 

Potential Clinical Implications 

Young, R. J. · Shatzkes, D. R. · Lalwani, A. K. · Babb, J. S. 


New York, NY 

PURPOSE 

A temporal bone CT study in a patient with intermittent tinnitus 

and hearing loss demonstrated absence of bone 

between the petrous internal carotid artery and the basal turn 

of the cochlea. Although the relationship between this 

patient’s symptomatology and this finding is uncertain, the 

potential implications with respect to increasingly popular 

cochlear implant surgery compelled us to analyze retrospectively 

a series of temporal bone CT scans in order to establish 

normative measurements for this region, which we 

termed the “cochlear-carotid interval” (CCI). 


Following IRB approval, two observers independently measured 

the bony interval between the cochlea and the petrous 

internal carotid artery on coronal images from 30 consecutive 

temporal bone CT studies. The 1 mm thick coronal 

images were acquired either directly or were reconstructed 

from an axial data set acquired at 0.75 or 0.6 mm slice thickness. 

All measurements were performed using electronic 

calipers on a Sienet MagicView VE 42 Siemens PACS station. 

Mixed model analysis of variance was used to evaluate 

differences between readers and sides with respect to the 

mean CCI while adjusting for age and accommodating the 

correlation among observations generated for the same subject. 

Monday

Monday 

RESULTS 

The right CCI ranged from 0.2-3.8 mm (mean 1.2 ± 0.8 mm, 

median 0.9) and the left CCI ranged from 0.2-5.0 mm (mean 

1.1 ± 0.9 mm, median 0.8). The results indicate that the CCI 

did not exhibit a significant association with subject age (p = 

0.1336) and that there were no significant differences 

between readers (p = 0.824) or sides (p = 0.350) in terms of 

mean CCI. 

Fig. 1. Coronal CT images demonstrate the thickness of bone 

between the basal turn of the cochlea and petrous carotid 

canal (cochlea-carotid interval), (A) CCI = 0 in our index 

patient (single arrow) and (B) CCI = 1.2 mm in another 

patient (double arrow). 

CONCLUSION 

The CCI varies widely between patients. Preoperative 

knowledge of thin or absent bone between the cochlea and 

petrous carotid may help prevent inadvertent penetration of 

the carotid canal during cochlear implant surgery. Analysis 

of larger data sets with clinical correlation may help determine 

any potential relationship of diminished CCI and tinnitus 

and/or hearing loss. 

KEY WORDS: Temporal bone, cochlea, carotid canal 


Duplication of the Internal Auditory Canal 

Gasparini, F. F. · Papsin, B. C. · Shroff, M. · Friedberg, J. · 

Padfield, N. · Blaser, S. 

Hospital for Sick Children 

Toronto, ON, CANADA 

PURPOSE 

To demonstrate a rare malformation which has significant 

implication in the surgical planning for cochlear implantation. 


Eight patients were identified retrospectively via a radiology 

report text search program. High-resolution petrous CT with 

direct axial and coronal views were evaluated. Parasagittal 

reconstruction and 3D reconstructions from the same data set 

were performed. 

RESULTS 

Six patients had unilateral, and two patients had bilateral 

involvement. Sensorineural hearing loss was profound in 

each of the involved ears. In some patients the malformation 

showed a clearly separated internal auditory canal (IAC) 

from an isolated facial nerve canal (FNC) with each canal 

coursing in a different orientation. One of the canals led to 

the geniculate turn of the FNC and the other followed its nor- 

14 

mal course through the labyrinth. In other patients the IAC 

and the FNC were separated only by a transverse megacrest. 

All involved ears, except one, had a narrowed cochlear nerve 

canal (CNC). In three patients, the contralateral IAC and the 

CNC were normal. In the other three patients (one with bilateral 

involvement) the contralateral IAC was hypoplastic. 

One patient had bilateral transverse megacrest with normal 

size IAC. One patient had a medially displaced contralateral 

FNC, but normal size IAC. Both patients with bilateral 

involvement presented with dysplastic vestibules and horizontal 

semicircular canals. 

CONCLUSION 

Duplication of the IAC is extremely rare and likely on the 

severe end of spectrum including transverse megacrest. IAC 

duplication in our patients was associated with profound sensorineural 

hearing loss, despite the variability in the size of 

the IAC. Severe hypoplasia of the cochlear nerve canal, a 

contraindication to cochlear implantation, was present in all 

patients with hypoplastic IAC, except one. 

KEY WORDS: Facial nerve canal, duplication, cochlear 

implantation 


Optimization of Reconstruction Parameters for 64-Slice 

CT of the Temporal Bone 

Diehn, F. E. · Witte, R. J. · Lane, J. I. · Lindell, E. P. · Primak, 

A. N. · McCollough, C. H. 

Mayo Clinic 

Rochester, MN 

PURPOSE 

High-resolution, isotropic, spiral imaging of the temporal 

bone is now possible with 64-slice CT. The purpose of this 

study is to optimize the reconstruction parameters for 64slice 

CT evaluation of the temporal bone in the axial and 

coronal planes. 


All patients were scanned using the Somatom Sensation 64 

(Siemens Medical Solutions; Forchheim, Germany). Spiral 

data were acquired with the head in neutral position. 

Reconstruction parameters included slice thickness, reconstruction 

increment (overlap), field of view size, and kernel. 

Reconstructions were performed in the axial and coronal 

planes. All images were reviewed by three board-certified 

staff neuroradiologists. Subjective comparisons of anatomical 

detail were performed for various structures, including 

the scalar lamina and the anterior and posterior crura of the 

stapes in the axial plane, and the inferior wall of the facial 

nerve canal in the coronal plane. 

RESULTS 

In all patients, the imaging consistently provided detailed 

visualization of the temporal bone. In the axial plane, optimal 

resolution of anatomical structures was seen with intermediate 

slice thickness (0.75 mm), 50% slice overlap, intermediate 

field of view (80 mm), and a sharp ultrahigh-resolution 

kernel. In the coronal plane, thinner slice thickness (0.49 

mm), smaller field of view (70 mm), and a sharp ultrahighresolution 

kernel resulted in optimal visualization of the 

anatomical structures.

CONCLUSION 

Sixty-four-slice CT imaging provides fast, reliable, high-resolution 

examination of the temporal bone. Optimization of 

reconstruction parameters results in consistent visualization 

of previously difficult to image structures in both the axial 

and coronal plane. 

KEY WORDS: Temporal bone, 64-slice CT, postprocessing 


Cerebrospinal Fluid Leaks: Correlation of High- 

Resolution Multiplanar Reformations with 

Intraoperative Findings 

La Fata, V. · McLean, N. · DelGaudio, J. · Hudgins, P. A. 

Emory University School of Medicine 

Atlanta, GA 

PURPOSE 

Cerebrospinal fluid (CSF) leaks from skull base defects are a 

debilitating problem, often resulting in significant morbidity. 

Preoperative evaluation of these patients is traditionally performed 

with thin-slice CT scan techniques and direct coronal 

acquisition. A cohort of patients with clinically suspected CSF 

leaks who underwent high-resolution axial CT scanning was 

reviewed retrospectively. Coronal and sagittal multiplanar reformations 

were performed for all examinations at a workstation. 

Surgical and/or endoscopic correlation was available for 

all patients. Both the site and the size of the bony skull base 

defects were correlated with the intraoperative findings. 


Sixteen patients with CSF leak underwent CT scanning, and 

these images were reviewed retrospectively. High-resolution 

CT scans were obtained using direct axial images at either 

0.625 mm or 1.25 mm increments, and multiplanar reformations 

were performed on a workstation. One patient underwent 

a screening sinus CT using 2.5 mm axial images. All 

CT scans were interpreted by two neuroradiologists who 

were blinded to the intraoperative findings. Determination of 

the site and size of the bony defects were arrived by consensus, 

using both the coronal and sagittal reformations. The 

size of the bony defect was measured using electronic 

calipers in the anterior-posterior and transverse dimensions. 

When available, MR scans were reviewed also to assess for 

the secondary signs of CSF leaks and skull base defects. 

Once the radiologic assessments were made, all cases were 

correlated with the intraoperative findings. 

RESULTS 

There were 13 women, 3 men, age range 16 to 83 years 

(average age 47 years). Defects were presumed to be postsurgical 

(6/16), post-traumatic (2 /16), spontaneous (5/16), 

tumor-related (2/16) or congenital (1/16). Preoperative CT 

identified 19 osseous defects in 16 patients, and 16/19 

defects were confirmed at surgery. Most defects were less 

than 10 mm in maximal dimension. Defects were seen at the 

ethmoid roof (7/19), or at cribriform plate/ethmoid roof 

(12/19). Three osseous defects identified on CT were not 

confirmed at surgery; these included defects of the ethmoid 

roof (1/3), inferolateral right sphenoid (1/3) and posterior 

wall of the left frontal sinus (1/3). Surgeons reported a trend 

to CT over-estimating the size of the defect, when compared 

to direct observation. 

15 

CONCLUSION 

Coronal and sagittal multiplanar reformations obtained from 

high-resolution CT using direct axial acquisitions performed 

remarkably well at preoperatively localizing skull base 

defects in patients who presented with CSF leaks. Skull base 

defects were most apparent on the coronal reformations. 

Although the bony defects that involved the posterior wall of 

the frontal sinus were not seen on the coronal images, these 

defects were well appreciated on sagittal reconstructions. 

When compared with intraoperative observation, CT overestimated 

the size of the bony defect in a significant portion of 

the cases, probably due to the fact that the skull base may be 

normally thin and even appear dehiscent without being associated 

with CSF leaks. We noted a trend of improved accuracy 

at estimating the size of the defect using the thinner 

slice acquisitions. 

KEY WORDS: Cerebrospinal, fluid, leak 


Evaluating Tumors and Tumor-Like Lesions of the Nasal 

Cavity, the Paranasal Sinuses, and the Adjacent Skull 

Base with Diffusion-Weighted MR Imaging 

White, M. L. · Zhang, Y. · Robinson, R. A. 

University of Iowa Carver College of Medicine 

Iowa City, IA 

PURPOSE 

The purpose of this study was to evaluate the utility of diffusion-weighted 

MR imaging in depicting the cellularity of 

tumors in the nasal cavity, the paranasal sinuses, and the adjacent 

skull base. We hypothesized that the most cellular tumors 

would have the lowest apparent diffusion coefficient (ADC) 

values. This information could be useful in the differential 

diagnosis of malignant tumors and benign tumor-like lesions. 


Twenty-one consecutive patients (12 men and 9 women, 

ranging in age from 8 to 87 years) with histologically proven 

malignant tumors or benign tumor-like lesions in the nasal 

cavity, the paranasal sinuses, and the adjacent skull base 

were examined using a 1.5 T MR unit. The b-values were all 

1000 s/mm 2 . The 16 malignant tumors included rhabdomyosarcoma 

(2), squamous cell carcinoma (8), malignant 

melanoma (2), lymphoma (1), and neuroblastoma (3). The 5 

benign lesions were mesenchymal proliferative process, glomus 

tumor, polyp, mucocele, and neuroma and fibrosis. 

Apparent diffusion coefficient maps were generated off-line 

on an ADW 4.0 GE workstation using FuncTool. Regions of 

interests were drawn in representative areas of the tumors. 

Mean ADC values were compared with cellularity derived 

from resected or biopsy material. 

RESULTS 

The mean ADC value in the malignant tumors was 10.06 

(range, 5.00 ~ 15.85) x 10 -4 mm 2 /s. By contrast, in the benign 

tumor-like lesions, the mean ADC value was 19.11 (range, 

12.27 ~25.20) x 10 -4 mm 2 /s. Despite having some overlap, the 

former was significantly lower than the latter (t-test, P < 

0.0005). In addition, ADC values were correlated inversely 

with the percentage of tumor cellularity. This relationship 

also was statistically significant. (Pearson correlation test, r 

= -0.623, P < 0.005). 

Monday

Monday 

CONCLUSION 

Our study revealed that the ADC values may be predictive of 

tumor cellularity and may be useful in differentiating malignant 

tumors from benign lesions. This information is not 

obtained with conventional MR imaging, and diffusionweighted 

MR imaging may play an important role in the 

workup and treatment of tumors of the nasal cavity, the 

paranasal sinuses and adjacent skull base. 

KEY WORDS: Sinonasal tumor, diffusion-weighted imaging, 

ADC 


Assessment of the Efficacy of Preoperative Embolization 

of Juvenile Angiofibromas with Respect to Blood Loss 

and Recurrence of Tumor 

Bapuraj, J. R. · Mitra, S. · Gupta, A. K. · Arya, V. · 

Khandelwal, N. · Suri, S. 

Postgraduate Institute of Medical Education and Research 

Chandigarh, INDIA 

PURPOSE 

To assess the efficacy of preoperative embolization of juvenile 

nasopharyngeal angiofibroma (JNA) with respect to 

blood loss during surgery and recurrence of tumor. 


Twenty-four clinically and radiologically diagnosed cases of 

JNA were included in this randomized and prospective 

study. Twelve patients were randomly selected for preoperative 

embolization, 12 other were not embolized. CT scans 

were done preoperatively in all patients. A repeat scan was 

obtained at 6-month follow up to assess recurrence. 

RESULTS 

Mean blood loss was 201.67 ml for embolized group while 

it was 788.75 ml for the nonembolized patients. This difference 

was found to be statistically significant (p < 0.01). The 

mean difference of Hb from preoperative to postoperative 

level was lower in the embolized group (0.9 gm/dl) than 

their nonembolized counterparts (2.1 gm/dl). The percentage 

rate of decrease of Hb from preoperative to postoperative 

level was lower in embolized patients (8.36) compared to the 

nonembolized ones (20.71). This difference was highly statistically 

significant (p < 0.01). Data were analyzed with 

respect to tumor grade. A significant difference was noted 

between the embolized and nonembolized patients with 

high-grade tumors but not between those with low-grade 

tumors. It is therefore, reasoned to infer that preoperative 

embolization of the branches of the external carotid artery 

appears to facilitate removal of high-grade tumors with minimal 

possible blood loss. The benefit of embolization in 

those with low-grade tumors is less obvious, probably due to 

less vascularity, and consequently, easier removal. 

Recurrence of tumor was defined as appearance of an 

enhancing mass in the review CT at 6 months in the sites of 

surgical exploration and removal. In the embolized group 

recurrence was present in 3 patients as compared to 5 patient 

in nonembolized group. However, reduction in the incidence 

of recurrences in the embolized group was statistically significant 

only when low-grade tumors were considered, not 

for high-grade tumors. Embolization and the technique of 

removal by endoscopy or open excision also were compared. 

16 

It was found that the blood loss following endoscopic 

removal averaged 229.37 ml, which is significantly less than 

that for open surgical excision (628.13 ml). This difference 

is statistically significant. 

CONCLUSION 

Preoperative angio embolization is useful for reducing intraoperative 

blood loss and the risk of tumor recurrence. Hence 

this may be adopted as a standard preoperative adjunct in all 

advanced cases of nasopharyngeal angiofibroma, referred 

for surgery. 

KEY WORDS: Angiofibroma – nasopharynx, neoplasms, 

embolization therapy 


Recurrent Aggressive Glomus Jugulare Tumor: 

Radiologic-Pathologic Correlation 

Johnson, M. H. · Moss, R. L. · Coehllo, D. · Martel, M. · 

Barry, C. · Sasaki, C. 


New Haven, CT 

PURPOSE 

An aggressive recurrent glomus jugulare tumor is presented 

and correlated with the intraoperative appearance as well as 

the gross, histologic, and immunohistochemical patholologic 

findings, in order to emphasize the importance of compartmental 

analysis of recurrent tumors prior to treatment. 


A 25-year-old female presented with recurrence of a glomus 

jugulare tumor initially diagnosed, surgically excised, and 

treated with adjuvant radiation therapy in 1998. Analysis of 

outside MR imaging and in-house contrast-enhanced CT 

showed recurrent tumor within the parapharyngeal, and 

carotid spaces, as well as within the internal jugular vein. No 

significant intracranial recurrence was demonstrated. 

Preoperative angiography demonstrated near occlusion of 

the ipsilateral vertebral artery, which provided neovascular 

supply to the superior aspect of the tumor bed at the skull 

base. The posterior division of the ascending pharyngeal 

artery sent large feeding branches to a large pedunculated 

tumor mass within the internal jugular vein. Anterior division 

of the ascending pharyngeal artery supplied a medial 

tumor compartment within the parapharyngeal space. 

Following embolization of the superior and lateral compartments 

of the tumor, the patient was taken to surgery, and the 

jugular vein exposed. The jugular vein containing tumor was 

removed “en-bloc” and directly corresponded with the preoperative 

imaging. The medial/anterior compartment was 

excised in the same operation. 

RESULTS 

Pathologic examination demonstrated paraganglioma (S-100 

negative) within the internal jugular vein, a 4 cm lateral neck 

tumor and a tumor infiltrated node within the parapharyngeal 

space. 

CONCLUSION 

Learning objectives: 1. Compartmental analysis of crosssectional 

imaging in recurrent glomus jugulare may demonstrate 

operability of seemingly unresectable tumors.

2. Angiographic compartments and selective embolization of 

tumor compartments facilitates resection. 3. The compartmentalized 

imaging and angiographic mapping correlates 

directly with the surgical findings and gross pathology. 

KEY WORDS: Glomus jugulare, radiologic-pathologic correlation, 

interventional 


10:00 AM - 11:38 AM 

Room 107 





Miscellaneous 



Moderators: Gary R. Duckwiler, MD 

Jacques G. Moret, MD 


Characterization of Cerebral Aneurysms for Assessing 

Risk of Rupture Using Patient-Specific Computational 

Hemodynamics Models 

Cebral, J. R. · Castro, M. A. · Burgess, J. E. · Pergolizzi, R. 

M. · Putman, C. M. 

Inova Fairfax Hospital 

Falls Church, VA 

PURPOSE 

The purpose of our study was to demonstrate the feasibility 

of using patient-specific 3 D rotational angiography image 

data from clinical studies to construct corresponding realistic 

patient-specific computational fluid dynamics models of 

cerebral aneurysms. Then using these models, characterize 

intraaneurysmal flow patterns and explore their possible 

associations to the clinical history of aneurysmal rupture. 


Realistic patient-specific anatomical models of cerebral 

aneurysms were constructed from 3 D rotational angiography 

images of 28 patients. The pulsatile flow simulations 

were performed using an implicit finite element formulation 

using physiologic flow conditions derived from phase-contrast 

MR measurements on a different normal subject. 

Visualizations of wall shear stress, oscillatory shear stress, 

peak pressure, and the intraaneurysmal flow velocity were 

produced for each aneurysm. The aneurysms then were classified 

into different categories depending on the distribution 

of mean wall shear stress (WSS), peak pressure (PP), oscil- 

17 

latory shear stress (OSI), complexity of the flow pattern and 

size of the flow impingement region. These categories then 

were correlated with the clinical history of hemorrhage. 

RESULTS 

Thirty-three aneurysms were studied in 28 patients ranging 

in age from 28 years to 85 years with a mean age of 49 years. 

Fourteen had documented subarchnoid hemorrhages that 

could be attributed to an intracranial aneurysm. These 

aneurysms consisted of 30 from the anterior circulation 

[internal carotid artery 17 (posterior communicating 3, cavernous 

3, paraclinoid 6, terminus 1, ophthalmic 2), anterior 

cerebral artery 2] and three from the posterior circulation 

aneurysms (basilar terminus 1, SCA 2). A total of 195 

anatomical and hemodynamic models were constructed and 

analyzed. No strong correlation with clinical history of rupture 

was observed for patterns of WSS, PP, or OSI. 

Unruptured aneurysms had large relative impingement zones 

in 88% and 92% of ruptured aneurysms had small relative 

impingement zones reaching a statistical significance using 

95% and 99% confidence intervals. The location of the 

inflow impingement showed a trend toward body and dome 

locations favoring unruptured aneurysms. Stable single vortex 

flow patterns were found in 75% of unruptured 

aneurysm, only reaching 90% confidence interval. Wall 

shear stress, PP, OSI, and impingement patterns were found 

to be independent of aneurysm size. Intraaneurysmal flow 

types show some positive correlation with size but all flow 

types were found in small and mid sized aneurysms. 

CONCLUSION 

Patient-specific realistic anatomical and hemodynamic visualizations 

of cerebral aneurysms can be used to characterize 

aneurysm flow dynamics for clinical studies. 

Intraaneurysmal flow patterns are influenced strongly by 

patient-specific geometry but only loosely correlating to 

aneurysm size. Small relative size of the inflow jet impingement 

zone to the aneurysm size is associated with a clinical 

history of aneurysmal rupture. 

KEY WORDS: Aneurysm, cerebral, hemodynamic models 

This paper was selected by the Eastern Neuroradiology 

Society (ENRS) to receive the Norman E. Leeds Award for 

best scientific paper at its 16th Annual Meeting held August 

20-22, 2004. Dr. Putman will be the presenter. 


Reconstruction of Normal Slipstream Flow in 

Intracranial Cerebral Arteries following Wide-Neck 

Aneurysm Embolization: Does Parent Vessel/Aneurysm 

Geometry Dictate Occlusion Device Choice? 

Imbesi, S. G. · Knox, K. · Kerber, C. W. 

University of California San Diego Medical Center 

San Diego, CA 

PURPOSE 

To evaluate and analyze the flow dynamics in intracranial 

parent vessel cerebral arteries following wide-neck 

aneurysm embolization with multiple iterations of parent 

vessel protection device placement. 

Monday

Monday 


We created multiple clear silicone elastic replicas of a sidewall 

and a bifurcation intracranial cerebral wide-neck 

aneurysm using the lost wax technique. The aneurysm dome 

measured 10 mm and the aneurysm neck measured 6 mm in 

both aneurysm types. The replicas were placed in a circuit of 

pulsatile non-Newtonian fluid and flows were adjusted to 

replicate human physiologic flow profiles, velocities, and 

volumes. Fluid slipstreams were opacified with isobaric dyes 

and images recorded using mini-DV digital video prior to 

aneurysm occlusion. Initially, coils were placed in the 

aneurysm sac without a parent vessel protection device. 

Then, coils were placed in identical copies of the aneurysm 

replica using nondetachable balloons, vascular stents, and 

new aneurysm neck bridge devices (Trispan) for parent vessel 

protection. Intravascular flow patterns then were evaluated 

again. 

RESULTS 

Prior to treatment, analysis of flow in the cerebral arteries 

and adjacent aneurysm showed undisturbed slipstreams parallel 

to the vessel sidewall in the parent vessel proximal to 

the aneurysm sac. Flow then entered the distal aneurysm 

neck, formed a vortex pattern within the aneurysm sac, and 

exited the proximal aneurysm neck. As flow reentered the 

adjacent vessel, disturbed slipstreams were created within 

the distal portion of the parent vessel. Aneurysm embolization 

was not possible (coil migration into the parent vessel 

lumen) without parent vessel protection nor with nondetachable 

balloon placement in the parent vessel across the 

aneurysm neck given the extremely wide aneurysm neck. 

Aneurysm embolization was only possible using intravascular 

stents and Trispan for parent vessel protection in both 

aneurysm types; however, optimal configuration (i.e., not 

only coil retention within the aneurysm sac but, specifically, 

the recreation of undisturbed parallel slipstream flow in the 

distal parent vessel) could only be achieved with the 

intravascular stent across the sidewall aneurysm neck and 

with the Trispan across the bifurcation aneurysm neck, but 

not vice versa. While the stent could hold the coil mass within 

the aneurysm sac in the bifurcation aneurysm even though 

the aneurysm neck was not covered completely and undisturbed 

slipstream flow was attained within the stented distal 

parent vessel branch, slipstream flow through the jailed distal 

parent vessel branch showed a continued markedly disturbed 

flow pattern. With the Trispan positioned across the 

neck of the sidewall aneurysm, the stem and proximal portion 

of the device protruded into the parent vessel lumen, 

which resulted in a continued markedly disturbed slipstream 

flow pattern within the distal parent vessel segment. 

Choosing the appropriate device based on the aneurysm 

geometry and eliminating this disturbed flow should 

improve cerebral perfusion, reduce the incidence of distal 

emboli, and possibly even prevent aneurysm reformation. 

CONCLUSION 

Use of clear silicone elastic replicas permits detailed analysis 

of individual intravascular flowing slipstreams. 

Knowledge of aneurysm flow dynamics and the recreation of 

normal parent vessel flow following aneurysm embolization 

should lead to a more efficacious and safer aneurysm obliteration. 

KEY WORDS: Vascular/intracranial, aneurysms, devices 

18 


Small Aneurysms Do Rupture: A Single Center 

Experience with Small Intracranial Aneurysms 

Shownkeen, H. · Hall, M. J. · Origitano, T. · Anderson, D. · 

Craig, E. 

Loyola University Medical Center 

Maywood, IL 

PURPOSE 

The 30-day mortality of aneurysmal subarachnoid hemorrhage 

(SAH) is approximately 45%. As a result of this high 

morbidity and mortality, preventing rupture is crucial and the 

basis of treating unruptured aneurysms. Determining the 

patient and aneurysm characteristics that result in increased 

SAH risk has been a topic of interest as the morbidity/mortality 

of treatment is not trivial. Prior studies, most notably 

the International Study of Unruptured Intracranial 

Aneurysms (ISUIA), have indicated a low rupture risk for 

small aneurysms. This review was compiled to highlight that 

small aneurysms do, however, rupture and in fact comprise a 

large percentage of aneurysms presenting with SAH at our 

institution. 


This is a retrospective review of patients with small 

aneurysms presenting at the author’s institution from 1997- 

2004. Aneurysms measuring less than or equal to 5 mm were 

defined as small. Medical records and radiologic studies and 

reports were assessed. Main areas of interest include the proportion 

of small aneurysms presenting with SAH, 30-day 

mortality, and the complications of intervention (both surgical 

and endovascular). 

RESULTS 

This study comprises a total of 111 small aneurysms in 104 

patients. Seventy-eight of these patients presented with SAH 

(75%). In 15 patients, unruptured small aneurysms presented 

subsequent to evaluation of a larger ruptured or unruptured 

intracranial aneurysm. The remaining aneurysms were 

discovered on incidental imaging studies. Interestingly, all 8 

patients with small posterior circulation aneurysms and 89% 

of posterior communicating artery aneurysms presented with 

SAH. All of these aneurysms were treated. The 30-day mortality 

and the morbidity/mortality of treatment will be evaluated 

and discussed in detail. 

CONCLUSION 

Results from prior studies have suggested the value of 7-10 

mm as the critical size for rupture. Data from the ISUIA suggests 

a very low risk of aneurysm rupture for UIAs less than 

10 mm. In a study evaluating patients with rupture of a previously 

unruptured aneurysm by Yasui, et al., however, a 

substantial number were less than 5 mm in size at initial 

diagnosis. A large proportion of aneurysms presenting at the 

author’s institution were less than or equal to 5 mm. Small 

aneurysms account for approximately 32% of all treated 

aneurysms at our institution. The majority of these 

aneurysms presented with SAH (especially when originating 

from the posterior communicating artery and posterior circulation). 

In fact of all 194 aneurysms presenting with SAH 

from 1997-2004, 78 were less than 5 mm (39.6%). 

Unfortunately, to date no small aneurysms have been followed 

prospectively at our institution and therefore the per-

centage of small aneurysms at risk of rupture cannot be 

determined. In light of the ISUIA study, which included a 

large amount of patients, this raises an interesting treatment 

dilemma. If small aneurysms tend to rupture at a low rate, 

why do these aneurysms constitute such a large proportion of 

aneurysms presenting with SAH at our institution and others? 

Additionally due to the historical risks of therapy, when 

is treatment of unruptured small aneurysms warranted? We 

will discuss our approach to treatment and the 

morbidity/mortality of small aneurysm therapy in detail. 

KEY WORDS: Aneurysm, small, intracranial 


Correlation between Angiographic and Histologic 

Features after Coil Embolization in Elastase-Induced 

Aneurysm in Rabbits 

Ding, Y. · Dai, D. · Lewis, D. A. · Danielson, M. A. · 

Kadirvel, R. · Cloft, H. J. · Kallmes, D. F. 

Mayo Clinic 

Rochester, MN 

PURPOSE 

To explore the relationship between angiographic and histologic 

features after coil embolization in elastase-induced rabbit 

aneurysms. 


One hundred six elastase-induced rabbit aneurysms 

embolized with various platinum coils were analyzed retrospectively. 

Digital subtraction angiography (DSA) was performed 

after embolization and before sacrifice. Aneurysms 

were harvested at 2 (n = 14), 4 (n = 33), 10 (n = 11), 16 (n = 

16), and 24 weeks (n = 32). H&E-stained slides were analyzed. 

Follow-up angiographic results (progressive occlusion, 

stable, recanalization, and coil compaction) were evaluated 

by comparing the aneurysm occlusion feature immediately 

after embolization and before sacrifice. Histologic features 

included the extent of thrombus organization (incomplete 

or complete thrombus organization) and shape of interface 

between aneurysm and parent artery (concave, flat, and 

convex). Concave interface was regarded as the sign of coil 

micro-compaction. Comparison between angiographic and 

histologic features was performed using Chi-Square test. 

RESULTS 

Seventy-six (72%) of 106 aneurysms remained angiographically 

stable at follow up, in which incomplete thrombus 

organization and micro-compaction were found in 44 (58%) 

and 42 (55%) of these cases, respectively. Twenty-four 

(23%) of 106 aneurysms showed coil compaction/recanalization 

angiographically, in which incomplete thrombus 

organization and micro-compaction were found in 21 (88%) 

and 22 (92%) of these cases, respectively. Progressive occlusion 

was seen in six cases, half of which had incomplete 

organization and micro-compaction. More coil micro-compaction 

(49/68, 72%) was found in aneurysms with incomplete 

thrombus organization, compared with aneurysms with 

complete thrombus organization (18/38, 47%) (p < .05). 

Complete thrombus organization was found more frequently 

in angiographically stable or progressively occluded 

aneurysms as compared to aneurysms showing angiographic 

compaction (58% versus 12%, respectively, p < .01). Micro- 

19 

compaction was found more frequently in angiographically 

recanalized aneurysms, compared with aneurysms showing 

stable or progressive occlusion aneurysms (92% versus 55%, 

respectively, p < .001). Coil compaction seen histologically 

was more prevalent than that seen angiographically (63% 

versus 23%, respectively, p < .01). 

Fig. A. Embolized aneurysm before sacrifice intraarterial 

DSA right anterior oblique view, demonstrating complete 

aneurysm occlusion. 

Fig. B. Aneurysm as in A showing concave interface 

between aneurysm and parent artery, indicating coil microcompaction. 

(H&E stain, original magnification, x 1.4). 

CONCLUSION 

In elastase-induced aneurysms, angiographically evident 

recanalization occurs in approximately one fourth of cases 

while histologic recanalization occurs in approximately two 

thirds of cases. Progressive occlusion is rare. Incomplete 

thrombus organization correlates with both angiographic and 

histologic compaction, suggesting that mechanical stabilization 

of aneurysms is important in healing. 

KEY WORDS: Aneurysm; embolization, histology; angiography, 

rabbit 


Unusual Increased Cortical Density in Immediate 

Postembolization CT of Patients with Intracranial 

Aneurysms 

Ozturk, A. · Saatci, I. · Erdogan, C. · Akmangit, I. · Pamuk, 

A. G. · Cekirge, H. S. 

Hacettepe University Hospitals 

Ankara, TURKEY 

PURPOSE 

The purpose of this study is to evaluate the “focal cortical 

hyperdensity” finding observed in immediate postembolization 

CT of some patients with intracranial aneurysms after 

uneventful endovascular treatment. 

Monday

Monday 


This study consisted of 93 consecutive patients with 100 

intracranial aneurysms, ruptured or unruptured, in whom a 

cranial CT was obtained following the endosaccular 

aneurysm treatment. Simultaneous MR imaging as well as a 

repeat CT after 4-6 hours were obtained in the patients with 

abnormal finding in the initial CT. All patients had CT 

immediately before the treatment. The study group consisted 

of 74 aneurysms which were treated using balloon remodelling 

technique and 26 aneurysms treated with no balloon 

assistance. The patients who were treated with parent artery 

occlusion, intraextra cranial by-pass surgery associated with 

endovascular method or who had intraprocedural apparent 

aneurysm rupture, thrombus formation or distal emboli, 

major catheter induced vasospasm were excluded as well as 

the patients who had acute or subacute infarct in the relevant 

territory on preembolization CT. 

RESULTS 

Cranial CT showed focal cortical hyperdensity (Fig. 1) in 40 

of 74 aneurysms (54%) which were treated with the aid of 

remodelling balloon, and 9 of 26 (34.6%), treated with no 

balloon assistance. Cortical hyperdensity was confined to the 

territory of the parent artery harboring the aneurysm and 

resolved in the repeated CT (Fig. 2). None of the patients 

were symptomatic. A statistically significant relationship 

was found between presence of this finding and increase in 

microcatheter time, increase in number of balloon inflation 

in patients who were treated with balloon assistance and also 

increase in total amount of contrast material in the entire 

group. 

CONCLUSION 

Focal cortical hyperdensity may occur secondary to changes 

in blood-brain barrier secondary to hemodynamic effects of 

balloon inflation and temporary cesation of blood flow. 

Augmented contrast enhancement in the cortex is probably 

due to increase in blood-brain barrier permeability secondary 

to reperfusion. The awareness of this finding is crucial to differentiate 

this clinically insignificant finding from possible 

hemorrhage that will affect the patient’s immediate postprocedural 

medical management (i.e., anticoagulation and/or 

antiaggregan treatment). 

KEY WORDS: CT, intracranial aneurysms, endovascular 

treatment 

20 


Long-Term Histopathologic and Immunohistochemical 

Comparison in Human, Rabbits, and Swine Aneurysms 

Embolized with Platinum Coils 

Dai, D. · Ding, Y. · Danielson, M. A. · Kadirvel, R. · Lewis, 

D. A. · Clot, H. J. · Kallmes, D. F. 

Mayo Clinic 

Rochester, MN 

PURPOSE 

To compare the histologic and immunohistochemical findings 

in experimental rabbit and swine aneurysms to that of a 

human aneurysm embolized with platinum coils, in order to 

clarify the cellular mechanisms of healing in these species. 


Elastase-induced aneurysms in rabbits (n = 6, harvested at 24 

weeks) and sidewall aneurysms in swine (n = 5, harvested at 

12 weeks) were created and embolized. A single human 

aneurysm that had been coiled 6 years prior to death was harvested 

following autopsy. Specimens were embedded in 

paraffin and stained with hematoxylin and eosin and Masson 

Trichrome staining. Immunohistochemical and immunofluorescence 

staining were performed for alpha smooth muscle 

actin (SMA), vimentin, and CD31. 

RESULTS 

The human aneurysm was filled with homogeneous, faintly 

staining, hypocelluar, amorphous tissue. Vimentin and SMA 

double immunofluorescent staining showed scattered, sparse 

cells within dome positive for vimentin but negative SMA, 

indicating chronic inflammatory cells or fibrocytes (Fig. 

1A). A thin layer of hypocellular tissue traversing the entire 

neck was observed, lined with a single layer of flattened cells 

that were positive for CD31, indicating endothelial cells. 

Collagen deposition was absent within aneurysmal dome 

and along the neck. Five of six rabbits showed loose, 

hypocelluar tissue, with faintly stained, amorphous matrix 

within the dome. Vimentin and SMA double immunofluorescence 

staining within dome showed scattered, sparse cells 

within dome positive for vimentin but negative SMA (Fig. 

1B). A thin layer of hypocellular tissue across the neck was 

observed in these five aneurysms, which was lined with single 

layer of CD31-positive endothelial cells. Collagen deposition 

was absent within aneurysmal dome and along the 

neck in all six aneurysms. Swine aneurysm samples showed 

highly vascularized fibrous tissue within the dome, with 

abundant, dense collagen bundles. The cells within this 

fibrous matrix were positive for SMA and vimentin, indicating 

myofibroblastic lineage. Light microscopy showed thick, 

dense, hypercellular tissue across the entire neck, positive 

for vimentin and SMA, indicating a well developed neointima.

Photomicrographs of human brain aneurysm (A) and rabbit 

aneurysm (B) show scattered, sparse cells within the 

aneurysm dome, which are positive for vimentin and negative 

for SMA. (Immunofluorescence double staining with 

antibodies for SMA and vimentin.) 

CONCLUSION 

The histopathologic reaction both in the dome and across the 

neck of rabbit, elastase-induced aneurysms treated with platinum 

coils is similar to that of human aneurysms. Swine 

aneurysms demonstrate dissimilar healing characteristics as 

compared to both rabbit and human aneurysms. 

KEY WORDS: Aneurysm; coil embolization, histopathology, 

immunohistochemistry 


Immunophenotypic Evolution of Cells Involved in 

Healing of Rabbit Aneurysms Embolized with Platinum 

Coils 

Dai, D. · Ding, Y. · Kadirvel, R. · Lewis, D. A. · Cloft, H. J. 

· Kallmes, D. F. 

Mayo Clinic 

Rochester, MN 

PURPOSE 

To demonstrate the immuophenotypic evolution over time of 

cells involved in the healing process of experimental rabbit 

aneurysms embolized with platinum coils. 


Elastase-induced aneurysms were created and embolized 

with platinum coils in rabbits. Aneurysms were collected at 

2 weeks (n = 2), 4 weeks (n = 6), and 24 weeks (n = 6) following 

embolization. All specimens were embedded in 

paraffin and stained with hematoxylin and eosin and 

immunohistochemistry. Multiple antibodies, including alpha 

smooth muscle actin (SMA), myosin, desmin, vimentin, and 

CD31, were employed. 

21 

RESULTS 

At two weeks, the majority of the aneurysmal lumen was filled 

with fresh thrombus. Scattered, sparse, spindled or satellitelike 

cells, distributed at the periphery of aneurysm lumen were 

noted, which were positive for SMA, myosin, and vimentin, 

indicating myofibroblastic differentiation. No CD31 positive 

cells were identified along the neck. Four weeks following 

embolization, most of the aneurysm lumens were filled with 

loose tissue, associated with small areas of poorly organized 

thrombus. All spindle cells within the loose tissue were strongly 

positive for SMA (Fig. 1A), with moderate reaction for 

myosin and vimentin, indicating myofibroblastic immuophenotype. 

Thin layers of fibrin mixed with scattered SMA-positive 

spindle cell were present along the aneurysm neck in five 

of six aneurysms. One of six aneurysms, with a very narrow 

neck, showed a thin layer of loose connective tissue lined with 

a single layer of CD31-positive endothelial. Twenty-four 

weeks following embolization, the aneurysm lumens were 

occupied completely with loose, mesh-like tissue. Most thinspindle 

cells within the loose tissue were negative for SMA 

(Fig. 1B), myosin and desmin, but positive for vimentin, indicating 

that SMA-positive myofibroblasts present at earlier time 

points had differentiated into fibrocytes. Thin layer of organized 

tissue lined with CD31 positive endothelial cells along the 

neck was found in five of six aneurysms. 

Photomicrographs of rabbit aneurysms embolized with platinum 

coils that shows spindled cells within the aneurysm 

dome at 4 weeks (A) are positive for SMA, but are negative 

for SMA at 24 weeks (B). Immunohistochemical staining 

with antibodies for SMA. (Original magnification x 100) 

CONCLUSION 

Myofibroblasts, the key cellular component involved the healing 

of rabbits aneurysms embolized with platinum coils, gradually 

lose their contractile elements over time. The resultant 

tissue is comprised of fibrocytes within a loose, connective tissue 

matrix. These findings may allow more rational design of 

modified coils aimed at improving aneurysm healing. 

KEY WORDS: Aneurysm; coil embolization, cell 

immunophenotype, rabbit 

Acknowledgment: This project was supported by NIH grant 

NS42646. 

Monday

Monday 


Histologic Comparison of HydroCoil Embolization 

System in Embolization of Aneurysms in Rabbits, 

Canines, and Humans 

Kallmes, D. F. 1 · Cloft, H. J. 1 · Cruise, G. 2 

1 2 Mayo Clinic, Rochester, MN, MicroVention, Inc., Aliso 

Viejo, CA 

PURPOSE 

To compare the histological features of experimental and 

clinical aneurysms embolized with HydroCoils. 


Rabbit elastase, canine bifurcation, and canine sidewall 

aneurysms were prepared according to the method of Altes 

(1), Graves (2), and German (3), respectively. The aneurysms 

were packed as densely as possible with HydroCoils. Followup 

evaluations ranged from 2 to 13 weeks. Human intracranial 

aneurysms were collected at autopsy, ranging from 2 to 4 

weeks postprocedure. All aneurysms were fixed in 10% neutral 

buffered formalin, embedded in methyl methacrylate, sectioned, 

and stained with hematoxylin and eosin. Outcome 

measures included: Durability [major compaction (0), minor 

compaction (1), stable (2), progressive occlusion (3)]; Sac 

organization [unorganized (0), mostly unorganized (1), mostly 

organized (2), organized (3)]; and Neck coverage [none (0), 

acellular (1), thin (2), thick (3)]. Total scores were obtained by 

summing the three component scores. Additionally, inflammation 

was compared between the experimental and human 

aneurysms. 

RESULTS 

At 2 weeks, rabbit elastase (n = 9) aneurysms were scored 4.7 

± 0.7. Two-week data were not available in the canine bifurcation 

or sidewall models. At 4 weeks, rabbit elastase (n = 9), 

canine bifurcation (n = 8), and canine sidewall (n = 4) 

aneurysms were scored 7.7 ± 0.8, 8.1 ± 0.8, 8.5 ± 0.6, respectively. 

At 10-13 weeks, rabbit elastase (n = 9), canine bifurcation 

(n = 14), and canine sidewall (n = 4) aneurysms were 

scored 7.2 ± 0.4, 7.9 ± 1.1, 8.0 ± 0.0, respectively. Two human 

aneurysms were harvested at 2 weeks postprocedure. Both 

aneurysms were scored at 5. Two other human aneurysms 

were harvested at 4 weeks postprocedure. One of these was 

scored a 4 and the other scored a 5. Significant inflammation 

was not observed in any experimental or human aneurysm. 

Mild degrees of inflammation generally was observed at short 

follow ups and subsided at longer follow-up times. 

22 

CONCLUSION 

In all three experimental aneurysm models, healing generally 

was completed at 4 weeks postprocedure, with stable or progressive 

occlusion, thin or thick neck coverings, and mostly 

organized to completely organized sacs. The tissue response did 

not change in the time course evaluated in this study, 10-13 

weeks. Rabbit and human aneurysms were similar at 2 weeks 

postprocedure. At 4 weeks, experimental aneurysms tended 

toward greater healing than the human aneurysms. Inflammation 

was equivalent, and minimal, in the experimental and human 

aneurysms. The development of experimental aneurysms that 

heal in a slower, more clinically relevant manner would improve 

the development of new aneurysm treatment therapies. 

REFERENCES 

1. AJR Am J Roentgenol 2000;174:349-354 

2. AJNR Am J Neuroradiol 1993;14:801-803 

3. N Engl J Med 1954;250:104-106 

KEY WORDS: Aneuysm, rabbit, hydrocoil 


Vascular Remodeling in Healing of Swine Aneurysm 

Embolized with Platinum Coils 

Kadirvel, R. · Ding, Y. · Dai, D. · Danielson, M. A. · Lewis, 

D. A. · Kallmes, D. F. 

Mayo Clinic College of Medicine 

Rochester, MN 

PURPOSE 

Endovascular embolization using platinum coils has been 

widely used as a alternative therapy for the treatment of 

intracranial aneurysms. But the mechanism of aneurysm 

healing is poorly understood. The purpose of this study was 

to characterize the vascular remodeling in swine aneurysms 

after embolization with platinum coils. 


Five side-wall aneurysms were created microsurgically in common 

carotid arteries in swine. These aneurysms were embolized 

immediately after creation using platinum coils by endovascular 

means. Angiography was performed prior to and immediately 

after embolization and at the time of tissue harvest. After 12 

weeks of implantation, aneurysm samples were collected for histologic 

and biochemical analysis. Normal venous tissue was 

used as control tissue for protein expression studies. 

RESULTS 

All five aneurysms were completely occluded angiographically 

at the time of embolization and at follow up. A fibrous 

membrane, on the order of 2000µ thick was seen at the neck 

of all five aneurysms. Three of five aneurysms showed highly 

vascularized fibrous tissue filled the aneurysms dome, 

with abundant, dense, disorganized collagen bundles. The 

cell bundles within fibrous matrix were positive for smooth 

muscle actin (SMA) and vimentin. Two of five aneurysms 

demonstrated dense chronic inflammatory tissue, primarily 

consisting of lymphocytes, histocytes, giant cells and thinwalled 

vessels within the dome. Thick, dense, hypercelluar 

tissue across the entire neck was observed in all aneurysms. 

Cells within this neointima were positive for SMA and 

vimentin. The expression of both MMP-2 (pro- and active 

forms) and MMP-9 (pro- and active forms) were found to be

high in aneurysms, whereas expression of their tissue 

inhibitors (TIMP-1 and - 2) was diminished significantly 

when compared with controls. VEGF and VCAM-1 expression 

were elevated in aneurysms. No significant changes in 

the expression of TGF -β and eNOS were observed. 

CONCLUSION 

Even in healed aneurysms in this model, ongoing vessel 

remodeling may be present, as evidenced by over expression 

of metalloproteinases and diminished expression of their 

inhibitors. Angiogenesis or vasculogenesis as well as inflammation 

are prominent in healing aneurysms in swine. 

KEY WORDS: Aneurysm, embolization, vascular remodeling 


Recanalization after Endovascular Treatment of 

Intracerebral Aneurysms 

Papanagiotou, P. · Grunwald, I. Q. · Roth, C. · Struffert, T. · 

Gül, G. · Politi, M. · Reith, W. 

Clinic for Diagnostic and Interventional Neuroradiology 

Homburg, GERMANY 

PURPOSE 

The aim of his study was to evaluate the risks of endovascular 

therapy, aneurysm regrowth, recanalization, and the need 

for reembolization. 


A prospective analysis was performed on 200 endovascularly 

treated aneurysms from 2000 to 1/2004. Seventy-nine 

asymptomatic and 121 ruptured aneurysms were treated. The 

risk of endovascular therapy, aneurysm regrowth, recanalization, 

and the need for reembolization was evaluated. 

RESULTS 

Mean observation time was 10 months. Complete occlusion 

(100%) in the initial intervention was achieved in 171 patients 

(85.5 %) an 80-95% occlusion rate in 21 aneurysms (10.5%). 

In 6 cases the occlusion rate was < 80% (4 %). One hundred 

and eighteen aneurysms were controlled after 10 months. 

Recanalization rate of all aneurysms in the first follow up was 

27/118 (22.8%). Eighty-three of 103 aneurysms (80.6%) with 

initial 100% occlusion rate remained completely occluded. 

Thirteen of 103 (12.6%) showed recanalization and 7/103 

(6.8%) a neck regrowth. In the group with 80-95% occlusion 

14 patients were reassessed: 1 showed spontaneous occlusion, 

in 7 cases (50%) the initial neck remained, in 6 cases (42.8%) 

recanalization increased. In one case the initial < 80% occlusion 

remained unchanged, in one case it increased. There was 

correlation between initial occlusion rate and recanalization. 

Bigger aneurysm size was associated with higher occurrence 

of recanalization. Aneurysm localization did not influence 

recanalization. Ten of 118 (8.4%) aneurysms were recoiled. 

From these one was the < 80% occluded aneurysm, 6 were 

from the 80-95% group, and the other 3 from the initially 

totally occluded group. 

CONCLUSION 

One in four aneurysms will show a recurrence. Initial occlusion 

rate seems to have an influence on the recanalization rate. 

KEY WORDS: Aneurysms, endovascular, recanalization 

23 


Efficacy of 3D Time-of-Flight MR Angiogram for a 

Follow Up of Embolized Cerebral Aneurysms 

Saguchi, T. · Murayama, Y. · Ishibashi, T. · Ebara, M. · Irie, 

K. · Takao, H. · Abe, T. 

Jikei University School of Medicine 

Tokyo, JAPAN 

PURPOSE 

A follow up of the embolized cerebral aneurysm with 

Guglielmi detachable coils (GDC) was performed mainly 

using craniograms and digital subtraction angiograms (DSA) 

so far. Recently, several authors have reported about an efficacy 

of the time-of-flight (TOF) MR angiogram (MRA) as a 

follow up after the embolized cerebral aneurysms. From the 

beginning of February 2004, we have been used 3D TOF 

MRA images for a less invasive and a useful follow up after 

GDC embolization. Then, they were compared with 3D DSA 

images that were obtained after GDC embolization on the 

3D workstation. We examined a usefulness of 3D TOF MRA 

as a follow up after GDC embolization. 


On the day after GDC embolization, 3D TOF MRA was performed 

for 50 patients. And as a follow up at our outpatient 

clinic, 3D TOF MRA was performed for 35 patients. We 

compared the 3D images between 3D TOF MRA and 3D 

DSA images after the embolized aneurysms. 

RESULTS 

There was a close correlation about the postoperative findings 

between 3D TOF MRA images that performed on the day after 

the operation and 3D DSA images that performed right after 

the operation. It was revealed that a recanalization was seen on 

3D TOF MRA images for 4 patients among thirty-five cases 

of the follow-up patients. In some cases, partial defects of vascular 

images were seen on 3D MRA images that were caused 

by an artifact of embolized coils. Those findings were never 

seen on 3D DSA images. Those were thought to be a limitation 

of a diagnosis using 3D TOF MRA. 

Monday

Monday 

CONCLUSION 

The findings of 3D images between 3D TOF MRA and 3D 

DSA that were reconstructed on the 3D workstation were 

closely correlated. Three-dimensional TOF MRA was very 

useful for a less invasive method of a diagnosis for a recanalization 

of the embolized aneurysms. 

KEY WORDS: 3D time of flight, MR angiography, aneurysm 


Paraplegia after Acute Intracranial Subarachnoid 

Aneurysm Rupture: Extention of the Subarachnoid 

Hemorrhage into the Lumbosacral Region 

Lee, S. · Klurfan, P. · Willinsky, R. A. · Gunnarsson, T. · 

terBrugge, K. G. 

The Toronto Western Hospital, University Health Network 


PURPOSE 

Extension of an intracranial subarachnoid hemorrhage into 

the lumbosacral region causing focal neurologic deficits is 

extremely rare. This report focuses on a patient with an 

acutely ruptured right posterior communicating artery 

aneurysm and extensive subarachnoid hemorrhage (Fischer 

grade IV) who presented with paraplegia. 


In addition to headache and a brief history of loss of consciousness, 

this 49-year-old female complained about acute 

onset of paraplegia (motor power; 0/5), hypoesthesia of both 

lower extremities and bilateral sciatica. There was no history 

of either lumbar puncture or trauma. 

RESULTS 

Her aneurysm was treated successfully with endovascular 

techniques and subsequent spinal MR imaging revealed a 

significant amount of subarachnoid blood in the lumbosacral 

region. Her paraplegia has improved gradually with conservative 

treatment and her 3-month follow-up neurologic 

examination revealed normal motor power in the lower 

limbs and minimal paresthesia in her right lower leg. A 3month 

follow-up spinal MR imaging demonstrated complete 

resolution of the subarachnoid hemorrhage in the lumbosacral 

region. 

CONCLUSION 

Although it is very rare, intracranial subarachnoid hemorrhage 

extension into the spinal canal in particular lumbosacral 

region should be included in the differential diagnosis 

of paraplegia after acute subarachnoid hemorrhage. 

KEY WORDS: Subarachnoid hemorrhage, paraplegia 

24 


Differents Stages of Endovascular Reconstruction of 

Intracranial Aneurysms with Stents 

Lylyk, P. · Haas, L. J. · Miranda, C. · Ferrario, A. · Pabon, B. 

· Musacchio, A. 

ENERI 

Buenos Aires, ARGENTINA 

PURPOSE 

The technique of endovascular reconstruction has evolved in 

the last years and new devices are now available. After 

development of balloon expandible stents for coronary use, 

the techniques have evolved. We still need a device that can 

fullfill the ideal conditions for intracranial circulation so that 

the risks and complications of the technique can be diminished. 

Based on the experience in animal and in vitro laboratory 

and the clinical experience with balloon expandible 

and autoexpandible stents, the authors analyzed the features 

of the appropriate stent for intracranial circulation. 


Between June 1996 and June 2004, 232 consecutive patients 

with sacular, disecant or fusiform aneurysms, were selected 

for intracranial stenting. This work was divided into four 

steps: 1. Computation models; 2. In vitro study: To analyze 

the stents in laboratory with relation to their navigability and 

flexibility, radiopacity; 3. In vivo (animal) study: To analyze 

anatomy and stents adaptability, radiopacity, navigability, 

delivery system, in the external carotid artery in animal laboratory; 

4. Clinical: 232 patients with intracerebral aneurysm 

were treated in our institution. Mean age was 49 years (range 

11-72 years). They were divided into 2 groups. Group 1: 

Coronary stent experience: included patients with balloon 

expandible stents. Group 2: Autoexpandible stent intracranial: 

included patients with autoexpandible stents. All 

patients had CT or MR scan and DSA preprocedure. Balloon 

expandible stents were implanted in 135 patients and autoexpandible 

stents in 97. Aneurysms group: 78.5% received 

combined treatment (stent and coils). Twenty-one and onehalf 

percent of patients had a complete occlusion of the 

aneurysms only with stents. 

RESULTS 

The patients were discharged under a rigid protocol with 

aspirin and clopidogrel treatment 72 hours before procedure 

and clopidogrel for an additional 90 days and aspirin continuously. 

The morbidity was 9.5% and mortality was 4.2%. 

CONCLUSION 

This series indicates that treatment with stents is a factual, 

effective, and safe technique for endovascular reconstruction. 

The selection of the type of stent depends on each case 

in particular and on the aneurysm selected. Technical 

improvements on delivery systems are mandatory to obtain 

better results during deployment phase. Preprocedural regimen 

of antiplatelet agents 72 hours before procedure is 

important in all patients. More trials and long-term follow up 

are needed to define the precise indications and to determine 

permanent vessel patency and aneurysms occlusion rate. 

KEY WORDS: Aneurysms, endovascular reconstruction, 

intracranial stents


10:00 AM - 11:30 AM 

Room 205 

(7d) ADULT BRAIN: Vascular; 

Extracranial 




Miscellaneous 



Moderators: J. Pablo Villablanca, MD 



Bilateral Vertebral Artery Duplication Demonstrated by 

MR Angiography of the Neck: Its Embryologic Origins 

and Implications 

Omojola, M. F. · Ionete, C. 

Creighton University Medical Center 

Omaha, NE 

PURPOSE 

To report a case of bilateral duplication of the proximal vertebral 

artery and describe the embryologic origins and clinical 

implications. 


An 83-year-old male was referred to us for cranial MR imaging 

with MRA of the head and neck because of mild cognitive 

impairment. There was no other significant clinical 

abnormality or significant past medical history. 

RESULTS 

His cranial MR images obtained on a 1.5 T system showed 

no significant abnormality other than incidental finding of a 

small right middle cranial fossa arachnoid cyst. Threedimensional 

time of flight cranial MRA is normal. The 2D 

time of flight neck MRA and 3D acquisition during intravenous 

gadolinium enhancement reveal bilateral proximal 

duplication of the vertebral arteries (Fig 1). A left proximal 

internal carotid artery with severe ulcerated stenosis also is 

present. On the right side, there are two separate origins of 

the vertebral artery from the right subclavian artery: one origin 

emanates proximally beyond the take off of the subclavian 

artery, makes a short loop and enters the carotid space 

staying in a medial relationship to the right common carotid 

artery while the second origin emanates from the subclavian 

artery adjacent to the origin of the right internal mammary 

artery, coursing straight and enters the foramen transver- 

25 

sarum at C7 vertebral level. Both vessels join in the foramen 

transversarum at C4-5 disk level to continue as the right vertebral 

artery. On the left side, there are two separate origins 

of the vertebral artery from the left subclavian artery: one 

emanates adjacent to the origin of the internal mammary 

artery, loops slightly forward and courses straight up behind 

the left common carotid artery while the other origin 

emanates from the subclavian artery just distal to the first 

and slightly larger enters the foramen transversarum at C7. 

Both vessels join in the foramen transversarum at C5-6 disk 

level to continue as the left vertebral artery. 

CONCLUSION 

This is the only report of proximal bilateral vertebral artery 

duplication. They were discovered by serendipity and probably 

has no known neurologic consequence. It is due to persistence 

of the fifth and sixth intersegmental arteries bilaterally. 

KEY WORDS: Vertebral artery duplication, MR angiography 


Pseudoaneurysm of the Anterior Spinal Artery in a 

Patient with Moyamoya: An Unusual Cause of 

Subarachnoid Hemorrhage 

Walz, D. M. · Woldenberg, R. · Setton, A. 

North Shore University Hospital 

Manhasset, NY 

PURPOSE 

To describe a pseudoaneurysm of the anterior spinal artery in 

a patient with moyamoya as the cause of recurrent subarachnoid 

hemorrhage and quadriplegia. 


A 58-year-old male presented with recurrent episodes of subarachnoid 

hemorrhage and subsequent quadriplegia. Initial 

diagnostic catheter angiogram of the intracranial circulation 

revealed diffuse occlusive vascular disease compatible with 

moyamoya and no definite associated aneurysm. As a result 

of progressively developing quadriplegia, MR imaging of 

the cervical region was obtained and revealed a vascular 

mass compressing the cervical cord with associated 

intraspinal hematoma. CT angiography confirmed a dense 

homogeneously enhancing vascular structure as the cause of 

Monday

Monday 

cord compression. Catheter angiography then was repeated 

with attention to the extracranial carotid and vertebral circulations. 

Findings included multivessel occlusion at the level 

of the skull base with extensive dural to pial collaterals. 

Occlusion of both vertebral arteries at the C1 level was 

noted. No intracranial aneurysms were identified. An 

enlarged anterior spinal artery was found off the left supreme 

intercostal artery that subsequently filled the basilar artery 

and both posterior cerebral artery territories. A 1.1 x 1.1 cm 

irregular bilobed pseudoaneurysm of the enlarged anterior 

spinal artery was noted with surrounding mass effect compatible 

with the above-mentioned hemorrhagic mass noted 

on the cross-sectional imaging studies (Fig. 1). 

RESULTS 

Successful endovascular occlusion of the pseudoaneurysm 

was achieved. 

CONCLUSION 

We present a highly unusual cause of subarachnoid hemorrhage 

in a patient with intra and extracranial occlusive vascular 

disease. Our case highlights the often rich and unique 

collateral networks that develop in association with occlusive 

disease as well as the need to evaluate the spinal axis 

and its vascular supply in cases of subarachnoid hemorrhage 

with a confounding clinical picture and without a definitive 

source of hemorrhage. 

KEY WORDS: Subarachnoid hemorrhage, moyamoya, 

pseudoaneursym 

26 


Prospective Evaluation of Complications in 23,649 

Consecutive Cases of Diagnostic Cervicocerebral 

Angiography 

Kaufmann, T. J. · Huston, J. · Mandrekar, J. · Schleck, C. · 

Kallmes, D. F. 

Mayo Clinic 

Rochester, MN 

PURPOSE 

Cervicocerebral angiographic imaging continues to shift 

from catheter angiography to noninvasive means. With the 

decreasing size of patient cohorts undergoing catheter 

angiography, the ability to carry out studies of the complications 

associated with the test will diminish. We report herein 

the largest study to date regarding the adverse event rate in 

cervicocerebral angiography. 


We prospectively have completed data forms on all patients 

receiving diagnostic cervicocerebral angiography at our 

institution, a tertiary and quaternary academic medical center, 

from 1981 through 2003. The data fields on the form 

include demographic data, comorbidities, indication for the 

examination, access site, procedure duration, number of 

catheters employed, type and volume of iodinated contrast 

material employed, presence or absence of trainee operator, 

and type of any neurologic, systemic, or local complication 

of the procedure. Clinical follow up was performed by a 

radiology nurse at 24 hours after every procedure, either by 

visiting inpatients and searching the medical record or by 

telephone interview with outpatients. Complications were 

recorded if they occurred within 24 hours of the angiogram, 

regardless of whether it was thought that the angiogram 

directly contributed to the complication. Descriptive statistics 

were generated of patient and procedural characteristics 

and of procedural complications. Univariate and multivariate 

analysis of complications was performed. 

RESULTS 

Records were reviewed for 23,649 consecutive cases, 

involving 20,015 patients. Mean patient age was 53, and 

55% were male. The most common indication for imaging 

was cerebrovascular disease (36%). Comorbidities included: 

hypertension (31%), serum creatinine greater than 1.2 

(16%), and frequent TIAs (10%). Retrograde, femoral arterial 

puncture was used in 98% of cases. A single catheter, 

alone, was used in 70% of cases. Mean procedure duration 

was 67 minutes. Access site hematoma occurred in 4%, and 

hematoma requiring surgery occurred in 0.08% of cases. 

Nausea/vomiting/hypotension occurred in 1%, anaphylaxis/circulatory 

collapse in 0.03%, and death in 0.05% (n = 13 

deaths; 11 of these related to neurologic conditions). 

Neurologic deficits occurred in 2.2% of cases (n = 523): 

transient in 1.8% (n = 419), reversible in 0.31% (n = 72), and 

permanent in 0.14% (n = 32). For all neurologic complications, 

including death related to neurologic events, independently 

correlating factors at multivariate analysis included 

patient age (p = .0008), history of frequent TIAs (p < 

.0001), and amount of contrast material used (p = .0205). 

The preprocedure imaging indication of cerebrovascular disease 

was more likely to correlate with neurologic complication 

of angiography.

CONCLUSION 

This study, comprising a prospective assessment of greater 

than 20,000 consecutive examinations, confirms the relative 

safety of diagnostic conventional cervicocerebral angiography. 

This true risk profile of cervicocerebral angiography 

should be weighed against its potential benefits in deciding 

between it and noninvasive imaging modalities. 

KEY WORDS: Angiography, complications, stroke 


MR Angiography of Radiation-Induced Extracranial 

Stenosis: Patterns and Correlation with Catheter 

Angiography 

Kane, A. G. 

Tripler Army Medical Center 

Honolulu, HI 

PURPOSE 

Radiation treatment of head and neck cancer can result in 

stenoses of the carotid and vertebral vessels. In order to identify 

such stenoses, noninvasive testing must distinguish radiation-induced 

stenosis from more frequent atherosclerotic 

stenoses in the carotid bifurcation. 


Seven high-grade (> 70%) stenosis in craniocervical vessels 

in patients who had undergone whole neck radiation therapy 

of head and neck cancer, and who subsequently were imaged 

with both MR imaging and catheter angiography, were 

reviewed. 

RESULTS 

All but one patient had smooth and long-segment common 

carotid stenoses (Figs. 1A, 1B arrows), and three of these 

had coincident proximal internal carotid bifurcation stenosis 

(two ulcerated) ipsilateral to the common carotid stenosis. 

Two patients had ipsilateral vertebral long-segment irregular 

stenoses. One patient treated with carotid stenting developed 

rapid restenosis within 4 months, underwent repeat angioplasty 

and is without restenosis 5 months later. 

27 

CONCLUSION 

Large vessel radiation injury commonly effects long segments 

of the common carotid artery, distinguishing this entity 

in location from more common diseases such as atherosclerotic 

stenosis, fibromuscular dysplasia, and dissection. 

Ulceration may be more common than in atherosclerotic 

lesions when radiation stenosis affects the carotid bifurcation. 

MR angiography, with its ability to rapidly image the 

entire craniocervical region, is ideally suited to screening 

these patients for potential intravascular therapy when radiation 

or encasement are suspected. 

REFERENCES 

1. Santoro A, Bristot R, Paolini S, Di Stefano D, Cantore G. 

Radiation injury involving the internal carotid artery. 

Report of two cases. J Neurosurg Sci 2000;44(3):159-164 

2. Gauvrit JY, Leclerc X, Gautier C, Pruvo JP. Techniques for 

evaluating the degree of carotid artery stenosis J 

Neuroradiol 2001;28(1):17-26 

3. Ohta H, Sakai N, Nagata I, Sakai H, Higashi T, Takahashi J. 

Bilateral carotid stenting for radiation-induced arterial 

stenosis. No Shinkei Geka 200129(6):559-563 

KEY WORDS: Radiation, vasculopathy, stenosis 


Endovascular Management of Extracranial Carotid and 

Vertebral Artery Dissections by Means of Stent- 

Supported Angioplasty 

Gomori, J. M. · Cohen, J. E. 

Hadassah Hebrew University Medical Center 

Jerusalem, ISRAEL 

PURPOSE 

Supraaortic dissection may lead to stenosis, occlusion, and 

pseudoaneurysms formation with subsequent embolic 

infarcts despite anticoagulation. Our purpose is to determine 

the therapeutic value of stent-supported endovascular reconstruction 

in a subgroup of patients who are poor candidates 

for medical therapy. 


The patient population was composed of 12 patients; 8 men 

and 2 women, with ages ranging from 17-64 years (mean 51 

years). Nine dissections were located on the ICA and 3 on 

the VA. Patients were considered for endovascular stenting 

Monday

Monday 

when symptomatic despite anticoagulation, contraindication 

for anticoagulation, clinical instability, or angiographic evidence 

of hemodynamic insufficiency. Before the procedure 

cranial CT/MR imaging and DSA of supraaortic vessels and 

cerebral parenchymography were performed. 

RESULTS 

The treatment significantly improved dissection-related 

stenosis from mean 75% stenoses to 8%. Two complete 

occlusions were recanalized. Multiple overlapping stents 

were used in 9 of the cases. No clinical complications were 

assessed in a mean follow-up period of 4 months (range 2 

months - 2 years). Postoperative morbidity: pneumonia, 1 

case. Postoperative mortality: 0%. 

CONCLUSION 

In selected cases of carotid and vertebral artery dissection, 

endovascular reconstruction by means of stents appears to be 

a safe and effective method of restoring arterial lumen with 

good clinical outcome. 

KEY WORDS: Dissection, stent, angioplasty 


High-Resolution Dynamic Time-Resolved MR 

Angiography: Initial Experience in the Neurovascular 

Patient 

Hopkins, J. K. · Cashen, T. · Walker, M. · Carr, J. · Futterer, 

S. · Parkinson, R. · Carroll, T. J. 

Northwestern University 

Chicago, IL 

PURPOSE 

Time-resolved contrast-enhanced MR angiography (CE 

MRA) is a promising technique, providing dynamic temporal 

information without significant degradation of spatial 

resolution. Rapid arteriovenous transit time and requirements 

for high spatial resolution make time-resolved imaging 

of neurovascular abnormalities particularly challenging. 

Additionally, spatial resolution is sacrificed incrementally in 

order to achieve an adequate frame rate. We have developed 

a CE MRA pulse sequence for high-resolution dynamic 

time-resolved MRA of the neurovascular system, combining: 

echo-sharing (TREAT), ultrashort TR and parallel imaging 

(GRAPPA). The purpose of this study is to evaluate this 

novel pulse sequence in a series of patients with a range of 

neurovascular abnormalities. 


Patients with suspected/known vascular disease were 

imaged using a 1.5 T whole body MR scanner (Avanto, 

Sonata, Siemens Medical Solutions, Erlangen, Germany) 

with a 2-channel neck coil and/or a 4-channel head coil. A 

3D multiphase TREAT (time-resolved echo-shared angiographic 

technique) pulse sequence was combined with parallel 

imaging (GRAPPA). The 3D TREAT sequence is a segmented 

k-space acquisition using the TRICKS variable rate 

k-space acquisition. Typical acquisition parameters were: 

(FOV = 300 x 135 x 64 mm, image matrix = 320 x 144 x 64, 

spatial resolution = 1.0 x 1.0 x 1.0 mm, temporal resolution 

= 6.0 s/frame, TR/TE = 2.44/0.91 ms, flip = 25, bandwith = 

980 Hz/pixel, R-L/A-P partial Fourier factors = 0.75/0.75). 

Gadolinium-based contrast material was administered as a 

28 

single dose in an antecubital vein at an injection rate of 4.0 

ml/s. 

RESULTS 

We have successfully imaged AVMs, STAMCA bypass, an 

intratumoral shunt, and stenosis of the carotid bifurcation. 

We have found the 6-second frame rate used for our standard 

submillimeter high-resolution CE MRA insufficient for high 

flow lesions, such as fistulae. In these cases, frame rates of 2 

to 3 seconds were desired and achieved through reducing the 

number of partitions, while maintaining in-plane resolution. 

Fig. 1 shows a series of coronal MIP images from a patient 

with intratumoral AV shunting (arrow) at temporal resolution 

of 4.5 seconds. 

CONCLUSION 

We have successfully combined echo-sharing and parallel 

imaging for high-resolution, high-frame rate neurovascular 

MRA. With continued improvements in temporal and spatial 

resolution, this technique could approach the gold standard 

examination, DSA. The potential to diagnose flow-related 

pathologies and assess or monitor treatment effects is exciting 

and could result in improvement in the morbidity and 

mortality associated with the typical work up of these types 

of lesions. 

KEY WORDS: Time resolved, contrast-enhanced MR angiography, 

carotid


Relationship between Carotid Artery Bifurcation 

Calcification and White Matter Changes 

Fanning, N. F. 1 · Walters, T. D. 2 · Symons, S. 1 · Fox, A. 1 

1Sunnybrook and Women’s College Health Sciences Center, 

Toronto, ON, CANADA, 2Hospital for Sick Children, 


PURPOSE 

Carotid bulb scores for calcified plaque have been proposed 

for future stroke risk, similar to acceptance of coronary 

artery calcification scores for future heart disease risk (1). 

Relative future stroke risk already has been correlated with 

white matter change severity. We sought to link associations 

between carotid calcification and white matter severity on 

CT to predict relative risk for future stroke, based on calcification 

grade. 


We retrospectively reviewed neuroradiologic findings in 221 

patients with unenhanced neck and brain CT for carotid bulb 

calcification and white matter changes. Degree of carotid 

calcification was scored by the Agatston method (2), calculating 

the calcific plaque load as the sum of all single calcific 

plaques in the carotid bulb, with each plaque calculated as 

a function of its Hounsfield density and area. White matter 

change severity was graded on CT according to the 

European Task Force for Age-Related White Matter Change 

scale (3). Each variable was measured in a blinded fashion. 

Demographic details, including age and gender, were recorded. 

Univariate and multivariate analyses were used to examine 

for evidence of association. 

RESULTS 

Both carotid calcification and white matter scores were 

strongly, and independently, associated with increasing age 

(r = 0.65, p < 0.001 and r = 0.67, p < 0.001 respectively). 

Despite apparent association between carotid calcification 

and white matter scores on univariate analysis, there was no 

independent effect evident after adjusting for age as a covariant 

(r = 0.006, p = 0.93). Gender had no independent effect 

on white matter scores; however, males had a marginally 

higher mean calcified carotid plaque load than females after 

controlling for age (p = 0.025). 

CONCLUSION 

Carotid bulb CT calcification scores do not independently 

predict severity of white matter changes. Future stroke risk, 

assessed by white matter severity scores, cannot be predicted 

from carotid calcific score. It is premature to suggest that 

screening CT for carotid calcification has potential benefit. 

REFERENCES 

1. Pletcher MJ, Tice JA, Pignone M, Browner WS. Using the 

coronary artery calcium score to predict coronary heart disease 

events: a systematic review and meta-analysis. Arch 

Intern Med 2004;164:1285-1292 

2. Agatston AS, Janowitz WR, Hildner FJ, Zusmer NR, Viamonte 

M, Detrano R. Quantification of coronary artery calcium 

using ultrafast computed tomography. J Am Coll Cardiol 

1990;15:827-832 

29 

3. Wahlund LO, Barkhof F, Fazekas F et al. A new rating scale 

for age-related white matter changes applicable to MRI and 

CT. Stroke 2001;32:1318-1322 

KEY WORDS: Carotid bulb calcification, white matter 

changes, stroke risk 


Validity of Millimeter Carotid Stenosis CT Angiography 

Measurements for Endarterectomy 

Bartlett, E. S. 1 · Walters, T. D. 2 · Symons, S. P. 1 · Fox, A. J. 1 

1Sunnybrook and Women’s College Health Sciences Center, 



PURPOSE 

Millimeter carotid stenosis measurements from high speed CT 

angiography (CTA) are direct, accurate, and have a linear relationship 

with NASCET-style ratios. Millimeter measurements 

can be used as guidelines for management of carotid stenosis 

in lieu of ratio calculations that NASCET and other trials have 

defined previously to support surgical versus medical management 

of carotid artery disease. The guidelines from these 

prior trials are based upon ratio measurements, since direct 

measurement of stenosis is not possible with standard catheter 

digital subtraction angiography (DSA). 


Two neuroradiologists separately reviewed 267 carotids 

imaged on CTA blinded to other information. The narrowest 

portion of carotid stenosis was measured in millimeters. 

Distal internal carotid (ICA) was measured well beyond the 

carotid bulb. NASCET-style ratio was calculated for each 

ICA, excluding suspected near-occlusion cases. Correlation 

coefficients were calculated to determine the degree of interobserver 

variation in mm measurements. ROC curves were 

utilized to determine the millimeter stenosis value equivalent 

to the 70% severe stenosis NASCET guideline with the 

greatest sensitivity, specificity, and predictive value. 

RESULTS 

Excellent interobserver correlation (0.78 to 0.89, 2-tail significance 

= 0.01) permitted averaging of millimeter stenosis 

and distal ICA measurements from the CTA luminogram for 

each carotid. These values were used to calculate mean percent 

stenosis. Regression analysis of mean percent stenosis 

as a function of mean millimeter stenosis proves a linear 

relationship between these values (Pearson’s correlation of - 

0.95, N = 136). ROC curve analysis demonstrates a direct 

millimeter measurement of 1.3 mm identifies severe carotid 

stenosis (70% as in NASCET) with a sensitivity of 88.2%, 

specificity 92.4%, and negative predicted value 98.2%. 

CONCLUSION 

Direct measurement of carotid stenosis by CTA can be utilized to 

determine the appropriateness of endarterectomy for severe 

symptomatic stenosis with relatively high sensitivity and specificity, 

based upon previously published percent stenosis guidelines 

from NASCET. With a negative predicted value of 98.2%, 

clinicians can be assured that patients with carotid stenosis measurements 

greater than 1.3 mm are without severe carotid disease. 

KEY WORDS: CT angiography, carotid stenosis, endarterectomy 

Monday

Monday 


Direct CT Angiographic Measurements of Vessels 

Alleviates Cumbersome Estimates for Ratio Calculations 


1Sunnybrook and Women’s College Health Sciences Centre, 



PURPOSE 

CT angiography of the carotid arteries allows for the direct 

measurement of diseased and normal vessels, alleviating the 

need for cumbersome and novel ratio calculations. This is 

especially true when regarding the ambiguities of some ratio 

calculations from studies such as the European Carotid 

Surgery Trial (ECST) and the Carotid Stenosis Index (CSI) 

method (1). The ECST method calculates a ratio based upon 

an estimate of an unseen carotid bulb. The CSI estimates the 

size of the carotid bulb by an equation: 1.2 x CCA diameter, 

based upon previously published anatomical relationships 

(2). 


Two neuroradiologists reviewed 261 carotid CTAs with 

carotid disease. Millimeter measurements were obtained of 

the residual stenotic lumen in the narrowest diameter, the 

actual carotid bulb diameter at the level of the greatest stenosis, 

and at the common carotid artery, measured between 3- 

5 centimeters prior to the carotid bifurcation. Interobserver 

variation correlation was calculated for all measurements. 

Pearson correlation coefficients compared the CSI estimate 

of the carotid bulb to the actual carotid bulb measurement 

from CTA. Ratio calculations of the stenosis were performed 

using the CSI carotid bulb estimate as denominator data and 

then repeated using the actual carotid bulb measurement as 

denominator data [(1-stenosis/carotid bulb)*100]. A pairedsample 

Wilcoxon Signed Rank test was performed to compare 

the results of these two ratio measurements per carotid. 

RESULTS 

Interobserver variability was good, ranging from 0.63 to 

0.81. The CSI estimate of the carotid bulb size routinely 

overestimated the actual measured carotid bulb value by an 

average of 1.5 mm. The overestimation was randomly distributed, 

as evidenced by the poor correlation between the 

CSI estimate and the actual bulb measurements (Pearson’s 

correlation coefficient = 0.4, N = 151). Paired-sample 

Wilcoxon Signed Rank test (2-tailed, equal variance) 

demonstrated a significant difference between the two sets of 

ratios with a Z-value of -9.51 (p < 0.001). 

CONCLUSION 

With the high-resolution anatomical data present in highspeed 

CTA, direct measurement of vessel diameter is possible. 

Cumbersome ratio calculations using visual or mathematical 

estimates of vessel diameter are no longer necessary. 

REFERENCES 

1. Bladin CF, Alexandrov AV, Murphy J, Maggisano R, Norris JW. 

Carotid stenosis index: A new method of measuring internal 

carotid artery stenosis. Stroke 1995;26:230-234 

30 

2. Williams MA, Nicolaides AN. Predicting the normal dimensions 

of the internal and external carotid arteries from the 

diameter of the common carotid. Eur J Vasc Surg 1986;1:91- 

96 

KEY WORDS: CT angiography, carotid disease 


Cross-Sectional Mm 2 Area of Carotid Stenosis CT 

Angiography 

Bartlett, E. S. · Symons, S. P. · Fox, A. J. 

Sunnybrook and Women’s Health Sciences Center 


PURPOSE 

New tools integrated in the current versions of PACS workstations 

allow for a more precise calculation of the carotid 

geometry. Carotid stenosis quantification has relied traditionally 

upon measurements from catheter angiography of 

the carotid artery. The validity of such measurements can be 

questioned since a stenotic carotid lumen often is shaped 

irregularly. Trials regarding this question have resulted in 

recommending digital subtraction angiography (DSA) imaging 

in at least 2 different planes and the provision of qualifying 

statements such as “narrowest diameter.” 


During an 8-month review of consecutive carotid CT 

angiography (CTA) in a single institution, two neuroradiologists 

evaluated 178 stenosed carotids in a blinded protocol. 

Carotid artery bulb stenosis was identified on axial CTA. 

AGFA Impax 4.5 volume tool (VT) utilizing Houndsfield 

units was used to estimate the cross-sectional area of the 

contrast luminogram. The narrowest diameter of a carotid 

stenosis was measured also by manually placing measurement 

calipers. Correlation coefficients were calculated to 

determine the degree of interobserver variation. Pearson correlation 

coefficients were calculated between the millimeter 

stenosis and the VT area, as well as between the VT area and 

the calculated area (based upon the narrowest diameter and 

assuming a circular residual lumen using a formula of: area 

= Pi x radius 2 ). 

RESULTS 

Excellent interobserver correlation (0.71 to 0.81, 2-tail significance 

= 0.01) permitted averaging of narrowest diameter 

stenosis measurement and the VT area values per carotid. 

There is excellent correlation between the VT area of the 

often irregular carotid stenosis and the narrowest diameter 

mm measurement (Pearson’s correlation of 0.77, N = 178). 

The VT area was generally greater than the calculated area 

by an average of 2.03 mm 2 . This difference is due to the often 

irregular shape of the residual lumen. Nonetheless, there is 

excellent correlation between the VT area and the calculated 

area (Pearson’s correlation of 0.76, N = 178).

Carotid measurement and example of volumetric tool: A: 

Measurement in centimeters of a right ICA with proximal 

stenosis. B: AGFA Impax 4.5 volume tool at the same level 

as in figure A, with perimeter and area in centimeters and 

squared centimeters, respectively. 

CONCLUSION 

Measurement of the narrowest portion of a carotid stenosis 

remains an acceptable method to characterize carotid disease. 

There is excellent correlation to more precise area calculations 

of an often irregular residual lumen. 

KEY WORDS: CT angiography, carotid stenosis, volume tool 


Identification of Carotid Near Occlusion by CT 

Angiography 


1 Sunnybrook and Women’s College Health Sciences Center, 

Toronto, ON, CANADA, 2 Hospital for Sick Children, 


PURPOSE 

CT angiography is a valuable tool for evaluation of carotid 

stenosis. However, near occlusion with secondary distal 

carotid artery reduction remains a confounder for calculation 

of NASCET-style percent stenosis. Using criteria to identify 

subtle findings of near occlusion (1), a recent review of 

NASCET and ECST angiograms diagnosed 262 near occlusions, 

only 16 having significant luminal collapse. 

Identification of near occlusion is essential prior to attempted 

percent stenosis ratio calculation, since these calculations 

yield fallacious numbers. The implications relate to proper 

stenosis quantification and treatment, since the risk of ipsilateral 

stroke in near occlusion is less than in severe stenosis (1). 


Two hundred and forty carotid artery CTAs for known or 

suspected carotid disease were evaluated independently by 2 

neuroradiologists with a blinded protocol. Besides identifying 

arteries with criteria for near occlusion, millimeter measurements 

were obtained at narrowest diameter of the stenotic 

bulb, distal ICA well beyond the tapering carotid bulb, and 

distal ECA prior to its terminal branching. Forty-three 

carotid arteries were labeled near occlusion. All interpretative 

disagreement cases were reviewed at consensus meeting. 

Interobserver variance was calculated for all measurements. 

Threshold values best predicting near occlusion were 

devised with ROC curve analysis for: 1) ICA stenosis, 2) distal 

ICA, 3) distal ICA:distal ICA, and 4) distal ICA:ECA. 

Combinations of variables also were evaluated. 

31 

RESULTS 

Based upon threshold values for the independently measured 

variables, the sensitivity range = 88.4-100; specificity = 83.9 

-92.9; PPV = 57.6-73.1 and NPV = 96.6-100. The range for 

paired permutations is: sensitivity = 76.3-84.2; specificity = 

93.8-97.1; PPV = 72.5-89.5 and NPV = 93.6-96.1. 

Fig. 1. The partially collapsed ICA called near occlusion on 

right is about the diameter of ECA, far narrower than usual 

relationship as seen on the left (arrow = distal ICA). 

CONCLUSION 

Threshold values are most helpful in providing guidelines to 

the CTA interpreter when assessing carotid artery disease 

and for the presence of near occlusion. Ultimate identification 

of near occlusion requires the judgment of the interpreter, 

with attention to the following criteria: 1) notable 

stenosis of the ICA bulb, and 2) distal ICA caliber reduction 

compared to: a) expected size, b) contralateral ICA, and c) 

ipsilateral ECA. 

REFERENCES 

1. Fox AJ, Eliasziw M, Rothwell PM, Schmidt MH, Warlow CP, 

Barnett HJM. Identification, prognosis, and management of 

patients with carotid artery near occlusion. AJNR Am J 

Neuroradiol in press 

KEY WORDS: Near occlusion, CT angiography, carotid disease 

Discussion 


10:15 AM - 11:45 AM 

Room 103 

(8) ELC Workshop B: Advanced 

PowerPoint 


— H. Christian Davidson, MD 

Monday

Monday 

Monday Afternoon 

1:00 PM - 1:30 PM 

Room 205 

(9) ELC Lecture B: High Speed 

Connectivity Update 


1:00 PM - 2:30 PM 

Theatre 

(10) Vascular Malformations of the 

Head and Neck (ASITN) 

(63) Imaging of Head and Neck Vascular 

Malformations 

— Anton N. Hasso, MD, FACR 

(64) Treatment of Hi-Flow Lesions 

— Wayne F. Yakes, MD 

(65) Treatment of Low-Flow Lesions 

— Patricia E. Burrows, MD 

Discussion 

Moderator: Mary E. Jensen, MD 

— Gerard J. Muro, MD 

Imaging of Head and Neck Vascular Malformations 

Anton N. Hasso, MD, FACR 



able to: 

1) Discuss the classification system of vasoformative anomalies. 

2) Review MR imaging findings in hemangiomas and vascular 

malformations. 

3) Construct a table listing the optimal treatment of vasoformative 

anomalies. 

32 


Vascular anomalies can be divided into hemangiomas and 

vascular malformations based on histologic features, natural 

history, and biologic behavior. Hemangiomas are benign 

endothelial tumors that appear during infancy, undergo a 

period of rapid proliferation, and eventually regress. 

Accordingly, hemangiomas can be subclassified as either 

proliferating or involuting. Vascular malformations, on the 

other hand, are developmental defects of the vasculature that 

are present at birth, grow in parallel with the individual, and 

never involute. These lesions are subdivided based on the 

predominant vascular channel that is affected, namely arterial, 

venous, capillary, lymphatic, or combined. MR imaging 

is critical in the characterization of vascular anomalies as it 

can determine hemodynamic flow properties, delineate 

extent of disease, and guide treatment. High-resolution, 3dimensional 

capabilities, and noninvasiveness make MR 

imaging the gold standard in the treatment planning of craniofacial 

vascular lesions. The hallmark imaging features of 

hemangiomas and vascular malformations are coupled 

closely to clinical decision-making strategies. Current therapeutic 

options include steroids, laser photocoagulation, percutaneous 

sclerotherapy, endovascular embolization, and 

surgical resection. The impetus for treatment of head and 

neck lesions is to improve cosmetic appearance, eliminate 

functional impairment, and preclude life-threatening complications. 

A number of effective treatments currently are 

employed in the management of vascular anomalies, including 

pharmacotherapy, laser photocoagulation, sclerotherapy, 

cryosurgery, embolization, and surgical resection. In this 

presentation, we will review the salient MR findings for various 

vascular anomalies and discuss strategies for selecting 

optimal treatments for lesions of the head and neck. 

REFERENCES 

1. Mulliken JB, Glowacki J. Hemangiomas and vascular malformations 

in infants and children: a classification based on 

endothelial characteristics. Plast Reconstr Surg 1982;69:412- 

422 

2. Meyer JS, Hoffer FA, Barnes PD, Mulliken JB. Biological classification 

of soft-tissue vascular anomalies: MR correlation. 

AJR Am J Roentgenol 1991;157:559-564 

3. Jackson IT, Carreno R, Potparic Z, Hussain K. Hemangiomas, 

vascular malformations, and lymphovenous malformations: 

classification and methods of treatment. Plast Reconstr Surg 

1993;91:1216-1230 

4. Van Aalst JA, Bhuller A, Sadove AM. Pediatric vascular 

lesions. J Craniofac Surg 2003;14:566-583 

5. Enjolras O, Mulliken JB. The current management of vascular 

birthmarks. Pediatr Dermatol 1993;10:311-313 

Treatment of Hi-Flow Lesions 

Wayne F. Yakes, MD 



able to: 

1) Be facile in the understanding of the various types of congenital 

types of vascular malformations. 

2) Be familiar with the rationale for endovascular treatment 

with the philosophy of permanent treatment. 

3) Be aware of various approaches for managing head and

neck vascular malformations with ethanol embolization. 

4) Understand the minimally invasive endovascular 

approaches to manage these complex and vexing lesions. 


Interventional Radiology and Interventional Neuroradiology 

have pioneered a minimally invasive therapeutic specialty to 

treat a wide variety of vascular and nonvascular lesions of the 

body, the brain, spine, spinal cord and head and neck areas. 

Interventional procedures routinely use minimally invasive, 

direct puncture, and transcatheter techniques to treat various 

conditions. In the head/neck and paraspinal area the vast 

majority of entities that are treated are largely of vascular origin. 

The extensive array of catheters, guidewires, embolic 

agents, digital imaging systems and the pharmaceutical products 

that commonly are used are a tribute to the hard work, 

insight, and imagination of the many dedicated investigators 

in this area. Because of significant laboratory research, clinical 

research, and extensive clinical experience, the judicious 

use of endovascular therapy is now commonplace in modern 

clinical practice. Now that it is firmly established as an essential 

therapeutic tool, its role will only continue to grow. In the 

head and neck and paraspinal region, embolotherapy procedures 

have become an essential therapeutic modality that has 

assisted our colleagues in otolaryngology, plastic and reconstructive 

surgery, neurosurgery, ophthalmology, vascular surgery 

and various pediatric surgical specialties. After the diagnosis 

has been established, the next major decision is to determine 

whether therapy is warranted. The interventional radiologist/interventional 

neuroradiologist should plan and direct 

the patient’s care with surgical specialists who are familiar 

with the management of hemangioma and vascular malformations 

and the problems they present. It is extremely important 

that appropriate surgical, medical, pediatric, and anesthesiology 

specialists be involved for optimal patient care. They 

would function much like a tumor board team in the management 

of cancer. I classify vascular malformations as either 

high-flow or low-flow lesions. The high-flow lesions are 

shunting malformations such as arteriovenous malformations 

(AVMs) and congenital or acquired arteriovenous fistulae 

(AVF). The low-flow lesions include venous malformations, 

lymphatic malformations, and mixed lesions. The arteriography 

of these lesions demonstrates normal arteries and normal 

capillary beds. The malformation is truly a postcapillary 

lesion. The terms “cavernous hemangioma," “vertebral body 

hemangioma," and “hepatic hemangioma” should be replaced 

with the term “venous malformation." If they were truly 

hemangiomas, they should not be present in the adult population. 

Vascular malformations are treated best where these 

patients are seen on a regular basis. The interventional radiologist 

or interventional neuroradiolgist who occasionally evaluates 

these patients will never gain enough experience or 

knowledge to manage these challenging lesions effectively. 

All too frequently the patient ultimately pays for the interventionalist’s 

initial enthusiasm, inexperience, folly, and lack of 

necessary clinician backup. To optimally manage these 

patients, a dedicated team should be in place. 

Interventionalists, combined with the various surgical and 

medical subspecialties, function together much like the tumor 

board team of specialists. When patients are seen and treated 

regularly, then experience can be gained, rational decisions 

made, complications appropriately managed, and patient care 

then is optimized. It cannot be emphasized enough that, as a 

group, vascular malformations pose one of the most difficult 

challenges in the practice of medicine. A cavalier approach to 

33 

their management always will lead to significant complications 

and dismal outcomes for the patients. 

REFERENCES 

1. Mourao GS, Hodes JE, Gobin YP, Casasco A, Aymard A, 

Merland JJ. Curative treatment of scalp AV fistulas by direct 

puncture and embolization with absolute alcohol. J 

Neurosurg 1991;75:634-637 

2. Yakes WF. Interventional neuroradiologic procedures in the 

head and neck: ENT perspective. In: English GM, (ed). 

Otolaryngology. New York: Lippincott-Raven publishers, 

1996;1-26 

3. Yakes WF. Diagnosis and management of AVMs. In: Haskal 

ZJ, Kerlan RK, Trerotola SO (eds). SCVIR Syllabus: Thoracic 

and Visceral Vascular Interventions. Fairfax: Society of 

Cardiovascular and Interventional Radiology Publishers, 

1996;314-322 

4. Yakes WF, Rossi P, Odink H. How I do it: arteriovenous malformation 

management. Cardiovasc Intervent Radiol 

1996;19:65-71 

5. Yakes WF. Endovascular management of high-flow AVMs. 

Semin Intervent Radiol 2004;21:49:49-58 

Treatment of Low-Flow Lesions 

Patricia E. Burrows, MD 

Dr. Burrows obtained her MD in Winnipeg, Manitoba, 

Canada and did fellowships in Pediatric Radiology, 

Pediatric Special Procedures, and Interventional 

Neuroradiology. She has been involved in imaging and treatment 

of vascular anomalies since 1985 and currently is the 

Co-Director of the Vascular Anomalies Program at 

Children's Hospital, Boston, where she has worked for the 

past 12 years. Her clinical and research activities are almost 

entirely related to endovascular treatment of vascular malformations. 

She has published over 140 original papers and 

50 reviews on the subject of pediatric vascular disease. 



able to: 

1) Diagnose various forms of low-flow vascular malformations 

by imaging and clinical characteristics. 

2) Determine appropriateness for treatment of patients with 

low-flow vascular malformations. 

3) Be familiar with the drugs and techniques that are useful 

in treating vascular malformations. 

4) Recognize complications of treatment of low-flow vascular 



Low-flow vascular malformations include venous malformations 

(VM), lymphatic malformations (LM) and capillary malformations 

(CM) as well as combined forms. Sclerotherapy or 

injection of sclerosing drugs into the vascular spaces is the current 

standard method of treatment. Commonly used sclerosants 

for venous malformations include ethanol, sodium tetradecyl, 

and other detergent type sclerosants. For lymphatic malformations, 

injection of ethanol, OK432, and doxycycline have been 

found to be effective. Venous malformations can be diffuse, 

focal, or multifocal. The soft tissues of the head and neck can 

be involved. The most extensive lesions involve the cerebral 

and cranial veins as well. Focal and multifocal lesions respond 

Monday

Monday 

best to sclerotherapy. Intraorbital venous malformations are 

almost impossible to treat safely by injection techniques 

because of the risk of orbital compartment syndrome and loss 

of vision. Venous malformations of the oral cavity and more 

distal airway require prolonged airway protection after sclerotherapy. 

Direct injection of supraglottic VM can be accomplished 

through a rigid laryngoscope. Varicocoele expansion of 

superficial neck veins can be treated but require more complicated 

techniques. Lymphatic malformations (LM) are classified 

as macrocystic or microcystic. Most cystic forms can be 

injected with sclerosant, using ultrasound guidance. Orbital 

LM with a visual impairment due to repeated bleeds and infections 

can be sclerosed cautiously, although there is a risk of 

orbital compartment syndrome. Complications include damage 

to skin and nerves, infection, orbital compartment syndrome, 

and airway obstruction. Patient follow-up surveys indicate 

good to excellent results in at least 75% of cases. 

Discussion 


1:00 PM - 2:30 PM 

Room 105/106 

(11) Neuropathies of the Head and 

Neck (ASNHR) 

(66) Upper Cranial Nerves (III-VI) 

— William P. Dillon, MD 

(67) Lower Cranial Nerves (IX-XII) 

— Wendy R.K. Smoker, MD, FACR 

(68) Brachial Plexopathy 

— Brian C. Bowen, PhD, MD 

Moderators: Wendy R.K. Smoker, MD, FACR 


Upper Cranial Nerves (III-VI) 



Upon completion of this presentation, participants will be able 

to: 

1) Recognize the anatomy and function of cranial nerves 3-6 

on MR imaging. 

2) Be familiar with the MR imaging protocol for and appearance 

of cranial nerve pathology. 

34 


Cranial nerve dysfunction is a disturbing problem that in most 

cases quickly brings the patient to medical attention. 

Symptoms are usually a manifestation of the particular nerve 

that is involved. The new onset of cranial nerve palsy may be 

the first sign of malignancy or other serious medical or vascular 

disease. For instance, acute third nerve palsy may be the 

harbinger of an expansile cerebral aneurysm; benign and 

malignant neoplasms, infection, noninfectious inflammatory 

abnormalities, radiation damage, and vascular infarctions may 

result in various cranial neuropathies. In addition, a number of 

idiopathic disorders may result in either single or multiple cranial 

neuropathies. Imaging of the cranial nerves has taken on 

more importance with the use of contrast-enhanced fat-suppressed 

MR techniques, which is the study of choice. Using 

MR imaging, the nerve itself is visualized throughout its intra 

and extracranial course. While a complete evaluation of cranial 

nerve dysfunction is beyond this short communication, 

the anatomical and pathologic features of dysfunction of cranial 

nerves 3-6 will be covered. There are several basic concepts 

that are useful to remember when designing a protocol 

for the patient with cranial neuropathy. Imaging should be performed 

in at least two planes encompassing the nucleus, the 

nerve, and its distal muscle of innervation. Imaging with both 

T1- and T2-weighted sequences with fat saturation is necessary. 

Contrast-enhanced, fat-suppressed T1-weighted imaging 

is essential for diagnosis of many disorders. Secondary signs 

of damage to cranial nerves, such as atrophy of denervated 

muscles, are often helpful signs. 

SUGGESTED READINGS 

1. Fischbein NJ, Dillon WP, Barkovich AJ, eds. A Teaching 

Atlas: Brain. Thieme;2000 

2. Wilson-Pauwels L, Akesson EJ, Stewart PA. Cranial Nerves: 

Anatomy and Clinical Comments. Toronto: B.C. Decker 

Inc.;1988 

3. Waxman SG, de Groot J. Correlative Neuroanatomy, 22nd 

edition. Connecticut:Appleton and Lange;1995 

4. Kelly WM, guest editor. Cranial Neuropathy. Neuroimag 

Clin North Am 1993;3 

5. Harnsberger HR. Handbook of Head and Neck Imaging. St. 

Louis:Mosby;1995 

Lower Cranial Nerves (IX-XII) 

Wendy R.K. Smoker, MD, FACR 



able to: 

1) Define the course (intracranial and extracranial) of the 

lower four cranial nerves. 

2) Name the muscles innervated by each of the lower four 

cranial nerves and identify them on cross-sectional images. 

3) Learn the imaging appearance of different stages of denervation 

atrophy. 

4) Name common pathologies that affect the lower four cranial 

nerves in their cisternal, skull base, high carotid space, 

and distal segments. 


Radiologic evaluation of the patient with deficits of the lower 

four motor cranial nerves requires knowledge of the course of

each nerve as well as muscles innervated by these nerves. 

Cranial nerves IX, X, and XI are complex nerves having both 

motor and sensory/parasympathetic fibers. However, symptoms 

of motor denervation are those that are typically appreciated 

and cause patients to present for radiologic evaluation. 

The hypoglossal nerve has complete motor function. 

Glossopharyngeal Nerve: Nucleus (ambiguus) in rostral 

medulla; exits brainstem between olives and inferior cerebellar 

peduncle; exits skull through pars nervosa of jugular 

foramen to enter high carotid space. Innervates stylopharyngeus 

muscle and contributes to pharyngeal plexus (along 

with the vagus nerve), providing innervation to levator veli 

palatini, superior and middle pharyngeal constrictor, 

palatopharyngeus, and palatoglossus muscles. 

Vagus Nerve: Nucleus (ambiguus) in rostral medulla; exits 

brainstem between olive and inferior cerebellar peduncle; 

exits skull base through pars vascularis of jugular foramen 

and courses within carotid sheath. Innervates inferior pharyngeal 

constrictor and cricothyroid muscles via the superior 

laryngeal nerve, the endolaryngeal muscles via the recurrent 

laryngeal nerve, and contributes to pharyngeal plexus. 

Spinal Accessory Nerve: Nucleus in anterior lateral gray of 

upper 5 or 6 cord segments; rootlets ascend through foramen 

magnum and exit skull base through pars vascularis of jugular 

foramen; Innervates trapezius and sternocleidomastoid 

muscles. 

Hypoglossal Nerve: Nucleus in rostral medulla, anteromedial 

to dorsal motor nucleus of X; rootlets exit brainstem in 

preolivary sulcus; coalesce to exit skull base through 

hypoglossal canal and course within high carotid space 

before coursing to terminate in sublingual space. Innervates 

intrinsic and extrinsic tongue muscles. The lower four cranial 

nerves may be affected by supranuclear, nuclear (brainstem), 

or infranuclear pathology. (This discussion does not 

address supranuclear pathology). Common nuclear pathology 

includes multiple sclerosis, infarction, neoplasms (primary 

and metastatic), syringobulbia, etc. Infranuclear 

pathology is subdivided into cisternal, skull base, carotid 

space, and distal segments. Nuclear, cisternal, skull base, and 

carotid space pathology often produces deficits of all four 

lower cranial nerves due to their close proximity in these 

locations. More distal pathology, however, often produces 

isolated cranial deficits. Various patterns of denervation atrophy 

often help distinguish proximal from distal pathology. 

Common cisternal pathology: meningitis, vertebrobasilar 

dolichoectasia, neoplasms (especially neurogenic tumors, 

and meningiomas). 

Common skull base pathology: trauma, neoplasms (especially 

metastases, paragangliomas, neurogenic tumors, and 

meningiomas), and infections. 

Common carotid space pathology: primary and nodal SCCa, 

lymphoma, neoplasms (especially paragangliomas and neurogenic 

tumors), infection, and vascular pathology (carotid 

dissection/aneurysm, jugular thrombophlebitis). 

Brachial Plexopathy 




able to: 

1) Identify plexus intrinsic and extrinsic anatomy, as dis- 

35 

played on MR images. 

2) Recognize the MR appearance of motor denervation. 

3) Identify the MR imaging features that are used to differentiate 

between normal and abnormal peripheral nerves. 

4) Recognize, and potentially avoid, pitfalls and artifacts in 

the acquisition and interpretation of plexus MR images. 


The clinical evaluation of plexopathy and peripheral neuropathy 

increasingly involves the use of high-resolution MR 

imaging to facilitate diagnosis and management. MR studies 

(MR neurography) of the brachial plexus are especially challenging 

because of the oblique course, complex intrinsic 

anatomy of the plexus components (roots, trunks, divisions, 

cords) and their extrinsic anatomical relationships (interscalene 

triangle, subclavian/axillary arteries/veins, and bony 

landmarks), as well as regional susceptibility effects and 

physiologic (vascular, respiratory) motion artifacts that can 

degrade image quality. Current imaging techniques emphasize 

the use of dedicated phase array coils, high spatial resolution 

acquisitions in the coronal (plexus in-plane) and sagittal 

or oblique-sagittal imaging planes (plexus in cross-section) 

of the involved plexus, predominantly spin-echo (T1weighted 

+ fat sat, T2-weighted + fat sat) and STIR 

sequences, and the use of a gadolinium contrast agent, in 

order to obtain diagnostic images of the component nerves 

themselves (approximately 2 - 6 mm diameter). In the 

patient with plexopathy, the MR study usually is requested to 

evaluate a clinically known or suspected (1) neoplastic or 

inflammatory mass (intrinsic versus extrinsic to plexus components), 

(2) traumatic injury (at sites ranging from neural 

foramina to cords and their branches), (3) entrapment or 

nontraumatic compressive lesion, or to provide evidence 

favoring (4) intrinsic structural abnormalities (hereditary 

hypertrophic neuropathy) or idiopathic conditions (brachial 

plexitis). MR imaging in the (5) posttreatment (surgery, 

radiotherapy) patient with a history of cancer usually is 

requested to help distinguish between radiation injury to the 

plexus and recurrent tumor. Examples of these clinical “indications” 

for brachial plexus imaging will be presented and 

discussed. The discussion also will include technical and 

anatomical considerations, such as magic angle effects in 

peripheral nerve imaging. 

REFERENCES 

1. Bowen BC, Pattany PM, Saraf-Lavi E, Maravilla KR. The 

brachial plexus: normal anatomy, pathology, and MR imaging. 

In: Bowen BC, Maravilla KR, Naidich T (eds). Imaging of 

the peripheral nervous system. Philadelphia, Elsevier Science 

(USA), Neuroimag Clin N Am 2004; pp59-85 

2. Chappell KE, Robson MD, Stonebridge-Foster A, Glover A, 

Allsop JM, Williams AD, Herlihy AH, et al. Magic angle 

effects in MR neurography. AJNR Am J Neuroradiol 

2004;25:431-440 

3. Bowen BC. Peripheral nerve imaging and the magic angle. 

Editorial. AJNR Am J Neuroradiol 2004;25:352-354 

Monday

Monday 


1:00 PM - 2:30 PM 

Room 107 

(12) NeuroNews Scientific Paper 

Session 

Moderators: TBD 


1:00 PM - 2:30 PM 

Room 201B 

(13) ASNR Business Center - Part I 

(69) Welcome and Introduction to Business Issues 

in Radiology 

— Gregory L. Katzman, MD 

(70) Net Present Value and Finance for Practicing 

Radiologists 

— Jonathan Breslau, MD 

(71) Managerial Accounting Applications in 

Radiology 

— Frank Lexa, VII, MD, MBA 

Moderators: Gregory L. Katzman, MD 


1:00 PM - 2:30 PM 

Room 103 

(14) ELC Workshop C: Website 

Creation for the Novice 


36 


1:30 PM - 2:00 PM 

Room 205 

(15) ELC Lecture C: Capturing and 

Using Image Files 


3:00 PM - 4:00 PM 

Room 103 

(16) National Library of Medicine 

(NLM) Workshop: PubMed ® /Medline 

Plus: Advanced Tips and Tricks 


3:00 PM - 4:30 PM 

Room 105/106 

— Barton F. Branstetter, IV, MD 

— Linda Milgrom 

(17a) HEAD AND NECK: Neck and 

Extracranial Carotid 



(17b) INTERVENTIONAL: Arteriovenous 

Malformations/Fistulae 

(17c) HEAD AND NECK: Neck, New Techniques, 

and Excerptas 

(17d) ADULT BRAIN: Infarction and Vasospasm 

Moderators: Anton N. Hasso, MD, FACR 

Wendy R.K. Smoker, MD, FACR

Paper 75 Starting at 3:00 PM, Ending at 3:08 PM 

Role of Diffusion-Weighted MR Imaging in Pediatric 

Head and Neck Masses 

Abdel Razek, A. A. · Hafez, M. · Bilal, M. · Monier, S. · 

Elmogy, S. · Rudwan, W. 

Mansoura University Faculty of Medicine 

Mansoura, EGYPT 

PURPOSE 

To determine the role of diffusion-weighted MR imaging in 

pediatric head and neck masses. 


This study included 37 patients (24 boys and 13 girls aged 3 

months-15 years: mean 6 years) with head and neck mass. 

Routine MR imaging and diffusion-weighted MR imaging 

were done on a 1.5 T MR unit (Symphony, Siemens). 

Diffusion MR imaging was done using a single-shot echoplanar 

imaging (EPI) with a diffusion-weighted factor, factor 

b of 0.500 and 1000 sec/mm2. The ADC map was reconstructed 

with calculation of apparent diffusion coefficient 

(ADC) values. Image analysis was performed qualitatively 

and quantitatively. The signal intensity was assessed visually 

on ADC maps and ADC value was measured in the suspected 

lesion. The final diagnosis proved by pathology, MR 

appearance, and clinical examination. 

RESULTS 

Adequate ADC maps were obtained in 34 patients. This 

study included malignant tumor (n = 13), benign lesions (n = 

21). Restricted diffusion (high SI at B = 1000 images and 

low SI at ADC map) was seen in 10 lesions. Free diffusion 

(low SI at B = 1000 images and high SI at ADC map) was 

seen in 16 lesions. Mixed pattern of diffusion (areas of low 

and high SI at B = 1000 images and ADC map) were seen in 

8 lesions. The mean ADC value of the malignant tumor (1.07 

± 0.17 X 10-3 mm2/sec) was smaller than that of benign 

lesions (1.99 ± 0.17 X 10-3 mm2/sec). There was significant 

difference in ADC values of benign and malignant pediatric 

head and neck tumors (p < 0.021). 

CONCLUSION 

Diffusion-weighted MR imaging is a new promising imaging 

modality for differentiating malignant pediatric head and 

neck tumors from benign lesions. It can be added to routine 

MR imaging of pediatric head and neck masses. 

KEY WORDS: Pediatric, diffusion 


Recurrent Head and Neck Neoplasms: Quantitative 

Assessment by Means of Dynamic MR Imaging with 

Analysis of Perfusion 

Maroldi, R. · Farina, D. · Piazzalunga, B. 

University of Brescia 

Brescia, ITALY 

PURPOSE 

To assess the role of dynamic MR sequence in the detection 

of recurrent head and neck neoplasms. 

37 


Forty-one patients previously treated for a head and neck 

tumor (28/41 squamous cell carcinoma, 10/41 adenoid cystic 

carcinoma, 1/41 adenocarcinoma, 1/41 olfactory neuroblastoma, 

1/41 rhabdomyosarcoma; 13/41 nasopharynx, 

12/41 paranasal sinuses, 10/41 oral cavity, 6/41 oropharynx,) 

had follow-up MR studies. Patients had been treated by surgery 

alone (3/41), RT (13/41). or both (25/41). In all patients 

a dynamic MR sequence (GE T1, TR813, matrix 2562; 

repeated acquisition of a single slice for 60 sec) was added 

to the standardized protocol. On images acquired, regions of 

interest (ROI) were placed on ICA, on the area of suspected 

relapse, and on normal mucosa and muscle. Slope of 

enhancement curves as well as average of maximum signal 

intensities (AMS) within ROI were statistically analysed (ttest). 

MR findings were confirmed by histology or subsequent 

follow-up studies. 

RESULTS 

Slope of enhancement curves of positive lesions (1.67 ± 0.6) 

was significantly steeper than in negative lesions (0.76 ± 0.6) 

(p = 0.00009). This finding was confirmed also when the 

steepest slope tracts of the curves were compared (p = 

0.00001). Significant difference was observed between AMS 

of recurrent (90 ± 18) and negative patients (66 ± 18) (p = 

0.0008). ROC curves analysis shows that a cut-off value of 

selective slope > = 1.4 enables to detect all recurrences with 

a 100% negative predictive value and 85.2% positive predictive 

value. 

CONCLUSION 

This dynamic sequence provides additional data on the 

enhancement pattern that are useful to discriminate recurrences 

from nonneoplastic tissues. 

KEY WORDS: Recurrence, head and neck neoplasm, MR 

imaging 


Reliable Fat Signal Suppression in Head and Neck 

Imaging Using a Three-Point Dixon Technique 

Barger, A. · DeLone, D. · Bernstein, M. · Welker, K. 

Mayo Clinic 

Rochester, MN 

PURPOSE 

MR imaging of the extracranial head and neck is made difficult 

by multiple factors, including unreliable fat suppression. 

The effectiveness of frequency-selective RF excitation or 

saturation pulses is dependent on main magnetic field (B0) 

homogeneity. Tissue-induced magnetic field inhomogeneity 

in head and neck imaging is substantial and is very difficult 

to remove completely by shimming, thus limiting the reliability 

of frequency selective fat saturation. Dixon techniques 

use multiple acquisitions with differing phase information to 

generate fat and water images (1). The original Dixon 

method is also unreliable in the presence of B0 inhomogeneity. 

However, modified Dixon techniques can account for 

effects of inhomogeneity-induced off-resonance (2, 3). In 

this work we investigated the feasibility of using a modified, 

three-point Dixon technique to obtain reliable fat-water separation 

in the soft-tissue neck. 

Monday

Monday 


Imaging was performed using a 1.5 T Signa scanner with 

EXCITE 11.0 software (GE Medical Systems, Milwaukee, 

WI). Coils used included five- and eight-channel neuro-vascular 

phased-array coils (Medrad, Indianola, PA, and MRI 

Devices, Gainesville, FL, respectively). A three-point Dixon 

protocol was performed on three volunteers and one patient. 

FSE T1-weighted scans were performed using TR 600, 

TEeff 11.5, 256 x 192 matrix, ETL 4. FSE T2-weighted 

scans were performed using TR 2553, TEeff 97.3, 256 x 192 

matrix, ETL 11. Scan time is quite manageable (generally 

several minutes per pack), and was proportional to the phase 

encoding matrix size divided by the ETL, the number of 

passes, the number of Dixon “points” (i.e., 3), and the repetition 

time. GE Dixon reconstruction software was used for 

postprocessing the acquired data. For volunteer studies, the 

Dixon FSE T1 was done without gadolinium; in the single 

patient study, the Dixon FSE T1 was done after intravenous 

administration of gadolinium. 

RESULTS 

Resultant separate fat, water, and combined images from the 

method were reviewed. In all subjects the water images 

obtained using three-point Dixon technique demonstrated 

effective, uniform fat suppression. In two of the cases the 

Dixon method caused some erroneous, nonanatomical mapping 

of fat signal, which was irrelevant to clinical interpretation 

in the one clinical case. 

CONCLUSION 

Three-point Dixon imaging promises to be an effective, 

robust, and practical method of fat suppression in the 

extracranial head and neck. The method does require three 

acquisitions for each phase encode, which imposes limitations 

of coverage and resolution to maintain a reasonable 

scan time. However, the signal-to-noise ratio is also 

improved and is roughly equivalent to a 2.7 NEX acquisition 

(2). In this preliminary investigation some erroneous mapping 

of fat signal was observed. Although this appears nonlimiting, 

a larger series of patients is needed. 

REFERENCES 

1. Dixon WT. Simple proton spectroscopic imaging. Radiology 

1984;153:189-194 

2. Glover GH, Schneider E. Three-point Dixon technique for 

true water/fat decomposition with B0 inhomogeneity correction. 

Magn Reson Med 1991;18(2):371-383 

3. Reeder SB, Wen Z, Yu H, Pineda AR, Gold GE, Markl M, Pelc 

NJ. Multicoil Dixon chemical species separation with an 

iterative least-squares estimation method. Magn Reson Med 

2004;1(1):35-45 

KEY WORDS: Dixon, fat suppression, head and neck 

38 


Radiologic Features of Solitary Fibrous Tumors of the 

Head and Neck 

Stambuk, H. E. · Ganly, I. · Carlson, D. · Ghossein, R. · Shah, 

J. P. · Patel, S. G. 


New York, NY 

PURPOSE 

To illustrate the salient radiologic features in the rare, generally 

benign, solitary fibrous tumors of the head and neck. 


Seven patients with a histopathologic diagnosis of solitary 

fibrous tumor (SFT) treated between the years 1990-2004 

were identified from institutional databases. Retrospective 

review of available CT (5) and MR (5) scans was performed 

by a CAQ certified neuroradiologist. Imaging findings were 

correlated with clinical data and pathologic features with the 

aim of identifying radiographic traits that would suggest the 

diagnosis of SFT. 

RESULTS 

Solitary fibrous tumors occurred more commonly in women 

than men (4:3). Age at presentation ranged from 46 to 78 

years (median 57 years). Symptoms at presentation varied 

according to the anatomical site of the primary tumor: palpable 

mass (3), nasal obstruction (2), arm pain (1), and ptosis 

(1). Most tumors presented as a slow-growing, painless 

mass with a history ranging in duration from 2 months to 3 

years. Anatomical sites of tumor included the sinonasal cavity 

(3), oral cavity (2), infratemporal fossa (1), and 

paraspinal soft tissues of the cervical spine (1). On imaging, 

all lesions were well defined including the two malignant 

lesions that had extended intracranially. Tumors ranged in 

size from 1.5 x 1.4 cm to 6.3 x 3.2 cm. All tumors were 

densely enhancing on imaging with tumor generally isodense 

to muscle on noncontrast CT and isointense to muscle 

on precontrast T1-weighted images. Most tumors were low 

to intermediate signal on T2-weighted images; however two 

had high-signal intensity. All lesions that were adjacent to 

bone caused expansile regressive remodeling. In addition, 

both patients with malignant SFT had areas of bone destruction 

one of which also had sclerosis of the fovea ethmoidalis, 

crista galli, and several ethmoid septa. All patients were 

managed by surgical resection. Five of the seven lesions 

were benign and two were malignant on histopathologic 

examination. The margins of surgical resection were positive 

in 3 patients (2 malignant and 1 benign SFT). None of the 

tumors recurred and all patients were alive with follow-up 

periods ranging from 2 to 76 months. 

CONCLUSION 

Solitary fibrous tumors of the head and neck region are rare 

and are most commonly benign. Although imaging findings 

are not diagnostic, the presence of an intensely enhancing, 

well defined mass with regressive remodeling of bone can be 

suggestive of SFT. Bone destruction is suggestive of malignant 

SFT, but even malignant lesions tend to remain well 

defined in appearance.

KEY WORDS: Mesothelioma/*diagnosis/pathology/surgery, 

paranasal sinus neoplasms/*diagnosis/pathology/surgery, 

neoplasm invasiveness 


Pathologic Changes of the Lateral Pterygoid Muscle in 

Patients with Temporomandibular Joint Disk 

Derangement: Objective Measures at MR Imaging 

Finden, S. G. · Enochs, W. · Rao, V. 

Thomas Jefferson University 

Philadelphia, PA 

PURPOSE 

The lateral pterygoid muscle, specifically the superior head, 

has been implicated in anterior dislocation of the disk in 

internal derangement of the temporomandibular joint (TMJ). 

In patients with occlusal disorders, the superior head of the 

lateral pterygoid muscle (SHLPM), which inserts on the disk 

and anterior capsule of the TMJ, has been postulated to 

undergo spasm and contraction, which results in forward 

traction and anterior displacement of the disk (1). In longstanding 

TMJ dysfunction, the SHLPM is expected to show 

subsequent edema and/or atrophy (2). The goal of the present 

study is to examine the corresponding changes of the 

SHLPM on MR imaging using objective measures of signal 

intensity and morphology. 


Patients with abnormalities involving the TMJ were identified 

through a retrospective review of MR imaging reports at 

our institution from January 1997 through August 2003. 

Relative signal intensities were measured for representative 

regions of interest (ROI) within the superior and inferior 

heads of the LPM bilaterally on T2-weighted, sagittal 

oblique images. In addition, relative thickness between the 

two muscle heads was measured as a morphologic parameter. 

Statistically significant differences in relative signal 

intensity and muscle thickness between the superior and 

inferior heads of the LPM then were identified and correlated 

with the presence or absence of anterior displacement, 

with or without reduction. 

RESULTS 

In patients with anterior disk displacement without reduction, 

17 of 19 joints (89.5%) demonstrated a significant 

increase in relative T2 signal intensity within the SHLPM 

compared to the inferior head. By contrast, no patients with 

anatomically normal disks demonstrated a statistically significant 

difference in signal between the superior and inferior 

heads of the LPMs. Variation in relative thickness 

between normal and abnormal muscles overlapped to an 

extent that precluded its use as a distinguishing parameter. 

CONCLUSION 

Pathologic alterations in the superior head of the lateral 

pterygoid muscle can be identified using ROI values as an 

objective measurement of relative T2 signal intensity. 

Comparing these values between the superior and inferior 

heads demonstrated a strong correlation (89.5%) between 

increased T2 signal intensity and pathologic alteration of the 

condylar head/disk relationship. This increased signal may 

reflect increased fluid signal related to muscle edema and/or 

fatty change secondary to atrophy. By contrast, relative 

39 

thickness between the two muscle heads as a morphologic 

measure is not useful in identifying this pathology. 

REFERENCES 

1. Wongwatana S, Kronman JH, Clark RE, Kabani S, Mehta N. 

Anatomic basis for disk displacement in temporomandibular 

joint (TMJ) dysfunction. Am J Orthod Dentofacial Orthop 

1994;105(3):257-264 

2. Yang X, Pernu H, Pyhtinen J, Tiilikainen PA, Oikarinen KS, 

Raustia AM. MR abnormalities of the lateral pterygoid muscle 

in patients with nonreducing disk displacement of the 

TMJ. Cranio 2002;20(3):209-221 

KEY WORDS: Temporomandibular joint, lateral pterygoid 


Extracranial Extramedullary Hematopoiesis in Two 

Teenagers with Sickle-Cell Disease 

Coopersmith, H. · Bello, J. 

Montefiore Medical Center 

New York, NY 

PURPOSE 

The purpose of this abstract is to describe intracranial 

epidural and extracranial soft tissue manifestations of 

extramedullary hematopoiesis (EMH) in two adolescent 

males homozygous for sickle-cell disease. 


Preceding vaso-occlusive crises in both patients suggest a 

direct causal relationship between the crisis, and the development 

of EMH. D.L. presented to the hospital with a 

bifrontal headache, and anterior chest pain. He was afebrile, 

and his lungs were clear to auscultation. D.L. was maintained 

on his routine daily doses of hydroxyurea and folic 

acid. Morphine, zithromax, prednisone, and cefzil were 

added to his treatment regimen. On hospital day (HD) #6, 

D.L.’s admission hemoglobin of 8.8 g/ml reached a low of 

6.8 g/ml. Four days later, he was discharged from the hospital, 

only to return the next day complaining of a severe retroorbital 

and bifrontal headache, and sternal chest pain. Three 

days later, D.L. developed bilateral parieto-occipital tender, 

fluctuant masses measuring 4 cm by 4 cm in size, as well as 

bilateral fronto-parietal swelling. On HD #5, MR imaging 

was performed. D.V. presented to the hospital with a productive 

cough, severe chest pain, and bilateral upper leg 

pain. After 24 hours, he developed a fever. A chest X-ray 

demonstrated a left lower lobe infiltrate. D.V. was started on 

intravenous cefuroxime, azithromax, and albuterol nebulizers 

in addition to the hydroxurea, folic acid, morphine, 

toradol, oxygen, and intravenous fluids he had been receiving 

already. Over his first three hospital days, D.V.’s hemoglobin 

dropped from 9.4 g/ml to 5.1 g/ml. He received 2 

units of packed red blood cells on HD #4, and was discharged 

from the hospital the following day. D.V. returned 1 

day later with swelling over the left side of his scalp, a history 

of fever to 101 degrees, and worsening bilateral leg 

pain. A CT scan of the brain was performed that day, as well 

as an MR image 5 days later. D.L.’s MR image showed diffuse 

expansion of the calvarium with multiple areas of calvarial 

high signal in the biparietal and bifrontal regions on 

T2, Flair, and T1-weighted sequences. In addition, there was 

an extracranial lobular soft tissue mass adjacent to the left 

Monday

Monday 

parietal bone, measuring 4.2 by 1.7 cm in size. The mass 

demonstrated high signal on both T1- and T2-weighted 

sequences. Small additional areas of high signal were present 

in the left parietal and right frontal epidural regions 

intracranially. D.V.’s CT scan demonstrated high attenuation 

regions in the left epidural space and left subgaleal space 

near the convexity measuring 3.1 cm and 5.9 cm in size, 

respectively. There was no evidence of bone erosion. The 

MR image showed high signal on T1- and T2-weighted 

sequences in the left frontal epidural space and the left subgaleal 

region, with minimal postcontrast enhancement. 

CONCLUSION 

Our case series is significant for several reasons. First, these 

may represent the first reported cases of EMH in the 

extracranial soft tissues. Second, there are few reported cases 

of EMH in patients homozygous for sickle-cell disease. 

Third, the time-course of EMH shadowed the time-course of 

the vaso-occlusive crises, suggesting a direct causal relationship. 

KEY WORDS: Extramedullary hematopoiesis, extracranial 

mass, sickle-cell disease 


Subarachnoid Perfluorocarbon Droplets after Attempted 

Retinal Detachment Treatment 

Malin, D. R. · Aulino, J. M. · Recchia, F. M. 

Vanderbilt University Medical Center 

Nashville, TN 

PURPOSE 

We present a case of a gunshot wound resulting in severe 

orbital trauma treated with vitreoretinal surgery and perfluorocarbon 

(PFC) liquid injection. Postoperative CT demonstrated 

migration of PFC into the subarachnoid space. 


A 22-year-old man suffered a gunshot wound to the face 

which produced multiple facial fractures and orbital trauma 

including retinal detachment. Surgical repair and exploration 

for a left open-globe injury was performed including vitreoretinal 

surgery with PFC liquid injection to treat retinal 

detachment. Optic nerve avulsion was discovered which rendered 

the treatment unsuccessful. Postoperative CT imaging 

demonstrated subarachnoid hyperdense PFC droplets. 

RESULTS 

Initial trauma head CT demonstrated extensive facial fractures 

and orbital trauma. After vitreoretinal surgery with 

PFC liquid injection to treat retinal detachment, unusual 

hyperdense small droplets were seen within the subarachnoid 

space and within the left intraconal orbit. 

CONCLUSION 

Perfluorocarbon liquids and silicone oil have been used as an 

aid in vitreoretinal surgery for the treatment of complex retinal 

detachments. As a highly dense substance, PFC may provide 

significant tamponade to displace subretinal fluid and 

blood and to stabilize and flatten the retina. Known complications 

include subretinal retention of PFC and migration of 

silicone oil along the optic nerve into the subarachnoid 

space. We do not believe that subarachnoid migration of PFC 

40 

has been reported previously. Our case illustrates orbital 

trauma with attempted retinal detachment repair resulting in 

subarachnoid migration of PFC droplets. 

KEY WORDS: Perfluorocarbon, subarachnoid, orbit trauma 


Multislice Carotid Wall Imaging at 1.5 T and 3.0 T 

Koktzoglou, I. 1 · Carroll, T. J. 1 · Walker, M. T. 1 · Morasch, M. 

D. 1 · Mani, V. 2 · Mizsei, G. 2 · Fayad, Z. A. 2 · Simonetti, O. P. 3 

· Li, D. 1 

1 Northwestern University, Chicago, IL, 2 Mount Sinai School 

of Medicine, New York, NY, 3 Siemens Medical Solutions, 

Chicago, IL 

PURPOSE 

Noninvasive MR imaging of the carotid artery wall has been 

shown to be capable of detecting and characterizing carotid 

atherosclerotic plaque at 1.5 T. However, achievable image 

signal-to-noise ratio (SNR) is limited by spatial resolution. 

One way to improve SNR is to image at higher magnetic 

field strengths. The objective of this work was to compare 

the SNR efficiency of carotid artery wall imaging at 1.5 T 

and 3.0 T using multislice turbo spin-echo (TSE) imaging. 


Six healthy volunteers (4 males, 2 females, mean age = 37 

years) were imaged on 1.5 T and 3.0 T whole-body scanners 

(Magnetom Sonata and Trio, Siemens Medical Solutions, 

Erlangen, Germany). Four channel phased-array coils consisting 

of two left and two right channels (1.5 T: Machnet 

BV, The Netherlands; 3.0 T: in-house coil) were used for signal 

reception. In each volunteer, 12 cross-sectional slices 

through the carotid bifurcation were imaged using fat-saturated 

proton density-weighted, T2-weighted, and T1-weighted 

regional saturation (RSAT) TSE. Imaging time (~4 min) 

and imaging resolution (0.47 x 0.47 x 3 mm 3 ) was matched 

on both 1.5 T and 3.0 T scanners. Carotid wall SNR and was 

measured at both 1.5 T and 3.0 T. 

RESULTS 

MR imaging was completed successfully in all volunteers. 

Mean carotid artery wall SNR at 1.5 T was 17 ± 4 (proton 

density-weighted), 11 ± 2 (T2-weighted), and 12 ± 3 (T1weighted). 

At 3.0 T mean carotid artery wall SNR was 36 ± 

18 (proton density-weighted), 23 ± 9 (T2-weighted), and 30 

± 13 (T1-weighted). Siganl-to-noise ratio efficiency at 3.0 T 

was roughly 2.2 fold higher than that of 1.5 T. Based on 

these results, three patient studies were performed at 3.0 T 

using RSAT TSE. Carotid plaques were clearly delineated in 

all patients. A T2-weighted image acquired from a patient at 

3.0 T is shown. The atherosclerotic plaque is well delineated, 

and exhibits what appears to be a fibrous cap (gray 

arrow) and a lipid core (white arrow).

CONCLUSION 

Carotid wall imaging benefits from higher SNR at 3.0 T. In 

this study the SNR gain at 3.0 T over 1.5 T was roughly 2.2 

fold. The higher SNR afforded by 3.0 T may allow for even 

higher resolution imaging of the carotid artery wall. Further 

studies with pathologic correlation are needed to validate the 

findings in the MR images. 

KEY WORDS: MR imaging, carotid artery, vessel wall imaging 


Evaluation of Intracranial and Cervical Arteries with 3D 

CT Angiography Using a 16-Detector Multidetector Row 

CT: Continuous Scanning of the Head and Neck Using a 

Single Injection of Contrast Medium 

Matsumoto, M. · Endo, Y. · Kodama, N. · Sakuma, J. · 

Suzuki, K. · Sasaki, T. 

Fukushima Medical University 

Fukushima, JAPAN 

PURPOSE 

Extracranial carotid atherosclerosis is well established as an 

important risk factor for the development of thromboembolic 

cerebrovascular disease. To detect cervical vascular lesions in 

50 patients with known or suspected intracranial cerebrovascular 

disease, we performed 3D CTA of the head and neck 

using a multidetector CT (MD CT) with 16 detectors. 


The head and neck were scanned continuously with a 20 s 

scan delay after a bolus injection of contrast medium 

(3ml/sec, total 100 ml). The data then were transferred to a 

workstation (ZIO M900; Amin Corp., Tokyo, Japan). A volume-rendering 

method was used for the extraction of vessels 

at a 140-200 HU threshold. We also created multiplanner 

reconstruction (MPR) images and maximum intensity projection 

(MIP) images of the cervical arteries and evaluated 

the visualization of the cervical arteries. 

41 

RESULTS 

We obtained excellent 3D CTA from the origin of the common 

carotid arteries to the superior sagittal sinus in all 

patients. A small aneurysm (1.2 mm in a diameter) was 

detected in this method. The quality of the 3D images of this 

method was equal or superior to that of conventional 3D 

CTA which target on head only. The CT numbers of the cervical 

carotid bifurcation, vertebral artery, and jugular veins 

at C4 level were 360.1 HU (n = 100), 322.7 HU (n = 100) 

and 189.8 HU (n = 100), respectively. Because of the high 

CT number of the carotid arteries, it was possible to extract 

arteries upon setting a threshold CT number. Using 3D CTA, 

whole course of the vertebral arteries were well demonstrated, 

especially at the cranio-vertebral junction. Since the CT 

number of the VA was 40HU lower than that of the carotid 

artery at the same level, we should set a scan delay late to 

evaluate the vascular lesions in the VA. To visualize the cervical 

segment of the VA within the transverse foramens of 

C6 to C2, we deleted vertebral arches at an imaginary line 

that linked transverse foramens of the cervical vertebrae. 

CONCLUSION 

The head and neck 3D CTA using the MD CT with 16-detectors 

provides valuable diagnostic information regarding vascular 

lesions of the intracranial and cervical arteries. This 

method facilitates the screening for vascular lesions in the 

cervical arteries as well as intracranial arteries in a single 

procedure. 

KEY WORDS: Multidetector CT, 3D CT angiography, multiplanar 

reconstruction 


Measuring Carotid Arterial Distensibility: Comparison 

of Measurements of Common Carotid Cross-Sectional 

Area on Catheter and CT Angiography 

Sherman, P. M. · Takhtani, D. 

The Johns Hopkins Medical Institutions 

Baltimore, MD 

PURPOSE 

We hypothesized that catheter angiography causes arterial 

distensibility due to direct injection whereas CT angiography 

does not. To measure the degree of distensibility, comparison 

was made between common carotid arterial area on both 

catheter and CT angiography. 


We retrospectively reviewed 25 common carotid arteries in 

15 patients in which both catheter angiography and neck CT 

angiography were performed during the past 2 years. 

Catheter angiograms were performed using hand injections. 

We included patients in whom both studies were performed 

within a 3-month period. There were 7 males and 8 females, 

ranging in age from 9 to 84 years. Common carotid diameter 

was measured on the lateral catheter angiogram image, 1 

centimeter below the carotid bifurcation. An intraluminal 

surface area was measured at the same location on the CTA 

using commercially available software after correcting for 

obliquity. The third or fourth cervical vertebral body was 

measured in the antero-posterior dimension at the midbody 

on the lateral carotid angiogram image and on the 3 D MIP 

lateral CT image to calculate the correction factor, using the 

Monday

Monday 

CT measurement as the reference standard. After applying 

appropriate magnification factor, the angiographic cross-sectional 

area was calculated using the formula π r2. 

RESULTS 

The arterial distensibility, defined as percent increase in area 

on catheter angiography compared to CTA, ranged from 12 

% to 58.4% in 23 arteries. In one common carotid artery in 

an 9-year-old girl who had complete internal carotid occlusion, 

the increase was 116%, and in another common carotid 

artery in a 52-year-old man with 90% internal carotid stenosis, 

the increase was 63.4%. 

CONCLUSION 

We demonstrated carotid distensibility in all vessels. This 

may be one of the factors in discordance between CTA and 

catheter angiography in the assessment of stenosis. This new 

way of measuring arterial distensibility also may find new 

applications in assessing arterial stiffness which has been 

found to predispose to atherosclerosis. Usually arterial distensibility 

is measured by doppler ultrasound. The degree of 

variation in distensibility in different subjects could be 

attributed to distal stenosis, differences in hand injection 

technique, inherent stiffness of varying degrees in the arteries, 

and intraobserver measurement differences. 

KEY WORDS: Carotid artery; distensibility, carotid angiography, 

neck CT angiography 


Zero Filling Interpolation Processing Technique Can 

Efficiently Suppress Segmental Stenosis Artifact on 

Small Artery of Anatomical Phantoms 

Lin, R. 1 · Wu, R. H. 1 · Xiao, Z. W. 1 · Liu, G. R. 1 · Kong, K. 

M. 1 · Lang, Z. J. 2 

1 Shantou University Medical College, Shantou, CHINA, 

2 Dalian Medical University, Dalian, CHINA 

PURPOSE 

The purpose of this study was to evaluate the zero-filling 

interpolation processing (ZIP) technique for contrastenhanced 

MR angiography (CE MRA). 


Phantoms of arteries were made with different lumen diameter. 

Gadolinium-enhanced three-dimensional MR angiography 

was performed on a GE 1.5 T scanner. The parameters 

of FSPGR pulse sequence were: flip angle 30°, TR 6 ms, TE 

1.4 ms, bandwidth 31.25 kHz, slice thickness 1.2 mm, matrix 

256 × 256. The sequence parameters were kept constant for 

the studies, whereas four selections were chosen: (1) with 

ZIP1024 and ZIP×4 techniques; (2) only with ZIP1024 technique; 

(3) only with ZIP×4 technique; (4) without ZIP technique. 

For image quality evaluation, MR maximum intensity 

projection (MIP) images were created. Signal-to-noise 

ratio (SNR) was measured on MIP images. Vessel edge was 

determined using full width at half maximum (FWHM) for 

lumen diameter calculation and calculated results were compared 

with the actual lumen diameter. The distinctness of the 

vessel edge and the artifacts on the phantoms were compared 

for all sequences. 

42 

RESULTS 

Three experienced radiologists made consensus evaluation. 

The FWHM results of lumen measurements for all the 

sequences with ZIP techniques were more accurate than that 

of the sequence without ZIP technique in all phantoms, no 

matter what was the size the artery. The vessel edge with 

ZIP1024 technique was more distinct. But the highest average 

SNR was obtained with the sequence without ZIP technique. 

The segmental stenosis artifacts on small artery of 

phantoms were only efficiently suppressed with ZIP×4 technique. 

CONCLUSION 

ZIP technique is excellent for CE MRA to obtain high quality 

MR angiography; it not only can improve the spatial resolution 

and the distinctness of the vessel edge on CE MRA, 

but also can efficiently suppress segmental stenosis artifact 

on small artery of phantoms. 

KEY WORDS: Zero 


Intraventricular Migration of Silicone Oil after 

Intravitreous Injection of Silicone 

Singla, R. · Magalhaes, A. C. A. · Kish, K. 

Wayne State University 

Detroit, MI 

PURPOSE 

Intravitreous silicone injection has become the standard for 

the treatment of retinal detachment. MR findings of intraocular 

silicone oil have been well described. However there are 

only few pathologic reports of the migration of silicone into 

the ventricles and to our best knowledge, no corresponding 

reports of CT imaging findings. This is an uncommon entity 

and it is important to recognize the radiologic findings of 

presence of intraventricular silicone and differentiate it from 

other intraventricular pathologies. 


A 43-year-old female with human immunodeficiency viral 

infection who had undergone intravitreous injection of silicone 

for retinal detachment as a complication of CMV retinitis 

presented with headaches. Noncontrast CT scan showed 

multiple small spherical areas in the nondependent portion of 

the lateral, third and fourth ventricles. Subsequently MR 

imaging was performed and a diagnosis of intraventricular 

silicone migration from the inraoccular injection of silicone 

oil was made based on the MR findings. 

RESULTS 

Noncontrast CT scan showed multiple spherical hyperdense 

lesions in the nondependent portion of the ventricles. 

Contrast-enhanced CT scans did not show any enhancement. 

Subsequent scanning in the prone position revealed migration 

of these lesions to the nondependent posterior portion. 

Nonenhanced MR imaging revealed hyperintense lesions in 

the nondependent parts of the ventricles on T1-weighted 

images and fluid-attenuated inversion recovery images. The 

lesion was hypointense when compared with vitreous in the 

globe and CSF in the ventricles on fat saturated T2-weighted 

images. A prominent chemical shift artifact was present 

along the frequency encoding direction that changed appro-

priately as the phase and frequency encoding directions were 

swapped. Retrospective consultation of the records of the 

patient revealed she had developed raised intraocular pressure 

after the silicone injection. 

CONCLUSION 

The migration of intravitreous silicone to the ventricles has 

been reported only recently and is not a well known entity. 

Most of the patients who had intraventricular silicone were 

reported to have developed raised intraocular pressure after 

the silicone injection. MR and CT findings of the intraventricular 

silicone are diagnostic. The mechanism of migration 

of intravitreous silicone into the ventricles is not well established; 

there are hypothesis to explain the migration but 

these need further evaluation to establish the facts. 

Consequences of migration of silicone into the ventricles are 

not known at this time; however, wide spread knowledge of 

radiologic findings and subsequent establishment of radiologic 

diagnosis can prevent further intervention and biopsy 

unless some adverse effects are noticed. 

KEY WORDS: Silicone oil, intraventricular, MR imaging 

43 


3:00 PM - 4:30 PM 

Theatre 

(17b) INTERVENTIONAL: 

Arteriovenous Malformations/Fistulae 



(17a) HEAD AND NECK: Neck and Extracranial 

Carotid 


and Excerptas 


Moderators: Philippe E. Gailloud, MD 



Ethanol Embolization of Tongue Vascular 

Malformations 

Yakes, W. F. 

Vascular Malformation Center 

Englewood, CO 

PURPOSE 

To determine the efficacy of ethanol embolization in management 

of tongue vascular malformations. 


Forty-seven patients (26 female, 21 male; mean age: 38 

years) presented to my service with tongue vascular malformations 

(8 AVMs; 39 venous-lymphatic). Forty-seven 

patients had undergone 61 failed previous procedures (embo, 

laser, surgery, steroid injection, alpha-interpheron, radiation). 

All patients had baseline arteriograms and MR images. 

All patients underwent transcath and/or direct puncture 

ethanol therapy. 

RESULTS 

Of 8 AVM patients, 5 cured and 3 treatment on-going (mean 

f/up: 40 months); of 39 patients with venous-lymphatic malformations, 

18 patients had dramatic reduction and 21 

patients’ therapy on-going with concurrent reductions (mean 

f/up: 60 months). One patient with AVM required additional 

surgery and 1 patient with mixed veno-lymphatic malformation 

required surgical debulking of excess tissues. Minor 

complications such as tongue blisters (9 instances) healed 

spontaneously; 3 tongue injuries healed spontaneously; 3 

infections responded to antibiotic treatment; 1 focal numbness 

resolved. 

Monday

Monday 

CONCLUSION 

Ethanol embolotherapy is a primary form of therapy to eradicate 

high-flow and low-flow vascular malformations of the 

tongue permanently at long-term follow-up. Rarely is concurrent 

surgery required. 

KEY WORDS: Vascular malformations, embolization, ethanol 


Unusual Imaging Findings following Ethanol Ablation of 

a Glomus Jugulare Tumor 

Wallace, M. A. · Rothfus, W. E. · Bartynski, W. S. 


Pittsburgh, PA 

Stroke is a known complication of alcohol ablation of 

meningohypophyseal arterial supply of indirect carotid-cavernous 

sinus fistulae and tumors from inadvertent distal ablation. 

However, there is no published data that describes the 

imaging characteristics of the ethanol during such complications. 

We describe a case of acute stroke following ethanol 

ablation of meningohypophyseal supply to a glomus jugulare 

tumor where the ethanol is visualized within the distal 

middle and anterior cerebral arteries on CT and MR imaging 

immediately following the procedure. The CT findings 

resemble intraarterial or subarachnoid air or fat and the MR 

findings are nonspecific. In vitro test images of dehydrated 

alcohol confirm the imaging characteristics found on the 

patient’s CT and MR imaging correlate to the characteristics 

of ethanol. Therefore, it should be recognized that intraarterial 

ethanol resembles an air or fat embolism on imaging following 

a complicated alcohol ablation procedure. 

KEY WORDS: Alcohol, stroke, ablation 

44 


Craniofacial Gunshot Wounds: A Review of 

Endovascular Management at a Level I Urban Trauma 

Center 

Gor, D. M. · Weigele, J. B. · Bronov, O. · Hurst, R. W. 

Hospital of University of Pennsylvania 


PURPOSE 

Craniofacial injuries account for 14% of all gunshot wounds. 

The relatively easy accessibility of high velocity militarytype 

weapons accounts for recent increases in numbers of 

craniofacial gunshot injuries seen at most major trauma centers. 

Advances in endovascular techniques currently allow 

many of these injuries to be treated acutely in a relatively 

noninvasive fashion with a positive inpact on management. 

The purpose of this review is to describe the angiographic 

features of craniofacial gunshot injuries seen in a Level I 

trauma center and to discuss their endovascular management. 


A retrospective review of all diagnostic carotid and vertebral 

angiograms from January 1997 to July 2004 was performed 

to identify gunshot injuries to the craniofacial region. 

Indications for diagnostic angiograms included active external 

or internal bleeding, expanding hematoma, or vascular 

injury detected by other imaging methods. Patients treated 

using endovascular techniques were identified. 

Epidemiologic data, zone of injury, vessel injured, type of 

injury, endovascular management employed, and immediate 

outcome were reviewed systematically. 

RESULTS 

Diagnostic carotid and vertebral angiograms were performed 

on 103 patients with craniofacial gunshot injuries. 

Endovascular intervention was performed on 27 patients. 

Ages ranged from 18 years to 62 years. Twenty-six patients 

were male with only one female. There were no zone I 

injuries, 7 zone II injuries, 15 zone III injuries, and 5 zone IV 

injuries. Injuries which were managed endovascularly, 

affected 5 lingual arteries, 6 facial arteries, 2 external carotid 

arteries, 6 internal maxillary arteries, 1 ascending pharyngeal 

artery, 1 ophthalmic artery, 1 middle cerebral artery, 5 vertebral 

arteries, and 4 internal carotid arteries. Types of injuries 

identified included lacerations of 16 arteries, complete transection 

of 6 arteries, dissection in 3 arteries, pseudoaneurysm 

in 7 arteries, vertebro-venous fistula in 3 patients, 

carotid-cavernous fistula in one patient, and complete occlusion 

of the vertebral artery in one patient. In 8 patients, PVA 

particles were used for embolization, nondetachable microcoils 

were utilized in 12 patients, and a combination of PVA 

and nondetachable microcoils were used in 4 patients. 

Guglielmi detachable coils were placed in 2 patients, and a 

neuroform stent was placed across an internal carotid artery 

dissection in one patient. The endovascular procedure was 

successful in acutely treating the injury in 25 of the 27 

patients. In 2 patients hemorrhage continued from ophthalmic 

artery collaterals after occlusion of the internal maxillary 

artery. Both patients had persistent hemorrhage into 

the nasopharynx.

CONCLUSION 

A relatively high percentage of craniofacial gunshot wounds 

evaluated angiographically in a Level I trauma center serving 

an inner city were identified as having potentially lifethreatening 

vascular injuries. Endovascular treatment was 

successful in acute management of the vast majority of significant 

vascular injury resulting from craniofacial region 

gunshot wounds. A variety of injuries ranging from vascular 

laceration, occlusion, pseudoaneurysms, and arteriovenous 

fistulae were treated definitively with good immediate outcome. 

In multisystem trauma, active hemorrhage from the 

craniofacial injury can be controlled quickly by endovascular 

methods when the patient needs to be stabilized to attend 

to other critical injuries. 

KEY WORDS: Vascular trauma, endovascular treatment 


History of the Diagnosis and Treatment of Spinal AVMs: 

From the First Spinal Myelogram in 1927 to 

Embolization in 2004 

Gupta, G. · Xavier, A. R. · Prestigiacomo, C. J. 

New Jersey Medical School/University of Medicine and 

Dentistry of New Jersey 

Newark, NJ 

The earliest descriptions of vascular anomalies in the 

preimaging era are case histories based on postmortem studies. 

Tracing the historical footprints of the various vascular 

malformations of the spine is complicated by a lack of consistent 

terminology among the early clinicians writing about 

this entity. We did an extensive literature search on the oldest 

available references and texts on the diagnosis and treatment 

of spinal AVMs dating back to late 1800s till today and 

how a new era was introduced after the invent of spinal 

angiography. 

1. Early Era of Imaging and Myelography: In 1927, Perthes 

reported the first spinal AVM preoperatively diagnosed by 

performing myelography. With this began a new era in the 

diagnosis and treatment of spinal vascular malformations. 

The volume of reported cases in the literature continued to 

grow. The lack of adequate imaging and similar presentations 

for some of these cases resulted in significant confusion 

regarding the etiology and subsequent classification of 

vascular spinal lesions. In 1941 Elsberg revisited the diagnosis 

and treatment of spinal vascular malformations in 

Surgical Diseases of the Spinal Cord, Membranes and Nerve 

Roots (Elsberg 1941). He further described experimental 

techniques such as lipiodoloscopy and Dr. Pool’s myeloscope 

as potential tools to help in the preoperative diagnosis 

of such lesions. Interestingly, though cerebral angiography 

was becoming routine, the concept of spinal angiography as 

a means of diagnosing spinal vascular pathology was not 

mentioned. 

2. Spinal Angiography and it was never the same…….. From 

a classification perspective, the classification of vascular 

malformations of the spine was predominantly a topographic, 

anatomical description as evidenced by Wyburn-Masons 

monograph of 1943. This “era” ended with the introduction 

of spinal angiography. Hensen and Croft were the first to 

publish a case of a spinal vascular malformation in 1956, 

when they diagnosed a cervical AVM on a vertebral injection. 

This was followed by several additional reports 

45 

describing spinal AVMs with aortography. However, not the 

first, Djindjian presented a large case series with excellent 

descriptions of his angiographic findings along with the clinical 

presentations, thus helping to further develop the pathophysiologic 

basis of these diseases. 

3. Selective Spinal Angiography was engineered by 

Djindjian. Di Chiro, Doppman, and others helped develop 

the precise preoperative delineation of the angioarchitecture 

of spinal cord vascular lesions. 

4. The Age of 3D Rotational Angiography, Gd-Enhanced 

MRA and Embolization: The treatment of spinal cord malformations 

is being expanded further with the use of interventional 

neuroradiology. With further improvements in 

spinal angiography, endovascular techniques and lately 

excellent embolization techniques under angiographic guidance, 

these lesions are embolized either as a primary treatment 

or as a complement to open microsurgical techniques. 

KEY WORDS: Arteriovenous malformations, history and heritage 


Hemorrhage Rate in Patients with Spetzler-Martin 

Grade IV and V Arteriovenous Malformations: Is 

Treatment Justified? 

Jayaraman, M. V. · Marcellus, M. L. · Do, H. M. · Steinberg, 

G. K. · Chang, S. D. · Marks, M. P. 

Stanford University 

Stanford, CA 

PURPOSE 

To evaluate the rate of hemorrhage in patients harboring 

Spetzler-Martin Grade IV and V intracranial arteriovenous 

malformations (AVMs). 


Medical records of consecutive patients presenting with 

Spetzler-Martin Grade IV and V AVMs were reviewed retrospectively 

for demographics, presenting symptom(s), number 

and time of hemorrhage events pretreatment, during 

treatment, and after treatment was deemed completed. 

RESULTS 

Fifty-one consecutive patients were identified, 30 male and 

21 female, harboring 44 Grade IV and 7 Grade V AVMs. 

Mean age was 28 years, range 3 to 77 years. Presenting 

symptoms were seizure in 19 (37.3%), hemorrhage in 14 

(27.5%), progressive neurologic deficit in 10 (19.6%), 

headache in 6 (11.8%) and incidental findings at imaging in 

2 patients (3.9%). The 51 patients accounted for a cumulative 

follow-up period of 450 patient years from time of initial 

diagnosis. Prior to the initiation of treatment, there were 

a total of 31 hemorrhages in 18 patients. Fourteen of these 

hemorrhages were the initial presenting symptom, and 17 

hemorrhages occurred in the interval between diagnosis and 

initiation of therapy, including 5 hemorrhages in 4 patients 

who did not present initially with hemorrhage. Average number 

of hemorrhages prior to treatment was 1.7 per patient 

(range 1-4). There were 11 hemorrhages after initiation of 

treatment, for a total of 42 hemorrhages. The hemorrhage 

rate for this population was 9.33% per year from time of 

diagnosis. 

Monday

Monday 

CONCLUSION 

In our analysis, the hemorrhage rate for Grade IV and V 

AVMs was 9.33% per year. This is higher than has been 

reported for AVMs overall, but similar to the bleed rate for 

basal ganglia and thalamic AVMs and supports the notion 

that treatment for Grade IV and V lesions is warranted. 

KEY WORDS: AVM, treatment 


Interobserver Variability in the Assessment of Brain 

Arteriovenous Malformation Angioarchitecture and 

Endovascular Treatment Results 

Iancu-Gontard, D. · Weill, A. · Guilbert, F. · Sillvagio, J. · 

Raymond, J. · Roy, D. 

Centre Hospitalier Universitaire Montréal, Notre Dame 

Montreal, PQ, CANADA 

PURPOSE 

Several angiographic features of brain arteriovenous malformations 

(BAVM) have been associated with an increase risk 

of hemorrhage. However interpretation of these features may 

not be consistent between observers. In preparation for an 

institutional review of 370 patients with BAVM, we conducted 

a study to determine the interobserver agreement in 

different angioarchitecture characteristics of the BAVM. 


Two experienced interventional neuroradiologists, independently 

reviewed pre and postendovascular treatment DSA 

from 50 consecutive patients. The angiograms were performed 

in a single institution and stored on an image archive 

system. Axial CT and/or MR images before and after treatment 

were included, to aid in localization and size measurements. 

Observers were not presented with precise definitions 

of the angioarchitectural features assessed. The following 

standard data were collected: Spetzler-Martin grade including 

eloquence of the adjacent brain, size and pattern of 

venous drainage; feeding arteries (one vs more arterial territories); 

aneurysm type if identified (circle of Willis, feeding 

artery, nidus); venous ectasia; high vs low flow malformation; 

nidus reduction after endovascular treatment (< 33%, 

33-66% and > 66%). Both observers were compared to each 

other for the interobserver agreement. To measure agreement 

kappa statistic (k) was used for nominal data and weighted k 

for ordinal data. Statistical Product for the Social Science 

(SPSS) was used for all analyses with the exception of confidence 

intervals for k and weighted k tests, which were calculated 

using Statistical Analysis Software (SAS). 

RESULTS 

Interobserver agreement was greater for characteristics such 

as Spetzler-Martin grade (weighted k = 0.7) including eloquence 

of the adjacent brain (k = 0.71), size (k = 0.75) or 

venous drainage pattern (k = 0.8), for the arterial feeders (k 

= 0.64), for aneurysms on circle of Willis (k = 0.76) and for 

the global reduction in nidus size (k = 0.74). Interobserver 

agreement was inferior for findings concerning hemodynamics 

(k = 0.3), presence of feeding artery aneurysms (k = 

0.19), intranidal aneurysms (k = 0.34) and venous ectasia (k 

= 0.5). 

46 

CONCLUSION 

Many long held angiographic characteristics of BAVMs are 

not reliably discriminated by different observers and as such 

should be used with caution for scientific assessments. In 

contrast, the Spetzler-Martin grading system and angiographic 

results of endovascular treatment can be used with 

substantial interobserver agreement. NB: Intervals for k 0.8 

to 1 = almost perfect; 0.6 to 0.8 = substantial; 0.4 to 0.6 = 

moderate; 0.2 to 0.4 = fair; 0 to 0.2 = slight; < 0 = agree less 

than the chance. 

KEY WORDS: AVM, endovascular, angiography 


Onyx Randomized Trial for Brain Arteriovenous 

Malformations 

Duckwiler, G. R. · Onyx Brain AVM Trial Investigators 

University of California Los Angeles School of Medicine 

Los Angeles, CA 

PURPOSE 

To evaluate the safety and efficacy of Onyx vs Trufill Nbutyl 

cyanoacrylate in brain arteriovenous malformations 

(BAVMs). 


This trial was a multicenter, prospectively controlled study 

involving patients diagnosed with a brain AVM of Spetzler- 

Martin grade of I-IV for which presurgical embolization was 

indicated. A total of 117 AVM patients were enrolled in this 

Study - 54 in the Onyx group and 63 in the Trufill group. 

Under the protocol, all BAVMs were to be resected surgically 

following embolization. In the event surgical resection 

was not performed, or if the AVM was resected only partially, 

follow up was to be conducted at 3 months, 12 months, 

and annually thereafter for 3-5 years. The embolization procedure 

for Trufill patients was performed using the standard 

techniques of the local investigator. For Onyx, embolization 

required the use of a special dimethyl sulfoxide compatible 

microcatheter. The primary efficacy endpoint was reduction 

of the AVM size equal to or greater than 50%, as measured 

by core lab. The safety endpoint compared serious adverse 

event rates between groups and clinical neurologic evaluations. 

Secondary efficacy endpoints were blood loss at surgery 

and surgical resection time. 

RESULTS 

At this time, 114 patients have completed evaluation by the 

core lab. Ninety-two percent of the Onyx patients and 82% 

of the Trufill patients achieved primary endpoint success. 

There were no significant differences in safety and no significant 

differences in the secondary efficacy endpoints 

between the two groups. Fourteen patients will require follow 

up annually per protocol. 

CONCLUSION 

Successful completion of the Onyx Randomized Trial for 

Brain Arteriovenous Malformations has been achieved 

which will compare the safety and efficacy of Onyx vs 

Trufill in brain arteriovenous malformations. 

KEY WORDS: Onyx, arteriovenous malformation, Trufill n- 

BCA


Embolization of Brain Arteriovenous Malformations 

with Ethylene Vinyl Alcohol Copolymer Dissolved in 

DMSO (Onyx Liquid Embolic System): Experience in a 

Prospective Series of 51 Patients 

Niemann, D. 1 · Molyneux, A. J. 2 · Moftakhar, R. 1 · Sneade, 

M. 3 · Cowsill, C. 3 · Coley, S. C. 4 

1 2 University of Wisconsin, Madison, WI, Frenchay Hospital, 

Bristol, UNITED KINGDOM, 3Radcliffe Infirmary, Oxford, 

4 UNITED KINGDOM, Royal Hallamshire Hospital, 

Sheffield, UNITED KINGDOM 

PURPOSE 

To determine the safety and effectiveness of Onyx Liquid 

Embolic System (LES) in the treatment of brain arteriovenous 

malformations (BAVMs). 


A consecutive series of 51 unselected patients with BAVMs 

presented to a single neurointerventionalist between January 

1999 and March 2003. The patients were evaluated neurologically 

and assigned Glascow Outcome Scale (GOS) score pretreatment 

and in follow up. MR imaging scans and digital subtraction 

angiography (DSA) were utilized to provide radiologic 

assessment and Spetzler-Martin grades were assigned. All 

the patients were treated with Onyx LES. Some of the patients 

had subsequent surgery or radiosurgery. Clinical and technical 

observations and complications were recorded. 

RESULTS 

Thirty-eight patients have completed treatment. Thirteen of 

38 (34%) have been treated by embolization alone. Six of 38 

(15%) patients have undergone surgical resection and 19 of 

38 (50%) have had radiosurgery postembolization. There has 

been no evidence of reopening of the embolized portions in 

the follow-up period, and there have been no instances of 

rebleeding. With one exception, no patient has had a worse 

GOS score in the follow-up period. There have been 4 permanent 

deficits making the permanent procedural morbidity 

5% (4 of 79 procedures). There have been no adverse toxic 

effects referable to the DMSO either clinically or on MR 

imaging. There have been no deaths in the follow-up period. 

CONCLUSION 

Onyx appears safe and effective, and offers new opportunities 

in the management of patients with BAVMs. 

KEY WORDS: AVM, embolization, Onyx 


Balloon-Assisted N-Butyl Cyanoacrylate Embolization of 

Arteriovenous Malformations 

Linfante, I. · Perlow, A. · Mawad, K. · Gounis, M. · Wakhloo, 

A. K. 

University of Miami 

Miami, FL 

PURPOSE 

Embolization with N-butyl cyanoacrylate (n-BCA) can 

achieve permanent occlusion in arteriovenous malforma- 

47 

tions (AVM). However, the use of such agent is not free of 

complications. Inflating a balloon positioned proximally to 

the site of n-BCA infusion may be of aid in interrupting 

blood flow and therefore better controlling the injection of 

the agent. Newly developed isoprene/polypropylene blend 

balloons allow navigation in tortuous vessels because of 

their high compliance. We therefore report our experience 

with such technique in 4 patients with complex arteriovenous 



Patient 1 is a 16-year-old man with a 30 x 15 cm left shoulder 

AVM with rapid shunting into the venous system. The 

malformation had supply from multiple branches of the subclavian, 

vertebral and bilateral external carotid arteries. 

Patient 2 is a 33-year-old woman with a complex and extensive 

intracranial dural AVM. The patient failed surgical ligation 

of both external carotid arteries. Patient 3 is a 68-yearold 

man with a condylar dural AVM with multiple feeders 

from external carotid and vertebral arteries. Patient 4 is a 43year-old 

man with Barrow type D carotid cavernous fistula. 

The patient had multiple feeders mostly from the left external 

carotid artery. 

RESULTS 

In patient 1 we used a 1:4 ratio of n-BCA/ethiodol (Cordis 

Endovascular, Miami Lakes, FL) in conjunction with a 7 x 7 

Hyperform balloon (Micro Therapeutics, Irvine, CA). The 

combination allowed 4-6 minutes n-BCA infusion into the 

nidus and the feeders. He was treated with multiple sessions 

with significant reduction of the lesion mass allowing the 

patient to undergo surgical resection. In patient 2 a 

Hyperglide 7 x 7 mm balloon (Micro Therapeutics, Irvine, 

CA) was introduced into the anterior 1/3 of the superior 

sagittal sinus. A 1:4 n-BCA/ethiodol ratio was infused over 

4-6 minutes via a 0.014” microcatheter positioned into the 

enlarged dural vein under balloon-induced flow arrest. In 

patient 3 a 7 x 7 mm Hyperform balloon was positioned in 

the inferior petrosal sinus to occlude the condylar draining 

vein. After balloon inflation and subsequent flow arrest, a 

2:1 mixture of ethiodol/n-BCA was infused over 4-6 minutes 

into the venous pouch of the condylar vein via a 0.014” 

microcatheter. The dural fistula then was completely obliterated. 

In patient 4, a 7 x 7 mm Hyperform balloon was positioned 

and inflated in the intercavernous portion of the cavernous 

sinus to arrest blood flow. A 1 to 3 mixture of n- 

BCA/ethiodol then was infused slowly into the left CS via a 

0.014” microcatheter. However, when the n-BCA column 

reached the balloon, it deflated although without complications. 

CONCLUSION 

In all 4 patients the balloon successfully assisted a slow, controlled 

infusion of the n-BCA. However, experimental 

results suggest that caution should be used when combining 

these two tools. In particular, the contact between the n-BCA 

and the isoprene/polypropylene bend of the balloon may 

cause its sudden deflation. A large series of patients is needed 

to evaluate the safety of this technique. 

KEY WORDS: Arteriovenous malformations, balloon-assisted, 

n-BCA 

Monday

Monday 


Strategies for Treatment of Cranial Dural Arteriovenous 

Fistulas 

Stefani, M. A. · Piotin, M. · Spelle, L. · Mounayer, C. · 

Moret, J. 

Fondation Ophtalmologique Adolphe de Rothschild 

Paris, FRANCE 

PURPOSE 

The authors discuss the treatment strategies for dural arteriovenous 

malformations (dAVM), considering anatomical 

features such as the presence of cortical venous reflux 

(CVR) and patterns of cerebral venous drainage and present 

a series of 136 patients treated at our institution. 


Intracranial dAVM treated between 1993 and 2003 were 

reviewed retrospectively. Data regarding clinical presentation, 

treatment, and outcome were noted. Angiographic features 

of the dAVM that were reviewed: patterns of arterial 

and venous anatomy, normal cerebral venous drainage, and 

the presence of CVR. 

RESULTS 

The overall mean ± SD age of the sample at presentation was 

58 ± 14 years, with no significant differences related to gender. 

Male individuals were more prone to present with hemorrhage 

(81.8% as opposed to 41.2% of the women). There was 

a tendency for hemorrhagic presentation among the first three 

decades of life (80%), with a clear predominance of nonhemorrhagic 

presentations on the last category of age (92.3%, p = 

0.018). Bleeding was the initial clinical manifestation in 24 

patients (17.6%). Ophthalmologic signs were present in 44 

cases and the most frequently found manifestations were 

proptosis, pulsatile bruits, visual deficits, pseudo-extraocular 

motor nerve palsies. Pulsatile tinitus and vertigo were noted 

initially in 51 patients. There were four reports of previous 

trauma, three cases with coincident cranial tumors (meningiome, 

glomic, and choleosteatoma), one case with the presence 

of an intracranial aneurysm, and 3 cases with the association 

of pial arteriovenous malformations. The presence of 

CVR was noted in 67 (49%) of the patients and was associated 

statistically with hemorrhagic presentation(OR = 31.4, 

95% CI 4.03 to 238.9). The two predominant locations for the 

dAVf were the transverse/sigmoid (38.2%) and the cavernous 

sinus (30.9%). The presence of CVR was related to the specific 

localization of the dAVf: all tentorial fistulas had CVR in 

contrast with none located at the condyle channel/sigmoid 

sinus. Cavernous locations had venous reflux on 9/42 (21.4%) 

of the cases and lateral sinus fistulas had CVR in 25/52 

(48.1%). The number of endovascular procedures varied from 

1 in 90 cases (66%) up to 8 a single, with different modalities 

treatment: endovascular (117 cases) and surgical (23 cases) 

approaches were employed. Cure was obtained in 60/67 

(90%) of the cases with CVR and in 48/69 (71%) in those 

without CVR. Complications occurred in 6 cases, including 4 

minor and 2 major events. Two patients harboring dAVM with 

cortical venous reflux refused to follow the recommended 

treatment; one patient with a huge dAVM died following the 

endovascular procedure. In four cases, small residual shunts 

were found at angiographic follow ups of dAVM with previous 

CVR and were treated conservatively. In 3 cases the 

dAVM disappeared spontaneously. 

48 

CONCLUSION 

Dural arteriovenous malformations should be treated considering 

the significant risks of hemorrhage related to the presence 

of CVR. Treatment strategies should focus on anatomical 

and clinical cure, achieved with low risk of complications 

when using a multidisciplinary approach, after careful 

understanding of the venous drainage, respecting the anatomy 

of the malformation and the normal cerebral venous 

drainage. 

KEY WORDS: Dural arteriovenous fistula 


Transvenous Approach through the Deep 

Leptomeningeal Veins in the Treatment of Dural 

Arteriovenous Fistulas 

Biondi, A. · Casasco, A. · Sourour, N. · Van Effenterre, R. · 

Chiras, J. 

Pitié-Salpêtrière Hospital-University Paris VI 

Paris, FRANCE 

PURPOSE 

Dural arteriovenous fistulas (DAVFs) with an exclusive retrograde 

leptomeningeal venous drainage are uncommon 

lesions associated with an aggressive natural history and 

poor prognosis. Dural arteriovenous fistulas can drain into 

superficial leptomeningeal veins draining the cortex of the 

cerebral or cerebellar hemispheres (cortical venous 

drainage) or into deep leptomeningeal veins draining noncortical 

structures (i.e., the diencephalon and brainstem). 

Dural arteriovenous fistulas with an exclusive retrograde 

leptomeningeal deep venous drainage usually are treated 

with transarterial embolization and/or surgery and only few 

cases treated with transvenous embolization have been 

reported. The purpose of our study was to evaluate the percutaneous 

transvenous approach through the deep (noncortical) 

leptomeningeal veins in the treatment of DAVFs. 


Retrospective chart analysis and radiologic studies evaluation 

were carried out in 8 patients with a DAVF treated by percutaneous 

embolization with retrograde navigation through the 

deep venous system (straight sinus, vein of Galen, basal vein, 

latero-mesencephalic vein). There were 6 men and 2 women 

ranging in age from 37 to 68 years. Clinical findings included 

hemorrhage in 5 patients (4 patients had recovered from hemorrhage 

while one patient presented a severe clinical status), 

symptoms of raised intracranial pressure in 2, and in another 

patient the DAVF was diagnosed by chance on a MR study 

performed because of hypoacusia. All patients had a superior 

petrosal sinus (SPS) DAVF. The exclusive retrograde leptomeningeal 

venous drainage involved mainly the latero-mesencephalic 

vein. Venous drainage through the SPS was never 

visualized. A venous pouch along the venous drainage was 

observed in all cases. A guiding catheter was positioned in the 

transverse/sigmoid or straight sinuses and a microcatheter was 

navigated through the deep venous system up to the fistula site 

(the stem of the petrosal vein) where coils were detached. 

RESULTS 

Transvenous route allowed DAVF occlusion in 7 cases and 

failed in one case in which a subsequent transarterial 

embolization allowed occlusion of the DAVF. After tranve-

nous embolization, 4 patients, who had recovered from the 

hemorrhage, remained asymptomatic. One patient with 

symptoms of raised intracranial pressure was clinically 

cured. The patient with the DAVF diagnosed by chance presented 

a transient partial internuclear ophthalmoparesis due 

to a slight bleeding during venous catheterization. The 

patient presenting with a severe clinical status due to DAVF 

hemorrhage had further bleeding due to intraprocedural perforation 

of a venous pouch. Despite this new bleeding no 

significant clinical changes were observed and his clinical 

status slightly improved a few days after ventriculoperitoneal 

shunt insertion. However this patient died 1 week 

after the procedure from cardiopulmonary complications. 

CONCLUSION 

Prompt therapy is mandatory in DAVFs with exclusive retrograde 

leptomeningeal deep venous drainage because of 

their poor prognosis. The retrograde navigation through the 

deep leptomeningeal veins requires caution because of the 

risk of vessel perforation especially at level of the venous 

pouches which are the weakest structures to pass. Although 

challenging, the transvenous approach through the deep leptomeningeal 

veins is feasible and effective in the treatment 

of selected cases of DAVFs. 

KEY WORDS: Dural fistula, transvenous, embolization 


CT Angiography of Spinal Vascular Malformation 

Near, L. 1 · Xavier, A. R. 1 · Gupta, G. 1 · Prestigiacomo, C. J. 1 · 

Farkas, J. 2 

1 New Jersey Medical School/University of Medicine & 

Dentistry of New Jersey, Newark, NJ, 2 Maimonides Medical 

Center, Brooklyn, NY 

PURPOSE 

CT angiography (CTA) is being used iincreasingly n the 

diagnosis of cerebrovascular diseases including intracranial 

aneurysms and cerebral arteriovenous malformations 

(AVM). CT angiography and postprocessed three-dimensional 

reconstructed images help in presurgical evaluation 

and therapeutic decision-making of cerebral AVMs. We 

would like to describe the case of a patient where the diagnosis 

of a spinal vascular malformation was made using 

CTA. 


A 15-year-old boy presented with recurrent nausea, vomiting, 

and headache. Clinical examination was significant for 

minimal weakness on his left side and hyper-reflexia with no 

ataxia or cranial nerve findings. His sensory examination 

was normal. MR imaging of the brain and the upper spinal 

cord was suggestive of an AVM with venous edema. CT 

angiogram showed dilated vascular channels running bilaterally 

from the left petrosal region to the cervical cord suggesting 

a spinal dural AVM. The patient underwent successful 

glue embolization of the left meningohypophyseal trunk 

with complete occlusion of the vascular malformation and 

no residual fistula. He made a rapid clinical recovery. 

RESULTS 

MR imaging of the brain and the upper spinal cord: 

Nonenhancing lesion, hyperintense on FLAIR and T2 

49 

sequences in the medulla oblongata with multiple signal 

voids around the medulla and the upper cervical cord (Fig. 

A). CT angiogram extending from the thoracic aorta to the 

circle of Willis: Source and 3D reconstructed images showed 

an enlarged left menigohypophyseal trunk feeding dilated 

vascular channels running inferiorly on both sides of the 

brainstem, extending to the C5 vertebral body level (Fig. B). 

These findings were confirmed by the cerebral angiogram. 

Postembolization angiogram: No residual vascular malformation. 

CONCLUSION 

CT angiography can be used to diagnose spinal AVMs. 

Whether CTA is a viable imaging alternative to conventional, 

catheter based, invasive angiography is under study. 

KEY WORDS: Dural AV fistula, spinal vascular malformation, 

embolization 


3:00 PM - 4:30 PM 

Room 107 

(17c) HEAD AND NECK: Neck, New 

Techniques, and Excerptas 




Carotid 




Moderators: Edward E. Kassel, MD 

Suresh K. Mukherji, MD 

Monday

Monday 


Ectopic Parathyroid Adenoma: A Retrospective Analysis 

of Radiographic Findings 

Urena, R. J. · Hudgins, P. A. · Weber, C. · Miller, J. · 

Mansour, K. · Esteves, F. 

Emory University School of Medicine 

Atlanta, GA 

PURPOSE 

The purpose of this presentation was to describe the most 

common location and imaging appearance, on parathyroid 

Sestamibi I 123 nuclear medicine scans and cross-sectional 

imaging, of ectopic hyperparathyroidism in a surgical series 

of 35 cases. The advantages and limitations of both imaging 

modalities were determined in this retrospective series. 


Thirty-five patients with suspected ectopic hyperparathyroidism 

comprised this retrospective series. Preoperative 

imaging was correlated with surgical findings in all patients. 

Preoperative diagnostic studies including CT, MR imaging 

and nuclear medicine were reviewed individually. CT and 

MR imaging and nuclear medicine studies then were 

reviewed together in patients who had both examinations. 

Intrathyroidal and thymic adenomas were not included in 

this series. 

RESULTS 

Twenty-four of 35 patients had primary hyperparathyroidism, 

8/35 had secondary and 3/35 had tertiary. 

Preoperative evaluation included: 29/35 had both a CT and 

nuclear medicine scan; 1/35 had MR and a nuclear medicine 

study; 1/35 had a CT scan alone, and 4/35 had only nuclear 

medicine scans. In 32/34 cases, the Sestamibi scan was positive. 

One nondiagnostic Sestamibi scan was degraded by 

artifact, and the second non-diagnostic scan was in a complicated 

patient with secondary hyperparathyroidism and 

multiple prior surgeries. Cross-sectional imaging was able to 

identify most cases; however, the addition of the nuclear 

medicine study increased sensitivity and specificity. On CT, 

the lesions enhanced intensely in approximately 60% of the 

cases, only minimally in the remaining lesions, and 2/30 had 

peripheral rim enhancement. The most common locations 

were para-esophageal or anterior mediastinum. Two unusual 

cases, one in the retropharynx and one right pre-carinal, were 

included. 

CONCLUSION 

This large, surgically proven series included patients with 

primary, secondary, and tertiary hyperparathyroidism. 

Lesions were detected almost always on Sestamibi I 123 scans, 

but high-resolution CT better localized adenomas and relationship 

to major vascular and neural structures. While most 

cases of ectopic hyperparathyroidism were localized by CT, 

not all lesions enhanced, and nuclear medicine study 

increased specificity. 

KEY WORDS: Parathyroid adenoma, hyperparathyroidism 

50 


Ultrasonography-Guided Fine Needle Aspiration Biopsy 

of a Newly Detected Nodule of the Thyroid Bed in 

Postoperative Patients for Thyroid Carcinoma 

Lee, J. · Lee, H. · Kim, H. · Lee, D. · Choi, C. · Kim, S. · Suh, D. 

Asan Medical Center 

Seoul, REPUBLIC OF KOREA 

PURPOSE 

To evaluate the diagnostic ability of ultrasonography-guided 

fine needle aspiration biopsy (USG FNAB) of a newly 

detected nodule of the thyroid bed for diagnosing recurrent 

carcinoma and to find ultrasonographic (USG) findings suggesting 

recurrent thyroid carcinoma in postoperative 

patients. 


We underwent USG FNAB of a newly detected nodule of the 

thyroid bed in 40 patients with a history of total thyroidectomy 

or lobectomy for thyroid carcinoma. The maximum 

diameter of the aspirated nodules was evaluated for the probability 

of getting an adequate specimen for cytologic diagnosis. 

We evaluated the USG findings of the nodules including 

the presence of microcalcification, an irregular margin, 

heterogeneity, and a taller than wide shape for the sensitivity, 

specificity, and positive predictive values for recurred 

malignancy. 

RESULTS 

The cytologic diagnosis from the USG FNAB was conclusive 

in 32 (80%) of 40 patients. Nineteen were diagnosed as 

recurrent tumor (all papillary carcinoma), and 14 of them 

were confirmed subsequently by surgery. Another 13 

patients diagnosed as benign (5 lymphoid hyperplasia, 4 

postoperative changes, and 4 residual benign thyroid tissue). 

The mean value of the nodule size was 0.83 cm with a range 

from 0.27 - 2.50 cm. The maximum diameter larger than 0.5 

cm was related significantly to the probability of getting an 

adequate specimen for cytologic diagnosis (p = 0.042). The 

rate of malignancy was significantly higher in nodules with 

an irregular margin (p < 0.001), dot-like heterogeneity (p = 

0.001), and a taller than wide shape (p = 0.016). Linear heterogeneity 

was only found in benign cytology groups with 

81.3% of negative predictive value. 

CONCLUSION 

Ultrasonography-guided fine needle aspiration biopsy of a 

newly detected nodule of the thyroid bed in postoperative 

patients for thyroid carcinoma showed conclusive results in 

80% of patients. The rate of recurrent malignancy was significantly 

higher in the nodules with an irregular margin, 

dot-like heterogeneity, and a taller than wide shape. 

KEY WORDS: Thyroid cancer, ultrasonography, fine-needle 

aspiration biopsy


Positron Emission Tomography/CT of Cystic Neck 

Masses: Trust Your CT 

Branstetter,IV, B. F. · Ferris, R. 

University of Pittsburgh 


PURPOSE 

Positron emission tomography (PET)/CT has become an 

accepted means of staging and monitoring head and neck 

cancer. In the setting of a neck mass without a histopathologic 

diagnosis, PET/CT also can be useful to distinguish 

benign from malignant cases. However, the PET component 

of the examination can be misleading in the setting of a cystic 

neck mass because the cystic component of the mass will 

not have FDG uptake. The purpose of this study is to determine 

whether PET/CT is a valuable modality for the evaluation 

of cystic neck masses. 


Four consecutive patients with cystic neck masses equivocal 

for malignancy were included in this study. After equivocal 

fine needle aspiration, each patient underwent a combined 

PET/CT evaluation. The PET component of each exam was 

evaluated by a nuclear medicine physician, and the CT component 

was evaluated by a head and neck radiologist; each 

had access to the full fused dataset. Discrepancies between 

the PET and CT readings were noted and compared to final 

pathology. 

RESULTS 

One of the four patients had a benign branchial cleft cyst, 

and three had squamous cell carcinoma. In each of the four 

cases, the PET reading was either equivocal or misleading, 

whereas the CT reading suggested the correct diagnosis. 

CONCLUSION 

Before a diagnosis of malignancy has been established, combined 

PET/CT of a cystic neck mass is of little value over CT 

alone. Cystic lymphadenopathy is evaluated inadequately 

with PET alone. 

KEY WORDS: PET/CT, neck cyst, neck mass 


Role of Apparent Diffusion Coefficient Map in 

Differentiating Benign from Malignant Parotid Tumors 

Abdel Razek, A. A. · Elhawarey, G. · Albadrawey, A. · 

Elserougy, L. · Kamel, Y. 



PURPOSE 

To determine the role of apparent diffusion coefficient map 

in differentiating benign from malignant parotid tumors. 


This study included 41 patients (21M, 20F aged 10-72 years: 

mean 47 years) with parotid mass. They underwent routine 

pre and postcontrast MR imaging and diffusion-weighted 

imaging on a 1.5 T MR unit (Symphony-Siemens). Diffusion 

51 

MR imaging was done using spin-echo type of single-shot 

echo-planar imaging (EPI) with a diffusion-weighted factor, 

factor b of 0.500 and 1000 sec/mm 2 . The apparent diffusion 

coefficient (ADC) map was reconstructed. The signal intensity 

was assessed visually on ADC maps and ADC value was 

measured in parotid lesion. The mean ADC values correlated 

with histopathologic results. 

RESULTS 

Adequate ADC maps were obtained in 39 patients. The 

benign tumors shows high SI (n = 23) or mixed SI (n = 4) 

and malignant tumors showed low SI (n = 11) or mixed SI (n 

= 3). The mean ADC values of benign tumors was 1.49 ± 

0.18 X 10 -3 mm 2 /sec and in malignant tumors was 0.89 ± 0.12 

X 10 -3 mm 2 /sec. The mean ADC value of the malignant 

tumor was significantly different than that of benign parotid 

tumors (P < 0.004). 

CONCLUSION 

Apparent diffusion coefficient map is a new imaging modality 

for differentiating benign from malignant parotid tumors. 

Also, it can be used for characterization of parotid tumors. 

KEY WORDS: Parotid, tumors, diffusion 


Role of Apparent Diffusion Coefficient Map in 

Differentiating Recurrent Head and Neck Tumors from 

Postradiation Changes 

Abdel Razek, A. A. · Soliman, N. · Kandeel, A. · Kamel, Y · 

Elshenshawy, H. 



PURPOSE 

To evaluate the role of apparent diffusion coefficient (ADC) 

map in differentiating recurrent head and neck tumors from 

postradiation changes. 


This study included 29 patients (21M, 8F aged 53-72 years: 

mean 67 years) with suspected clinically recurrent head and 

neck tumor after complete course of radiotherapy. Ten of 

these patients underwent previous operation. They underwent 

routine postcontrast MR imaging. Diffusion-weighted 

imaging was done on a 1.5 T whole body scanner 

(Symphony, Siemens) using spin-echo type of single-shot 

echo-planar imaging. The parameters used were TR/TE: 

4200/38 msec, B value = 0, 500 and 1000 sec/mm 2 . From 

these images apparent diffusion coefficient (ADC) maps 

were generated. The signal intensity was assessed visually 

on ADC maps and ADC value was measured in suspected 

lesion. The mean ADC values correlated with histopathologic 

results. Statistical analysis was done. 

RESULTS 

Adequate ADC maps were obtained in 28 patients. The 

smallest recurrent tumor detected was 1.2 cm. On ADC map, 

the recurrent tumor showed low (n = 17) or mixed (n = 4) 

signal intensity and postradiation changes revealed relatively 

high (n = 5) or mixed (n = 3) signal intensity. The mean 

ADC value of the recurrent tumor (1.19 ± 0.23 X 10 -3 

mm 2 /sec) was smaller than that of postradiation changes 

Monday

Monday 

(2.42 ± 0.11 X 10 -3 mm 2 /sec). There was statistical difference 

in mean ADC values between recurrent tumor and postradiation 

changes (p < 0.001). 

CONCLUSION 

Apparent diffusion coefficient map is a new promising imaging 

modality for differentiating recurrent head and neck 

tumors from postradiation changes. So, ADC map can be 

used to follow up head and neck cancer patients after radiation 

treatment. 

KEY WORDS: Diffusion, recurrence 


Diffusion-Weighted Imaging of Paraganglioma: 

Preliminary Experience in Diagnosis and Differential 

Diagnosis 

Aschenbach, R. · Esser, D. · Basche, S. 

HELIOS-Klinikum Erfurt GmbH 

Erfurt, GERMANY 

PURPOSE 

The aim of this work was to evaluate diffusion-weighted 

imaging in diagnosis and differential diagnosis of paragangliomas 

of the head and neck region. 


As paragangliomas are characterized by a strong vascularization 

and possible infiltration of skull base sometimes differential 

diagnosis is difficult. Using diffusion-weighted 

imaging paragangliomas show a typical high signal and low 

apparent diffusion coefficient. 

RESULTS 

Similar appearing lesions in conventional MR imaging of 

skull base (e.g., neurinomas, meningeomas, and giant cell 

tumors) show significant different apparent diffusion coefficient. 

Diffusion-weighted imaging also can be helpful in 

diagnosis of recurrent paraganglioma after surgery or stereotactical 

radiation and differentiate scar tissue from recurrent 

paraganglioma. 

CONCLUSION 

In conclusion we presumed that diffusion-weighted imaging 

is a new tool in diagnosis and characterization of skull base 

lesions and their differential diagnosis. 

KEY WORDS: Diffusion-weighted imaging, paraganglioma, 

head and neck imaging 


Migration of a Temporal Fossa Hemangioma 

Bradbury, M. S. · Karimi, S. · Boyle, J. 


New York, NY 

PURPOSE 

To report a case of migrating hemangioma of the temporalis 

fossa and discuss surgical implications and relevant anatomy. 

52 


The patient, a 36-year-old male, presented with a history of 

an asymptomatic palpable mass in the left temporalis fossa, 

overlying the lateral orbital ridge. The patient had undergone 

an initial failed resection of the lesion. Using an endoscopic 

procedure, a 2-inch incision was made above the left orbit. 

The tumor could not be visualized, and the incision was converted 

subsequently to a larger hemi coronal incision, with 

dissection proceeding along the periosteum and temporalis 

muscle. Even given this greater exposure, the tumor was 

unable to be located, and the procedure was aborted. The 

patient subsequently presented to our institution for further 

management. A 2.5 cm mass was eventually excised from 

the buccal space via an intraoral approach and a gingivobuccal 

incision. 

RESULTS 

The initial outside MR imaging study revealed a 2.5 cm well 

circumscribed, homogeneously enhancing mass within the 

left temporalis fossa, superficial to the muscle (Fig. 1). 

Following the initial endoscopic procedure, MR imaging 

was performed. No mass was identified within the left temporalis 

fossa, although a mass was seen within the buccal fat, 

suggesting that the mass had migrated from its original location 

(Fig. 2). A final MR image was obtained after intraoral 

resection and revealed no evidence of residual tumor. 

CONCLUSION 

This case illustrates the importance of selecting the proper 

surgical approach given the relevant imaging findings, as 

well as highlights critical anatomical considerations. 

Successful excision of the tumor warranted proper surgical 

technique, namely a hemi-coronal incision with scalp flap 

creation. The initial superficial location of this freely mobile 

lesion mandated that care should have been taken not to violate 

fascial integrity, given the increased potential for migration 

deep to the disrupted fascia. 

KEY WORDS: Migrating hemangioma, temporalis fossa mass 


Intraorbital Foreign Bodies: CT and MR Findings 

Rotblat, M. · Lee, H. 

New Jersey Medical School/University of Medicine & 

Dentistry of New Jersey, University Hospital 

Newark, NJ 

PURPOSE 

An orbital foreign body can be overlooked because a penetrating 

wound may be accompanied by edema, emphysema,

and minimal inflammatory changes in the early clinical 

course. We describe the CT and MR findings of variable foreign 

bodies in the orbit for establishing the early detection. 


Seven patients with orbital foreign bodies are presented. Two 

patients’ orbits contained wood, one a pencil, one a ball point 

pen, one shrapnel from an explosion, one pellet from a BB 

gun, and finally one with a glass fragment. One patient had 

both CT and MR examinations. Six patients had CT examinations. 

RESULTS 

The wooden foreign bodies were of variable densities on CT. 

They also contained linear areas of air within the wood internally. 

In two patients the wooden foreign body was not identified 

initially. These patients had follow-up examinations at 

22 and 60 days. Both patients subsequentially developed 

orbital infections. The follow-up examinations in these individuals 

again showed persistent linear intraorbital emphysema 

with new inflammatory changes. The CT of the pencil 

shows a cylinder of air with a central hyperdensity. The 

remaining foreign bodies were hyperdense on CT without 

orbital emphysema. 

CONCLUSION 

Prompt surgical removal of foreign bodies in the orbit is 

essential. Identification of wooden foreign body is a challenge 

with cross-sectional imaging. CT and MR examinations 

of wooden foreign bodies were characterized by trapping 

air within the wood which is linear in shape. This is 

seen on the initial study and can persistent for up to 2 

months. Therefore, in the setting of trauma, orbital emphysema 

without an associated orbital wall fracture should raise 

suspicion for the presence of an intraorbital wooden foreign 

body. All other foreign bodies are hyperdense on CT and easily 

identified. 

KEY WORDS: Orbit, foreign body, wooden-glass 


MR Imaging and CT Findings in Bilateral Retinal 

Detachment Associated with Vogt-Koyanagi-Harada 

Syndrome 

Camacho, A. C. · Chaljub, G. · McGehee, B. E. · Ethridge, K. P. 

University of Texas 

Galveston, TX 

MR and CT images demonstrating bilateral retinal detachment 

and chorioretinitis in a patient with Vogt-Koyanagi- 

Harada disease will be shown. These specific findings can 

aid the clinician in making the diagnosis of this relatively 

rare autoimmune disease. Review of the diagnostic criteria 

for the disease will be outlined along with imaging findings. 

KEY WORDS: Retinal detachment, Vogt-Koyanagi-Harada, 

MR imaging 

53 


Ameloblastic Carcinoma of the Mandible 

Dunfee, B. L. · Sakai, O. · Pistey, R. · Barest, G. 

Boston University Medical Center 

Boston, MA 

PURPOSE 

Ameloblastic carcinoma (AC) is a rare, malignant epithelial 

odontogenic tumor that histologically retains the features of 

ameloblastic differentiation and exhibits cytological features 

of malignancy in the primary or recurrent tumor. This tumor 

may develop within a preexisting ameloblastoma or arise de 

novo or from odontogenic cyst. This case report describes 

CT and MR findings of an ameloblastic carcinoma of the 

mandible with a complete review of clinical presentation, 

imaging findings, gross and histopathology. 


A 16-year-old male presented to the otolaryngology clinic 

with a several-month history of a progressively growing tender 

mass on the right mandible. He denied any difficulty 

breathing or eating; however, he was unable to fully occlude 

his teeth. On physical examination, the patient was well 

nourished and developed with an obvious right mandibular 

and facial mass swelling. The football-shaped mass was firm 

on palpation and extended up the right ramus and canine 

tooth. However, no significant lymphadenopathy was appreciated. 

RESULTS 

CT imaging revealed a 6.0 x 8.0 cm, expansile mass displacing 

the posterior body and angle of the right mandible with 

associated cortical disruption. An erupted molar tooth was 

noted also in this region. There was no definite evidence of 

invasion of other adjacent structures. The mass demonstrated 

peripheral enhancement with contrast imaging. However, 

there was no suggestion of metastatic disease. On MR imaging, 

the expansile mass enhanced heterogeneously with contrast. 

The scattered, unenhanced areas of the mass were 

thought to represent necrosis versus cystic degeneration. The 

margins of the mass were clear with normal fatty marrow 

identified about 2.5 cm in length from the articular surface. 

Although the mass crossed the midline and displaced the 

tongue and other adjacent structures, there was no evidence 

of infiltration or invasion of the surrounding soft tissues. The 

vascular and nerve structures appeared unaffected. On gross 

pathologic review, the tumor was well circumscribed with 

multiple white lobules on cut sections. Histology revealed 

sheets of atypical spindled epithelial cells exhibiting nuclear 

pleomorphism. These regions were separated by areas of 

hyalinized fibrous stroma. 

Monday

Monday 

CONCLUSION 

We report CT and MR findings in a case of ameloblastic carcinoma 

of the mandible. Ameloblastic carcinoma shares similar 

radiologic findings with ameloblastomas. More aggressive 

features such as cortical disruption and solid components 

should raise the possibility of ameloblastic carcinoma 

despite the rarity of this tumor. 

KEY WORDS: Ameloblastic carcinoma, mandible, odontogenic 

tumors 


Central Nervous System Metastases Arising from 

Carcinoma Ex Pleomorphic Adenoma of the Parotid 

Gland 

Sheedy, S. P. · Welker, K. M. · DeLone, D. R. · Gilbertson, J. R. 

Mayo Clinic 

Rochester, MN 

PURPOSE 

To present the clinical and radiographic features of a rare 

case of biopsy proven brain and spinal cord metastases arising 

from carcinoma ex pleomorphic adenoma of the parotid 

gland. 


The patient was a 36-year-old male with a known history of 

grade 4 carcinoma ex pleomorphic adenoma of the right 

parotid gland with local nodal and extranodal extension. He 

had undergone right total parotidectomy with right modified 

radical neck dissection and postoperative adjuvant radiation 

therapy. Nineteen months after completion of radiotherapy, 

he presented to the emergency department with a 3-week 

history of malaise, fever, gait imbalance, and fatigue. 

Despite treatment for pneumonia, his symptoms progressed 

and he began to experience a decline in his mental status and 

generalized weakness. On physical exam, he was found to 

54 

have palpable nodules in the left submandibular region and 

in the right supraclavicular region. He was admitted for further 

workup including FNA of a left submandibular lymph 

node, which revealed metastatic grade 4 (of 4) adenocarcinoma. 

RESULTS 

MR imaging of the brain demonstrated multiple enhancing 

nodular lesions throughout both cerebral and cerebellar 

hemispheres, the basal ganglia, right thalamus, cerebral 

peduncles, pons, medulla, and right optic tract. Several 

demonstrated an unusual lamellated pattern of enhancement. 

Many of these lesions were associated with significant vasogenic 

edema. There was no abnormal diffusion restriction. 

MR imaging of the cervical and thoracic spine showed an 

enhancing intramedullary nodular lesion at T1. Additionally, 

there were patchy regions of increased T2 signal in the spinal 

cord extending from the level of C2 to T3. Because the 

patient had a prior history of extraplumonary tuberculosis 

involving the genitourinary tract, a stereotactic biopsy of the 

right frontal lobe was performed to exclude disseminated 

tuberculosis as the cause of the above findings. The biopsy 

revealed high-grade carcinoma consistent with metastasis 

from carcinoma ex pleomorphic adenoma of a parotid primary. 

CONCLUSION 

Although pleomorphic adenomas are common benign 

tumors of the parotid glands, they undergo malignant transformation 

in up to 25% of untreated cases. It is not unusual 

for carcinoma ex pleomorphic adenoma to recur locally or to 

metastasize to regional or distant sites (1,2). However, to our 

knowledge, there are no reports in the radiology literature of 

brain and spinal cord metastases arising from carcinoma ex 

pleomorphic adenoma of the parotid gland. This report illustrates 

a potential devastating consequence, central nervous 

system metastases, due to malignant transformation of a 

pleomorphic adenoma. Consequently, this case strengthens 

the argument for surgical excision of benign parotid pleomorphic 

adenomas whenever feasible. 

REFERENCES 

1. Olsen KD, Lewis JE. Carcinoma ex pleomorphic adenoma: a 

clinicopathologic review. Head Neck 2001;23:705-712 

2. Czader M, Eberhart CG, Bhatti N, Cummings C, Westra WH. 

Metastasizing mixed tumor of the parotid: initial presentation 

as a solitary kidney tumor and ultimate carcinomatous 

transformation at the primary site. Am J Surg Pathol 

2000;24:1159-1164 

KEY WORDS: Parotid, carcinoma, metastases 


Multiple Myeloma Involving the Thyroid Cartilage 

Kalina, P. · Lane, J. 

Mayo Clinic 

Rochester, MN 

PURPOSE 

To describe the CT, MR, and PET appearance of a multiple 

myeloma involving the thyroid cartilage.


A 60-year-old male presented with 3 months of increasing 

hoarseness. He has a history of multiple plasmacytomas 

including maxillary sinus, frontal sinus, and nasal cavity. 

RESULTS 

Neck MR imaging revealed a 4 cm mass arising from the larynx, 

centered on the right thyroid cartilage extending from 

the hyoid bone to the cricoid cartilage. CT confirmed 

involvement of the right lamina of the thyroid cartilage. F-18 

FDG PET scan with CT fusion images demonstrated 

increased metabolic activity in the right thyroid cartilage. 

Follow-up CT and PET at 4 months following treatment 

revealed considerable decrease in the size and metabolic 

activity of the right thyroid cartilage mass. 

CONCLUSION 

Extraosseous involvement of the larynx/thyroid cartilage by 

multiple myeloma (extramedullary plasmacytoma) is 

exceedingly rare. Few sporadic case reports exist. 

Ossification of the thyroid cartilage with formation of a marrrow 

space may explain this unusual finding. We describe 

the pre and posttreatment CT, MR, and PET appearance of 

this atypical manifestation. 

KEY WORDS: Plasmacytoma, larynx 

Discussion 


3:00 PM - 4:44 PM 

Room 205 

(17d) ADULT BRAIN: Infarction and 

Vasospasm 




Carotid 




and Excerptas 

Moderators: Pamela W. Schaefer, MD 


55 


Perfusion CT Evaluation of Cerebral Vascular 

Autoregulation in a Large Series of Acute Stroke Patients 

Wintermark, M. 1 · Flanders, A. E. 2 · Velthuis, B. 3 · Pedraza, 

S. 4 · Goldsher, D. 5 · van Leeuwen, M. 3 · Anderson, J. 6 · 

Nesbit, G. 6 · Meuli, R. 7 

1University of California San Francisco, San Francisco, CA, 

2Thomas Jefferson University Hospital, Philadelphia, PA, 

3University Medical Center Utrecht, Utrecht, THE 

NETHERLANDS, 4Hospital Doctor Josep Trueta, Girona, 

SPAIN, 5Rambam Medical Center, Haifa, ISRAEL, 6Oregon Health & Science University, Portland, OR, 7Lausanne University Hospital, Lausanne, SWITZERLAND 

PURPOSE 

In acute stroke patients, cerebral vascular autoregulation 

attempts to compensate for reduced local brain perfusion 

consecutive to vascular occlusion. It involves recruitment of 

collaterals and vasodilatation, and is the hallmark of tissue at 

risk or penumbra. Perfusion CT (PCT) has been suggested in 

pilot studies as able to assess cerebral vascular autoregulation 

in acute stroke patients. The purpose of this study was 

to evaluate this hypothesis in a large series of patients. 


As part of a prospective international multicenter trial, 130 

acute stroke patients were enrolled. Inclusion criteria were: 

admission to the emergency room with symptoms suggesting 

hemispheric stroke lasting 12 hours or less, and no evidence 

of intracerebral hemorrhage. An acute and/or follow-up MR 

scan confirming the final diagnosis of hemispheric ischemic 

stroke was obtained. CT and MR angiography were performed 

to assess arterial recanalization or persistent arterial 

occlusion. Diffusion-weighted and perfusion-weighted 

imaging when available, were evaluated to delineate the 

acute and final infarct core, and the tissue at risk. Perfusion 

CT maps were assessed for areas with reduced regional cerebral 

blood flow (rCBF), low regional cerebral blood volume 

(CBV), and increased mean transit time (MTT), both in a relative 

(comparison of the ischemic hemisphere to the healthy 

one) and an absolute quantitative way. ROC analysis was 

used to determine the best PCT map, and the best threshold 

for each map. 

RESULTS 

Ninety-five patients currently have been enrolled in the 

study. The PCT parameter showing the best correlation for 

the identification of ischemic tissue with preserved autoregulation 

is increased relative MTT (area under the curve = 

0.92). The optimal threshold is a MTT value increased to 

145% of the contralateral value. The PCT parameter showing 

the best correlation for the identification of ischemic tissue 

with altered autoregulation is lowered absolute rCBV 

(area under the curve = 0.89), with an optimal threshold at 

2.0 [cc x 100 g-1]. 

CONCLUSION 

Perfusion CT provides insight into cerebral vascular autoregulation 

in acute stroke patients. It features distinct patterns 

possibly affording identification of tissue at risk or penumbra. 

KEY WORDS: Acute stroke, autoregulation, perfusion CT 

Monday

Monday 


Modified Aspect Score on Diffusion/Perfusion MR 

Imaging Correlates Strongly with National Institute of 

Health Stroke Scale Score in Acute Stroke 

Schaefer, P. W. · Mehta, A. · Camargo, E. · Schwamm, L. H. 

· Koroshetz, W. J. · Gonzalez, R. G. · Lev, M. H. 

Massachusetts General Hospital, Harvard Medical School 

Boston, MA 

PURPOSE 

The Alberta stroke program early CT score (ASPECTS) is a 

grading system developed to assess early ischemic CT 

changes in patients with acute middle cerebral artery (MCA) 

stroke. This CT ASPECTS score correlates highly with acute 

and follow-up neurologic assessment tests. Our purpose was 

to determine if a modified ASPECT score (mASPECT) - 

based on degree of involvement in the entire MCA territory, 

and obtained from MR perfusion and diffusion-weighted 

images (PWI/DWI) - could accurately predict admission 

clinical status, measured by the NIH stroke scale score 

(NIHSS). 


Forty-nine patients (18 female, 31 male; ages 16-98 years, 

mean 69 years) with acute MCA stroke, and assessed with 

MR DWI/PWI within 12 hours of symptom onset, were 

evaluated. Two neuroradiologists obtained mASPECT 

scores on DWI, cerebral blood volume (CBV), cerebral 

blood flow (CBF), and mean transit time (MTT) maps. Ten 

regions were evaluated: 1) lentiform nucleus, 2) internal capsule, 

3) caudate nucleus, 4) insula region, 5) inferior frontal 

gyrus region, 6) anterior temporal lobe region, 7) posterior 

temporal lobe region, 8) middle frontal gyrus region, 9) precentral 

central gyrus region, and 10) postcentral gyrus/inferior 

parietal lobe region. Each region was rated on a 1-4 

scale: 1 = uninvolved, 2 = possibly involved, 3 = < 50% 

involved, 4 = > 50% involved. For each patient and map, 

scores for each region were added to obtain a mASPECT 

score. An experienced neurologist obtained NIHSS scores 

from medical records. Spearman correlation coefficients for 

NIHSS versus DWI, CBV, CBF and MTT mASPECT scores 

were obtained for all patients and for those who presented 

within 6 hours. 

RESULTS 

Eighteen patients were imaged within 6 hours (early), and 31 

patients between 6-12 hours of stroke onset. NIHSS scores 

ranged from 0-29 (mean 8.5). mASPECT scores ranged from 

12-32 (mean 18.3) for DWI; 12-33 (mean 18.5) for CBV; 12- 

39 (mean 21.2) for CBF; and 12-38 (mean 21.6) for MTT 

maps. Interobserver agreement was good, with at most a 3point 

deviation for all mASPECTS scores. The NIHSS and 

mASPECT scores correlated highly for all maps - particularly 

for the 0-6 hour group - as follows: for CBV maps, r = 

0.75 all, r = 0.87 early; for DWI maps r = 0.72 all, r = 0.87 

early; for CBF maps r = 0.71 all, r = 0.86 early; for MTT 

maps r = 0.70 all, r = 0.85 early. 

CONCLUSION 

The mASPECT score is a reliable grading system for assessing 

acute ischemic changes on MR DWI/PWI in acute MCA 

stroke patients. Admission mASPECT score correlates highly 

with admission NIHSS score, and provides an excellent 

56 

assessment of acute clinical status, which may be valuable in 

the management of acute stroke patients. Future investigation 

in a larger patient number may reveal differences in 

mASPECT scores between infarct “core” (measured by DWI 

and CBV maps) and ischemic “penumbra”(measured by 

CBF and MTT maps), especially if those scores are location 

weighted. Future investigation also may reveal if the 

mASPECT score is useful for predicting clinical outcome. 

KEY WORDS: MR perfusion, NIHSS, aspects 


Detection of Parenchymal Hemorrhage in Acute 

Ischemic Stroke: CT versus Echo-Planar Gradient-Echo 

MR Imaging 

Mikulis, D. J. 1 · Taylor, K. 1 · Farb, R. 1 · Willinsky, R. 1 · 

Rowan, S. 1 · Kassner, A. 2 · Silver, F. 1 

1The Toronto Western Hospital, Toronto, ON, CANADA, 

2University of Toronto (University Health Network), 


PURPOSE 

This study compares CT vs echo-planar (EPI) gradient-echo 

MR imaging in the detection of parenchymal hemorrhage in 

patients with acute ischemic stroke (AIS). 


Following a noncontrast CT obtained in the emergency 

room, 33 consecutive patients underwent MR imaging that 

includes an EPI gradient-echo sequence (single shot, 2D, 23 

slice, TR 2125, TE 30). Thirty-three consecutive patients 

presenting with hyperacute ischemic stroke were studied. 

Six patients had no follow-up imaging. Three patients had 

follow-up CT only. One patient had follow-up MR imaging 

only. Twenty-three patients completed follow-up CT and 

MR imaging between days two and three postadmission. 

One patient was excluded from further analysis as there was 

a primary hemorrhage detected on admission CT and MR 

imaging. Patients were scanned on a GE 1.5 T MR imaging 

system using a neurovascular head coil. One hundred and 

twenty-six studies were randomized by scan type (CT or MR 

imaging) and imaging session (admission or follow-up). All 

identifying information was removed from the images prior 

to review. All cases were reviewed separately by three neuroradiologists. 

Reviewers recorded a binary response - hemorrhage 

present or absent. The true presence of hemorrhage 

was determined by the final reports of the attending staff 

radiologists who had all imaging studies available for 

review. 

RESULTS 

Time between admission CT and MR imaging = 3.2 ± 3.9 

hours; and between follow-up CT and MR imaging = 4.9 ± 

7.5 hours. Time from stroke onset to initial CT was 4.2 ± 3.6 

hours and MR imaging 4.7 ± 2.9 hours. In 10 patients the 

exact timing between ictus and imaging was indeterminate. 

Number of hemorrhages seen on CT not detected on MR 

imaging = 0. Number of hemorrhages seen on MR imaging 

not detected on CT = 8 (see Table). All three reviewers had 

concordant responses in 90.2% of CT studies and in 89.2% 

of MR imaging studies indicating high interobserver reliability 

(kappa = 0.84 +/- 0.07). Five of 32 patients had evidence 

of hemorrhage on admission. In 3 of those patients,

the hemorrhage was seen on MR imaging only. The timing 

from ictus to MR imaging detection in these cases was 3 

hours, 4.8 hours, and indeterminate. Each was confirmed on 

follow-up imaging showing evidence of progression. 

Admission p

Monday 


Fiftey-one patients with acute ischemic stroke were studied 

by diffusion- and perfusion-weighted imaging within 3.0 ± 

0.8 hours, on day 1 and days 5-8. Patients were treated either 

conservatively (aspirin and low molecular weight heparin; n 

= 15), with early hemicraniectomy (n = 2), by IV (n = 32) or 

IA thrombolysis (n = 1). One patient was treated within a 

randomized, placebo controlled, double blinded acute stroke 

trial. After realignment of the image data sets, the parameter 

maps of the apparent diffusion coefficient (ADC), cerebral 

blood flow and blood volume (CBF, CBV), and mean transit 

time (MTT) were analyzed in the area of later HT. The 

degree of local diffusion and perfusion impairment in the HT 

area was compared with the entire diffusion and perfusion 

abnormality. Reperfusion status was separately assessed for 

the entire perfusion abnormality and the HT area. 

RESULTS 

Hemorrhagic transformation was observed in 19/51 patients 

(37.2 %) within 8 days after symptom onset. Areas destined 

to hemorrhagic transformation revealed a more severe 

decrease of ADC (to 70 ± 13%; p < 0.01), CBV (to 31 ± 

26%; p < 0.001) and CBF (to 28 ± 19%; p < 0.001) compared 

to the entire perfusion abnormality. Local reperfusion 

in the HT area was observed in 18/19 patients. We did not 

find differences in NIHSS on days 5-8 in patients who developed 

HT vs patients who did not. Patients showing HT were 

older (p = 0.001). 

CONCLUSION 

Our data support the hypothesis that hemorrhagic transformation 

in stroke patients is the result of local reperfusion 

within the region with the most severe perfusion impairment 

(Fig.). Thus, the hemorrhagic transformation per se will 

cause no additional harm to the patient since the tissue probably 

is infarcted in most of the cases. The presence of hemorrhagic 

transformation did not coincide with a worse clinical 

outcome. The prediction of hemorrhagic transformation 

without discrimination into space-occupying parenchymal 

haematoma and hemorrhagic infarction seems of limited 

clinical value. 

KEY WORDS: Stroke, hemorrhage, thrombolysis 

58 


MR Imaging Changes Brain Attack Therapy Plans 

Hsu, D. P. · Wang, G. · Schoenberg, A. · Zaidat, O. O. · 

Suarez, J. I. · Landis, D. M. D. · Lanzieri, C. F. · Tarr, R. W. 

· Sunshine, J. L. 


Cleveland, OH 

PURPOSE 

To define the benefit of ultrafast MR imaging in the selection 

of thrombolytic therapy for acute stroke patients. 


The multidisciplinary Brain Attack Team (BAT) consists of 

the neurology, neuroscience intensive care unit, neurosurgery 

and interventional neuroradiology which is alerted 

for patients with presumed acute neurologic deficits. These 

patients undergo neurologic evaluation, noncontrastenhanced 

head CT and when appropriate head MR imaging. 

The MR imaging consists of axial grase T2, diffusion, and 

bolus-tracking T2*-weighted images. Diffusion trace and 

time-to-peak perfusion maps are generated on the console. 

Quick 3D time-of-flight MRA can be performed of the 

intracranial and cervical vessels as well. One hundred and 

seventeen consecutive patients in which the BAT was called 

and who progressed through MR imaging were reviewed retrospectively. 

A total of 97 patients with full data available are 

included in the analysis. Data reviewed include date, time of 

symptom onset, time of institutional arrival, and time of MR 

imaging. An overall clinical impression was scored prior to 

MR imaging and a presumptive treatment plan generated. 

Clinical assessment included distinction of TIA from acute 

stroke, and between small and large vessel disease. The clinical 

impression and plan were compared to those formulated 

following input of MR imaging and the consistencies tabulated. 

RESULTS 

The age ranged from 32-92 years with a mean of 68, the 

female/male ratio was 52:45 and the most common risk factor 

was hypertension (63/97). The prevalence of previous TIA 

was 12% and previous stroke was 19%. The most encountered 

symptoms were weakness 91%, language deficit 49%, and 

altered sensorium 46%, with the clinical impression of stroke 

in all these. Large vessel etiology (LVD) was most likely clinical 

in 78%, a cardiac source for 29%, and small vessels 

(SVD) for 16%. Twelve percent showed CT evidence of acute 

ischemic change. Ninety percent (87/97) had MR imaging 

findings of acute cerebral infarct: LVD in 80% (78/97) with 

embolic origin for 99% of these, 10% 10/97 SVD. In 12% 

(12/97) there was discrepancy between the impression and 

plan formulated from clinical and CT data alone compared 

with the final treatment plan generated in combination with 

the MR imaging data. Ten of 12 were to receive therapy based 

on clinical evaluation but did not because of MR imaging 

findings: 6 had clinical LVD but had negative MR imaging, 2 

had small matched perfusion/diffusion-weighted imaging 

defects, 1 had a large matched perfusion/diffusion-weighted 

imaging defect, and 1 had small vessel disease. The other 2 

were changed from nontreatment or IV only plan to intrarterial 

therapy. These two presented with clinical impressions of 

SVD and then MR imaging disclosed LVD amenable to 

intraarterial thrombolysis.

CONCLUSION 

In this series of 117 consecutive patients, the impression of 

the condition was often consistent between clinical and MR 

imaging-based interpretation, however there were notable 

exceptions. Twelve (12%) had their therapy altered significantly. 

Ten patients who would have received thrombolysis 

based on initial assessment were found to have no MR 

abnormality or an abnormality that would not benefit from 

thrombolysis. Two (2%) of the patients had a treatment plan 

change to thrombolysis based on MR imaging findings. 

KEY WORDS: Stroke, perfusion-weighted imaging 


Acute Ischemic Stroke MR Imaging: Rapid Assessment 

of Anatomy, Hemorrhage, Penumbra, Major Vessels, and 

Early Defects in the Blood-Brain Barrier 

Mikulis, D. J. 1 · Kassner, A. 2 · Rowan, S. 1 · Silver, F. 1 


2University of Toronto, University Health Network, Toronto, 

ON, CANADA 

PURPOSE 

A comprehensive and time-efficient MR imaging protocol is 

reviewed for acute stroke management. This study aims to 

prove the feasibility of the protocol in a clinical setting. 


After an initial screening CT, 33 consecutive patients underwent 

an MR imaging outlined in Table 1. Imaging was performed 

on a 1.5 T GE Signa system using a neurovascular 

head coil. Three separate 15 cc injections of Gd-DTPA were 

required for sequential acquisition of permeability, perfusion, 

and contrast-enhanced (CE) MRA. Twenty-seven patients 

received all 3 CE sequences. Five patients received only two 

injections as the permeability sequence was omitted. Follow 

up with CT and MR imaging was performed at 48-72 hours. 

Images were assessed for the presence of parenchymal hemorrhage 

(blinded assessment by 3 neuroradiologists), and 

quality of CEMRA (subjective assessment by 1 neuroradiologist). 

Contrast-enhanced MRA quality was assessed only in 

patients receiving all 3 injections. Image quality was scored on 

a scale of 1 (poor) to 5 (excellent). Diagnostic confidence was 

rated as nondiagnostic, low, and high. 

RESULTS 

All patients successfully completed the admission MR imaging. 

Five hemorrhages were detected on the MR gradientecho 

sequence (later confirmed on follow-up imaging) with 

3 of 5 missed on the admission CT. No CT positive MR negative 

hemorrhages were observed. Diagnostic quality of 

CEMRA was rated as high in 82%, low in 7%, and nondiagnostic 

in 11% (MRA failed for technical reasons). Contrastenhanced 

MRA overall quality = 3.52 which is above average 

(3.0). Permeability was increased significantly within 

the region of restricted water diffusion that outlined the 

ischemic injury (reported in a separate study). 

Acute stroke protocol 

Series 1 2 3 4 5 6 7 8 

Sequence Localizer FSE T1 EPI EPI EPI GRE GRE EPI CEMRA - 

(ETL 3) Diffusion FLAIR (multi-shot) Permeability Perfusion Arch to 

(single-shot (single-shot) (31 phases) (single shot circle of 

EPI B = 1000) 25 phases) Willis 

Scan time 35 sec 35 sec 32 sec 39 sec 20 sec 4:46 min 44 sec 2:38 min 

59 

CONCLUSION 

This MR imaging protocol was able to rapidly assess hemorrhage, 

diffusion, perfusion, the great vessels from the aortic 

arch to the circle of Willis, and early defects in the bloodbrain 

barrier (BBB). The short acquisition time (10:49 minutes 

total for 8 sequences) and the need for 3 injections did 

not significantly compromise image quality. The MRA was 

judged slightly better than average despite the presence of 

preexisting Gd-DTPA. The information obtained was always 

diagnostic. The perfusion sequence may even benefit from 

previously injected Gd-DTPA where disruption of the BBB 

is present (1). The protocol was well tolerated and can be 

performed on a standard clinical MR imaging system with 

EPI capability. Finally, detecting early BBB defects eventually 

might assist in triaging patients if a link between these 

defects and hemorrhagic risk can be established. Since the 

clinical utility of permeability is under investigation, further 

time savings can be achieved by removing this acquisition 

from the protocol yielding a total acquisition time of only 

6:03 minutes. 

REFERENCES 

1. Kassner A, et al. JMRI 2000;11(2):103-113 

KEY WORDS: MR imaging, ischemic stroke, protocol 


A Novel Self-Expanding Nitinol Stent System for Stroke 

Prevention in Intracranial Atherosclerotic Disease: The 

Wingspan Trial 

Bose, A. 1 · Hartmann, M. 2 · Henkes, H. 3 · Ringleb, P. 2 · Sit, S. 4 

1 New York University School of Medicine, New York, 

NY, 2 University of Heidelberg Medical Center, 

Heidelberg, GERMANY, 3 Robert Janker Klinik 

Neuroradiology, Bonn, GERMANY, 4 Boston Scientific 

Corporation, San Leandro, CA 

PURPOSE 

Patients with high-grade, symptomatic intracranial (IC) atherosclerotic 

lesions who failed antithrombotic therapy are 

known to be at high risk of recurrent stroke but unfortunately 

have no other treatment options. Previous attempts to use balloon-expandable 

coronary stents (BES) to treat these patients 

yielded mixed results, largely due to rigid, inflexible designs 

and problems with access. The Boston Scientific self-expanding 

Wingspan TM Stent System (WST) is a highly flexible, selfexpanding 

(SE) stent system specifically designed for the 

treatment of IC atherosclerosis. The purpose of this prospective, 

multicenter, single-arm trial is to assess safety and performance 

of WST in patients with severe, symptomatic IC 

lesions who are refractory to medical therapy. 


Major entry criteria were recurrent stroke attributed to target 

lesions with stenosis > 50% and < 14 mm in length, failed 

medical therapy, and a Modified Rankin score (MRS) of < 3. 

Patients with acute, evolving stroke with thrombolytic treatments 

within 24 hours of entry were excluded. Intracranial 

lesions were predilated with the PTA balloon followed by 

WST placement for further remodeling. PTA balloons were 

undersized to no more than 80% of parent vessel diameter 

with slow inflation at nominal pressure (6 atm) to minimize 

vessel injury. Control angiogram was performed (WASID 

Monday

Monday 

method) to assess degree of stenosis at baseline, postprocedure 

and 6-month follow up. The primary endpoint is the 

composite rate of ipsilateral stroke/death at 30 days. 

Secondary endpoints are incidence of all stroke, composite 

ipsilateral stroke/death, angiographic evidence of vessel dissection, 

stent restenosis/migration, and all adverse events at 

6-month follow up. 

RESULTS 

A total of 45 patients were enrolled at 11 international centers. 

Patients were 51 to 80 years of age, 73% were male, and 

all had MRS of 3 or 2. Target vessel diameter ranged from 2 

- 4.8 mm with lengths from 1.25 to 14 mm; 51% of the 

lesions were in the anterior and 49% in the posterior circulation. 

Degree stenosis ranged from 55 to 99%. All 45 patients 

(100%) had a history of stroke at entry. The study is expected 

to complete in February 2005 when all 45 patients have 

reached their 6-month follow up. Complete results on clinical 

and angiographic outcomes at baseline, postprocedure, 

30-day and 6-month follow up will be presented. 

CONCLUSION 

In symptomatic patients with intracranial atherosclerotic disease 

who fail antithrombotic therapy, intracranial angioplasty 

and stenting may be a viable treatment alternative to BES. 

KEY WORDS: Intracranial atherosclerosis 


Are Intrathecal Steroids Administered at Extended Time 

Points Effective in Focal Cerebral Ischemia? A MR- 

Controlled Study in Rats 

Goericke, S. L. 1 · Engelhorn, T. 1 · Speck, U. 2 · Blechschmid, 

N. 1 · Becker, W. P. 1 · Forsting, M. 1 · Doerfler, A. 1 

1 Institute of Diagnostic and Interventional Radiology, 

Essen, GERMANY, 2 Humboldt University of Berlin, 

Berlin, GERMANY 

PURPOSE 

Our former study showed the intrathecally administered 

steroid triamcinolonacetonide (TCA) 30 minutes after 

ischemia to reduce infarction volume significantly. The aim 

of this MR-controlled study was to evaluate the neuroprotective 

efficacy of TCA administered intrathecally on infarction 

volume in acute focal cerebral ischemia in rats at different 

time points. 


Focal cerebral ischemia was induced in 75 Wistar rats using 

an endovascular occlusion technique of the middle cerebral 

artery (MCAO). Triamcinolonacetonide (0.012 mg/kg body 

weight) was administered into the cisterna magna of 15 rats 

each 0.5 hour, 1 hour, 2 hours, and 4 hours after MCAO. 

Fifteen animals received equivolumetric saline intrathecally 

30 minutes after MCAO. Infarction volume was calculated 

24 hours after MCAO by TTC-staining in all animals. 

RESULTS 

MR imaging confirmed successful MCAO and intrathecal 

administration in all animals. Compared to controls (18.03 ± 

2.79%), infarction volume was significantly reduced using 

TCA at a dose of 0.012 mg/kg 30 minutes after MCAO 

(12.86 ± 5.39%, p = 0.003), 1 hour after MCAO (9.68 ± 

60 

4.53%, p < 0.0001), 2 hours after MCAO (10.57 ± 6.21%, p 

= 0.004) and 4 hours after MCAO (9.05 ± 5.15%, p < 

0.0001) with corresponding infarction volumes in diffusionweighted 

MR images. 

CONCLUSION 

Intrathecally steroid triamcinolonacetonide administered 

0.5, 1, 2, and 4 hours after MCAO may significantly reduce 

infarction volume in permanent focal cerebral ischemia in 

rats. Further studies, focusing on long-term clinical outcome 

and combined therapies such as additional thrombolysis, are 

necessary to assess the therapeutic value. 

KEY WORDS: Cerebral infarction, glucocorticoids, neuroprotection 


CT Perfusion as an Indicator of Vasospasm Status 

Postsubarachnoid Hemorrhage: CT Angiography versus 

CT Perfusion 

Fountain, A. J. · Lekht, I. · Westman, D. A. · Baumgarten, D. A. 

Emory University 

Atlanta, GA 

PURPOSE 

Vasospasm is a major cause of morbidity and mortality in 

patients after subarachnoid hemorrhage (SAH) due to 

aneurysmal rupture. Vasospasm typically affects patient in 

the 4-10 days after SAH. Various techniques are available in 

the evaluation and early detection of vasospasm. Clinical 

examination along with transcranial doppler ultrasound 

(TCD) has been used to quantify the severity of vasospasm. 

The advent of CT angiography (CTA) enabled clinicians to 

visualize intracranial arteries around the circle of Willis and 

quantify the severity of vasospasm with better precision of 

the anatomical area involved. The combination of clinical 

examination, TCD, and CTA are used routinely in Neuro 

ICU patients for making clinical decisions about the treatment 

of patients status post SAH. It is assumed that patients 

who have significant vasospasm on CTA will go on to an 

infarction if left untreated. The purpose of this study is to 

correlate CT perfusion variables [mean transit time (MTT), 

cerebral blood flow (CBF), cerebral blood volume (CBV)] 

with CTA in patients after SAH due to aneurysmal rupture. 


We retrospectively reviewed 13 patients at various times 

after SAH with a clinical picture compatible with 

vasospasm. The patients were evaluated with CTA followed 

immediately by CT perfusion. Information on spasm was 

available in 169 vascular segments in these 13 patients. 

Correlative information on MTT, CBF, and CBV was available 

for all of these segments. A Chi-square test was utilized 

to evaluate the correlation between spasm on CTA and the 

perfusion variables. 

RESULTS 

Preliminary evaluation of the data demonstrates a significant 

correlation between vasospasm findings on CTA and MTT 

abnormalities on CT perfusion in corresponding vascular territories 

(p < .0001). No correlation was found between CTA 

and CBF or CBV.

CONCLUSION 

CT perfusion is used widely for evaluation of acute stoke. Its 

role in evaluation of vasospasm is unknown at this time; it is 

assumed that vasospasm identified on CTA produces measurable 

perfusion abnormalities in similar vascular territories. 

This study demonstrates that the process is more complex 

and while in some instances it may be true that significant 

vasospasm measured on CTA produces significant measurable 

abnormalities on CT perfusion, at the same time there are 

other findings at CT perfusion that do not correlate with 

findings at CTA. Regional vasodilation, gradual recruitment 

of nearby arteries to supply brain parenchyma affected by 

vasospasm, differences between the time of onset of 

parenchymal perfusion abnormalities, and vasospasm of 

major arteries around the circle of Willis or small patient 

population are some of potential explanations of the discordance 

of findings present in this study. 

KEY WORDS: CT perfusion, vasospasm, CT angiography 


Vasospasm after Subarachnoid Hemorrhage: Relevance 

of Perfusion CT and CT Angiography for Diagnosis and 

Management 

Wintermark, M. · Ko, N. U. · Liu, S. · Criqui, G. · Smith, W. 

S. · Dillon, W. P. 

University of California San Francisco 

San Francisco, CA 

PURPOSE 

Vasospasm after subarachnoid hemorrhage is classically 

evaluated by transcranial doppler sonography (TCD) and 

digital subtraction angiography (DSA). Because TCD is not 

completely satisfactory for the selection of patients for 

angiography, other noninvasive imaging modalities are 

investigated. The goal of this study was to evaluate the relevance 

of perfusion CT (PCT) and CT angiography (CTA) 

with respect to diagnosis and management of vasospasm. 


Twenty-seven patients with subarachnoid hemorrhage, 

admitted to our neurovascular unit and at risk of vasospasm, 

were identified retrospectively as having had PCT CTA and 

TCD within a time interval of 12 hours from each other, 

and/or PCT CTA and DSA within a time interval of 12 hours 

from each other. The patients’ charts were reviewed for the 

diagnosis of vasospasm on the basis of TCD and/or DSA, 

and for the kind of treatment they underwent (triple H, IV 

nimodipine, IA verapamil, angioplasty). In each patient, 

CTA, TCD, and DSA examinations were reviewed independently 

for vasospasm in 18 corresponding vascular territories. 

PCT values were measured in the same territories. 

RESULTS 

A total of 40 PCT CTA, 38 TCD and 35 DSA examinations 

were obtained in our 27 patients. Seven patients were found to 

have no vasospasm. Eighteen patients were suspected of 

vasospasm on TCD. Only 11 were confirmed by DSA. Six 

patients underwent endovascular therapy using either IA verapamil 

or angiography. A total of 123 territories among 630 

were identified as showing vasospasm by DSA (gold standard). 

CTA and PCT-derived mean transit time (MTT) represented 

the best combination for the diagnosis of vasospasm (positive 

61 

predictive value = 97%, negative predictive value = 91%), and 

was better than CTA alone or MTT alone. A rCBF value < = 40 

[cc x 100g-1 x min-1] represented the best indicator for 

endovascular therapy, with a negative predictive value of 95%. 

CONCLUSION 

A CT survey with PCT and CTA represents a reliable screening 

test in patients with suspected vasospasm, and can be 

used as a tool in the selection of the best strategy for the 

management of these patients. 

KEY WORDS: Vasospasm, perfusion CT, CT angiography 


Role of CT Angiography in Detection of Vasospasm following 

Aneurysmal Subarachnoid Hemorrhage 

Fassihi, A. A. · Kim, A. K. Y. · Giannotta, S. L. · Larsen, D. 

W. · Teitelbaum, G. P. · Kim, P. E. 

Keck School of Medicine, University of Southern California 

Los Angeles 


PURPOSE 

We studied the role of CT angiography (CTA) to determine 

the suitability of patients for angioplasty to treat vasospasm 

following aneurysmal SAH between March and August 

2004. 


Fourteen patients presented on average post-bleed day 2 and 

underwent admission brain CTA. Average Fisher and Hunt- 

Hess Grades were both 3. Twenty-one aneurysms were identified 

(43% MCA; 33% ACOM). All patients underwent coil 

embolization (n = 4) or clip-ligation (n = 10). Eight patients 

developed clinical signs of vasospasm and underwent a total 

of 12 CTAs which were compared with admission CTAs. 

RESULTS 

Transcranial doppler results compared favorably to clinical 

exam and CTA results for detection of vasospasm. On 10 of 

12 CTA scans, patients’ vasospasm did not justify emergency 

angioplasty due to the distal location or degree of 

vasospasm. In two cases, moderate to severe vasospasm was 

determined by CTA in vessels amenable to angioplasty and 

patients underwent successful balloon dilation. Follow-up 

CTA postangioplasty revealed mild vasopasm recurrence but 

sustained patency of angioplastied vessels 1 week later. CT 

angiography facilitated the time from detection of clinical 

vasospasm to determining the degree and location of 

vasospasm and obviated the need for stand-by neurointerventional 

and anesthesia teams in case angioplasty proves 

necessary. After dilatation, angioplasty sites can be followed 

by CTA for restenosis. Average nonionic-contrast use for 

CTA was 119 ml compared to 90 ml for angiography. 

CONCLUSION 

Analysis of the location and degree of vasospasm by CTA 

allows for rapid triage of vasospasm patients to either medical 

or endovascular management while often avoiding the 

risks of diagnostic neuroangiography. 

KEY WORDS: CT angiography, vasospasm, subarachnoid 

hemorrhage 

Monday

Monday 


Multidetector CT Angiography in the Evaluation of 

Patients with Neurovascular Stents: Correlation with 

Conventional Angiography 

Stockmann, T. M. 1 · Villablanca, J.P. 2 · Jahan, R. 2 · 

Duckwiler, G. R. 2 · Sayre, J. 2 

1 2 Hospital of St. Raphael, New Haven, CT, University of 

California Los Angeles Center for the Health Sciences, Los 

Angeles, CA 

PURPOSE 

To evaluate the efficacy and accuracy of multidetector helical 

CT angiography (CTA) in the evaluation of patients neurovascular 

stenosis postendovascular stenting with correlation 

to catheter angiography. 


Both CTA and DS angiography (DSA) were performed poststenting, 

prospectively evaluating the following: vessel 

stented, diameter of patent artery, length and diameter of 

stent, residual/recurrent stenosis, intimal hyperplasia, plaque 

burden, and presence of procedural complications such as 

stent thrombosis and arterial dissection. CT angiographic 

images were evaluated using orthogonal or curved oblique 

2D multiplanar reformatted (MPR) images. CT angiographic 

measures used internal digital caliper with predefined 

accuracy, using a full-width at half-maximum of the slice 

sensitivity profile orthogonal to the long axis of the artery. 

DS angiography measurements were made using a digital 

caliper with guide catheter or supraclinoid internal carotid 

artery standard diameter as reference. All CTA measurements 

were made with the readers blinded to the DSA 

results. 

RESULTS 

Thirty-five stents (6 types) were deployed into 29 arteries: 

16 cervical ICA, 8 CCA, 2 petrous ICA, 2 vertebral, 1 subclavian. 

Demographics: M/F 12/7, mean age 66 years (range 

46-86). Average stented vessel diameter was 4.5 mm (range 

2.8-7.5 mm). Stent length and diameters agreed closely 

between multidetector CTA and DSA, with an average of 

10% difference for length and 18.7% for diameter, with a 

tendency to multidetector CTA underestimation. 

Concordance for patency was 100%. Sensitivity of multidetector 

CTA for residual or recurrent stenosis was 1.0, specificity 

0.79. Residual/recurrent stenosis rates were 35% for 

DSA (range 10-65% severity) and 52% for multidetector 

CTA (range 10-71% severity) with an average 12% difference 

between DSA and multidetector CTA stenosis rates; 

generally multidetector CTA overestimates. Intimal hyperplasia 

was detected by multidetector CTA in 29% and by 

DSA in 9%. Intrastent atherosclerotic plaque was noted in 

37% of stented cases by DSA and in 23% by CTA. 

Complications were seen in 16% of patients by both multidetector 

CTA and DSA, except for 1 case of a cavernous 

carotid fistula missed on multidetector CTA, and one case of 

nonocclusive eccentric intraluminal thrombus missed by 

DSA. 

CONCLUSION 

Multidetector CTA is a viable noninvasive technique for the 

evaluation of patients following neurovascular stenting. 

Multidetector CTA has a high sensitivity and moderately 

62 

high specificity for the detection of intrastent residual or 

recurrent stenosis as compared to DSA. There was a high 

concordance between multidetector CTA and DSA for intimal 

hyperplasia, intrastent atherosclerosis, and complication. 

KEY WORDS: Multidetector CT angiography, neurovascular 

stents, conventional angiography 


4:40 PM - 6:10 PM 

Room 105/106 

(18) Orbital Imaging (ASHNR) 

(124) Proptosis 

Proptosis 

Deborah R. Shatzkes, MD 

— Deborah R. Shatzkes, MD 

(125) Ocular and Orbital Tumors 


(126) Orbital and Ocular Trauma 

— Alisa D. Gean, MD 

Moderators: Deborah R. Shatzkes, MD 




able to: 

1) Describe the role of the radiologist in the clinical setting 

of proptosis. 

2) Differentiate the various imaging patterns found in the 

proptotic patient. 

3) Enumerate the most common lesions within each of these 

patterns. 

4) Formulate a limited and useful differential diagnosis 

based on the imaging pattern and available clinical information. 


The role of the radiologist in the clinical setting of proptosis 

is not to make the diagnosis, but to search for the cause. 

Knowledge of clinical information including (at minimum) 

the patient’s age, the duration of symptomatology and the 

presence or absence of pain will aid greatly in narrowing the 

very broad differential diagnosis. Utilizing a pattern 

approach, this differential can be divided broadly into three 

categories: alteration in bony orbital contour, enlargement of

structures that belong in the orbit, and presence of structures 

that don’t belong in the orbit. Within the first category, alteration 

in bony orbital contour, the most common entities are 

expansile osseous lesions such as metastases and meningioma, 

as well as expansile processes within the paranasal 

sinuses, such as mucocele and polyposis. Within the second, 

enlargement of structures that belong in the orbit, thyroid 

orbitopathy accounts for the majority of cases, with additional 

vascular and primary ocular causes. The third category, 

presence of structures that don’t belong in the orbit, 

includes a very wide differential of neoplastic, infectious, 

inflammatory, congenital and traumatic lesions, many of 

which are covered in the subsequent lectures of this session. 

By combining clinical information with lesion categorization, 

as notes previously, the radiologist can produce a limited 

and clinically relevant differential diagnosis. 

Ocular and Orbital Tumors 


Dr. H. Christian Davidson is an Associate Professor at the 

University of Utah, where he also serves as Executive Vice 

Chairman in the Department of Radiology. Dr. Davidson 

trained as a resident in radiology at Stanford University, and 

as a fellow in neuroradiology at the University of Utah. Dr. 

Davidson has special interest in head and neck imaging, and 

is author of many original publications, as well as co-author 

of the definitive reference text "Diagnostic Imaging: Head 

and Neck." 



able to: 

1) Design a strategy for the imaging approach to orbital and 

ocular lesions using MR imaging, CT, and US. 

2) Understand spatial- and pathology-based classifications 

of orbital and ocular lesions. 

3) Describe the clinical and imaging features of common 

orbital and ocular lesions. 


The imaging approach to orbital and ocular masses will be 

discussed. Clinical indications and protocol strategies for 

MR imaging, CT, and US scanning will be presented, including 

the relative strengths and weakness of these modalities. 

The spatial approach to classifying orbital and ocular masses 

will be discussed, as well as pathologic categorization that 

provides a coherent organization to the two dozen or so most 

common lesions that occur in the orbit. Individual lesions 

will be discussed, including relevant clinical and pathologic 

features, as well as imaging features with many examples of 

typical and variant presentations. 

REFERENCES 

1. Davidson HC. The Orbit. In: Harnsberger HR (ed): Diagnostic 

Imaging: Head and Neck. September 2004. AMIRSYS, Salt 

Lake City, UT. 

2. Harnsberger HR, Wiggins RH, Davidson HC (eds). Pocket 

Radiologist: Head and Neck Top 100 Diagnoses. November 

2001. AMIRSYS, Salt Lake City, UT. 

63 

Orbital and Ocular Trauma 




able to: 

1) Review normal ocular and orbital anatomy. 

2) Discuss the imaging options for evaluating trauma to the 

orbit. 

3) Illustrate the imaging manifestations of the various 

injuries. 

4) Discuss the clinical relevance and treatment options for 

the injuries. 


The seminar will begin with a review of normal ocular and 

orbital anatomy. This will be followed by a discussion of the 

imaging approach and the imaging manifestations of orbital 

trauma. Particular emphasis will be given to the clinical relevance 

of the imaging findings as well as to potential treatment 

options. The injuries to be discussed will include globe 

rupture, ocular FB, lens dislocation, orbital “tenting," orbital 

hematocyst, Terson’s syndrome, CCF, chiasm shearing, and 

multiple fractures that involve the orbit (e.g., LeFort II/III, 

zygomatic-complex, optic canal, medial/lateral blow-out). 


4:45 PM - 6:15 PM 

Theatre 

(19) Advanced Imaging Seminar - 

Diffusion 

(127) Diffusion and Stroke 

— Howard A. Rowley, MD 

(128) Diffusion Tensor Imaging 

— Aaron S. Field, MD, PhD 

(129) Diffusion - Parallel and PROPELLER 

— Marshall S. Sussman, PhD 

Panel Discussion 

Moderators: Howard A. Rowley, MD 


Monday

Monday 

Diffusion and Stroke 




able to: 

1) Become familiar with the temporal evolution and differential 

considerations of diffusion-weighted imaging in 

stroke. 

2) Review trial data using perfusion-weighted imaging-diffusion-weighted 

imaging mismatch to select patients for 

acute stroke treatment beyond 3 hours. 


Diffusion-weighted imaging (DWI) has become a critical 

component in the evaluation of stroke and TIA. Diffusion 

sequences are rich data sets containing the native echo-planar 

T2 (b = 0), 3-direction averaged DWI (trace), and apparent 

diffusion coefficient (ADC) maps. Diffusion-weighted 

imaging starts to become positive within minutes and may 

remain bright for about a month, while ADC typically normalizes 

at about 10 days. This temporal evolution allows distinction 

of acute vs subacute infarction, and also helps identify 

reperfusion. Diffusion-weighted imaging is not only the 

most sensitive way to detect stroke but is also positive in 

many TIA patients. While sensitive for ischemia, DWI is not 

specific for it: differential considerations include infection, 

highly cellular tumors, blood clots, and other conditions. 

Clinical trials and off-label stroke protocols have begun to 

use perfusion-diffusion mismatch as a standard method to 

select patients for thrombolysis or mechanical clot retrieval 

beyond the usual 3-hour treatment window. The recently 

published DIAS trial showed positive clinical outcomes, low 

bleeding rates, and excellent recanalization using IV 

desmoteplase in a 3-9 hour time window with at least a 20% 

PWI-DWI mismatch (Hacke et al., Stroke 2005). Several 

additional trials are currently in progress to test the validity 

of the PWI-DWI mismatch hypothesis. 

Diffusion Tensor Imaging 

Aaron S. Field, MD, PhD 



able to: 

1) Review principles of diffusion tensor imaging (DTI). 

2) Summarize common methods of DTI acquisition and 

postprocessing. 

3) Review anatomy of major cerebral white matter (WM) 

tracts. 

4) Recognize manifestations of WM disease on DTI. 


Diffusion tensor imaging (DTI) has enabled unprecedented 

visualization of white matter (WM) pathways in the CNS 

and is currently making its way into clinical practice. This 

presentation will review the principles underlying DTI, with 

an emphasis on concepts over rigorous mathematics. 

Common approaches to DTI acquisition and postprocessing 

will be described, with an emphasis on echoplanar imaging 

(EPI) and the mapping of diffusion anisotropy and fiber tract 

64 

orientation. Alternatives to EPI and advanced postprocessing 

will be covered briefly but the greater emphasis will be on 

methods currently or soon to be widely available. The anatomy 

of major cerebral WM fiber tracts readily depicted by 

DTI will be reviewed. Finally, common patterns of diseasealtered 

WM pathways and pitfalls encountered in clinical 

application will be discussed, with an emphasis on tumor 

imaging. 

REFERENCES 

1. Dong Q, Welsh RC, Chenevert TL, et al. Clinical applications 

of diffusion tensor imaging. J Magn Reson Imag 2004;19:6-18 

2. Jellison BJ, Field AS, Medow J, et al. Diffusion tensor imaging 

of cerebral white matter: A pictorial review of physics, 

fiber tract anatomy, and tumor imaging patterns. AJNR Am 

J Neuroradiol 2004;25:356-369 

3. Field AS, Alexander AL. Diffusion tensor imaging in cerebral 

tumor diagnosis and therapy. Top Magn Reson Imag 

2004;15:315-324 

Diffusion - Parallel and PROPELLER 

Marshall S. Sussman, PhD 

Dr. Marshall S. Sussman currently holds an Assistant 

Professor appointment in the Department of Medical 

Imaging at the University of Toronto, as well as an Associate 

Researcher appointment at the University Health Network 

(UHN). He received his PhD from the Department of 

Medical Biophysics at the University of Toronto in the field 

of MR imaging. His research interests include applications 

of novel diffusion-weighted imaging methods, the development 

of motion compensation strategies for MR imaging, 

and real-time MR imaging techniques. He has published or 

submitted 12 papers, 28 abstracts, and has 7 patents issued 

or pending. He was a finalist in the Young Investigator 

Competition of the 2002 International Society of Magnetic 

Resonance in Medicine. 



able to: 

1) Recognize the limitations of conventional diffusionweighted 

MR imaging techniques. 

2) Illustrate diffusion-weighted parallel imaging techniques, 

and to appreciate their benefits. 

3) Illustrate the diffusion-weighted PROPELLER technique, 

and to appreciate its benefits. 


Diffusion-weighted (DW) MR imaging provides valuable 

diagnostic information in a variety of neurologic applications, 

including stroke, white matter degeneration and development, 

and brain tumors. Currently, most clinical DW MR 

scans are performed using the single-shot EPI pulse 

sequence. This pulse sequence acquires the entire DW image 

in a single data-acquisition period lasting about 100 ms. The 

advantage of single-shot acquisition is that motion artifacts, 

which are a major impediment to achieving high-quality DW 

images, are minimized. Unfortunately, the long 100 ms dataacquisition 

period makes this pulse sequence particularly 

sensitive to image artifacts such as geometric warping, signal 

dropout, and blur. Such problems are particularly acute in

anatomical regions which contain large magnetic susceptibility 

differences; such as the sinus cavities. Recently, two 

new DW imaging techniques, parallel imaging and PRO- 

PELLER, have been developed that have the potential for 

overcoming these difficulties. Parallel imaging techniques 

[e.g., SMASH (1) and SENSE (2)] allow one to reduce the 

overall data-acquisition time. When used in conjunction with 

single-shot EPI, this has the effect of reducing sensitivity to 

the aforementioned image artifacts. As an added benefit, by 

reducing the echo time (TE), the T2 “shine-through” effect, 

which often confounds image interpretation, also is reduced 

with parallel imaging. Diffusion-weighted PROPELLER (3) 

is a multishot fast spin-echo pulse sequence. Unlike conventional 

pulse sequences, it uses a novel circular k-space trajectory. 

The major advantage of this pulse sequence over 

conventional single-shot EPI is that it is much less sensitive 

to the types of image artifacts mentioned previously, due to 

the fact that it is spin-echo based. On the other hand, because 

DW PROPELLER is a multishot sequence, it is more susceptible 

to motion artifacts. However, because of DW 

PROPELLER’s unique k-space trajectory, its overall sensitivity 

to motion is less severe than conventional multishot 

pulse sequences. In this presentation, an overview of the previous 

concepts underlying conventional single-shot EPI, parallel 

imaging, and PROPELLER DW imaging will be discussed. 

A comparison highlighting the relative advantages 

and drawbacks of each pulse sequence will be performed. 

REFERENCES 

1. Sodickson DK, et al. Magn Reson Med 1997;38:591-603 

2. Pruessmann K, et al. Magn Reson Med 1999;42:952-962 

3. Pipe JG. Magn Reson Med 2002;47:42-52 


65 

Monday

Monday 

Notes:

Tuesday Morning 

8:00 AM - 12:00 PM 

Room 201B 

(20) ASNR Research Grant Writing 

Seminar: Practical Tips on Getting 

Your Grant Funded - Part II 


8:00 AM - 9:00 AM 

Room 205 

— Janet S. Rasey, PhD 

(21) ELC Lecture D: Multimedia 

Conference Room 2005 

— Venkata Natarajan, PhD 

67 






8:00 AM - 9:30 AM 

Room 105/106 

(22) Intracranial Atherosclerotic 

Disease (ASITN) 

(132) Stroke Risk and Medical Treatment of 

Intracranial Arterial Stenosis 

— Helmi Lutsep, MD 

(133) Surgical Treatment of Intracranial Arterial 

Stenosis 

— Colin P. Derdeyn, MD 

(134) Angioplasty for Intracranial Arterial Stenosis 

— Michael P. Marks, MD 

(135) Stenting for Intracranial Arterial Stenosis 

— John D. Barr, MD 


Moderator: John D. Barr, MD 

Stroke Risk and Medical Treatment of Intracranial 

Arterial Stenosis 


Dr. Lutsep is an Associate Professor in the Department of 

Neurology at Oregon Health & Science University. Dr. 

Lutsep completed her medical school and residency at the 

Mayo Clinic and completed a behavioral neurology fellowship 

at the University of California at Davis as well as a fellowship 

in cerebrovascular disease at Stanford University. 

She is a member of the stroke team and is the Associate 

Director of the Oregon Stroke Center. Dr. Lutsep’s research 

interests involve the development of therapies for stroke prevention, 

acute treatment and stroke recovery, with a focus on 

the use of devices in clinical stroke treatment. She also studies 

stroke imaging and the cognitive effects of focal brain 

lesions. Dr. Lutsep is the author of numerous publications 

Tuesday

Tuesday 

and presentations. She serves on the executive board of the 

Western States Stroke Consortium and is a chief editor of the 

eMedicine online neurology textbook. 



able to: 

1) Be able to review stroke rates associated with intracranial 

stenosis and identify characteristics that may increase this 

risk. 

2) Be able to discuss the efficacy of aspirin compared to warfarin 

treatment in the randomized Warfarin Aspirin 

Symptomatic Intracranial Disease (WASID) trial. 


Symptomatic intracranial stenoses are present in 8-10% of 

patients. Higher rates are seen in Chinese, African American, 

and Hispanic populations compared with Caucasians and, in 

the anterior circulation, in patients with cortical symptoms or 

signs. Two prospective trials provided data of stroke risk in 

patients with intracranial stenosis on medical therapy. The 

first was the medical treatment arm (primarily aspirin) of the 

extracranial-to-intracranial bypass trial published in 1985. In 

131 patients with 70% or more middle cerebral or internal 

carotid artery stenosis inaccessible to endarterectomy, stroke 

rates were 7-10% per year over an average follow up of 4.65 

years. Patients with a symptomatic 50% or more stenosis of 

the intracranial carotid, middle cerebral, basilar or vertebral 

artery were included in the Warfarin Aspirin Symptomatic 

Intracranial Disease (WASID) trial presented in 2004. This 

prospective trial showed high recurrent stroke rates of 11- 

12% in 1 year in both the aspirin and warfarin arms. Trial 

data for characteristics that might carry higher stroke risk, 

such as higher degrees of stenosis, are forthcoming. 

Although a retrospective study had suggested that warfarin 

could reduce stroke rates due to symptomatic intracranial 

stenosis of 50% or more compared to aspirin, the prospective, 

multicenter WASID trial showed that warfarin (INR 2- 

3) had no more efficacy than aspirin (1300 mg/day) and was 

associated with increased rates of major hemorrhage: 8% in 

the warfarin group and 3% in the aspirin group. Because of 

the hemorrhage risks and monitoring requirements of warfarin, 

antiplatelet agents thus arerecommended, although 

only aspirin has been assessed in this patient population. The 

high rates of stroke on medical therapy suggest that it is not 

sufficient for the prevention of stroke due to intracranial 

stenosis. In the first prospective intracranial (and extracranial 

vertebral) stent trial, Stenting of Symptomatic 

Atherosclerotic Lesions in the Vertebral or Intracranial 

Arteries (SSYLVIA), strokes occurred in 4/61 (6.6%) of 

patients in the first 30 days. While another 4/55 (7.3%) of 

patients had strokes later than 30 days, all occurred in those 

with recurrent stenosis of at least 50%, one of which was in 

the only patient not stented. This trial used a bare stent 

(Neurolink, Guidant Corporation), which carries a potentially 

higher rate of restenosis and recurrent stroke than a treated 

stent. 

68 

Surgical Treatment of Intracranial Arterial Stenosis 




able to: 

1) Identify the three primary surgical options for revascularization 

in patients with atherosclerotic intracranial stenosis. 

2) Analyze the advantages and disadvantages of each procedure, 

relative to medical therapy, in this patient population. 


There are three primary options for surgical revascularization 

for intracranial occlusive disease (stenosis and occlusion): 

Endarterectomy, direct arterial bypass, and indirect 

arterial bypass. We will review these three procedures and 

the evidence supporting their use in different patient populations. 

Endarterectomy of intracranial vessels is technically 

challenging but feasible. This technique has been reported in 

case reports. Direct arterial-to-arterial bypass, either superficial 

temporal artery to middle cerebral artery or other higher 

flow procedures, has received much more study. In the 

EC/IC bypass trial, patients with atherosclerotic intracranial 

artery stenosis who underwent surgical bypass had a higher 

risk of stroke than patients treated medically (1). The application 

of this procedure for patients with complete carotid 

artery occlusion currently is being investigated in a randomized 

trial (the Carotid Occlusion Surgery Study) (2). Finally, 

indirect revascularization procedures, such as encephalodural 

synangiosis (EDAS), have been used with some success in 

patients with intracranial stenosis or occlusion associated 

with moyamoya phenomena (3). Whether the EDAS procedure 

may be of benefit in patients with atherosclerotic occlusion 

remains undetermined. The applications of both these 

surgical techniques commonly involve patient selection 

based on hemodynamic evaluation, and these methods will 

be reviewed as well (4). 

REFERENCES 

1. The EC/IC Bypass Study Group. Failure of extracranialintracranial 

arterial bypass to reduce the risk of ischemic 

stroke: Results of an international randomized trial. N Engl 

J Med 1985;313:1191-2000 

2. Grubb RL Jr, Powers WJ, Derdeyn CP, Adams HP Jr, Clark 

WR. The carotid occlusion surgery study. Neurosurgical 

Focus 2003;14:1-9 

3. Yilmaz EY, Pritz MB, Bruno A, Lopez-Yunez A, Biller J. 

Moyamoya. Indiana University Medical Center Experience. 

Arch Neurol 2001:58;1274-1278 

4. Derdeyn CP, Grubb RL, Jr, Powers WJ. Cerebral hemodynamic 

impairment: Methods of measurement and association 

with stroke risk. Neurology 1999;53:251-259

Angioplasty for Intracranial Arterial Stenosis 




able to: 

1) Describe the risks and benefits of angioplasty in the management 

of symptomatic intracranial atherosclerosis. 

2) Identify those patients who may possibly benefit from this 

procedure. 


A significant number of patients with symptomatic intracranial 

stenoses suffer recurrent symptoms (strokes and/or transient 

ischemic attacks) despite medical therapy with 

antithrombotic drugs. Percutaneous transluminal angioplasty 

has been used to treat this group of patients. This lecture will 

review the procedural success rates and complication rates of 

angioplasty for this disease. It also will discuss the available 

mid-term and long-term clinical follow up when symptomatic 

patients are treated with this procedure. These results 

will be compared to available natural history data in order to 

understand the potential impact of angioplasty on recurrent 

stroke rates in this group of patients. There will be a discussion 

of methodology for patient selection, and also the 

angioplasty procedure. In addition selected cases will be 

used to illustrate treatment decisions as well as limitations of 

the procedure. There also will be a brief discussion of future 

directions for endovascular therapy in the treatment of 

patients with symptomatic intracranial atherosclerosis. 

Stenting for Intracranial Arterial Stenosis 




able to: 

1) Review patient selection criteria for intracranial stenting. 

2) Review published reports of intracranial stenting. 

3) Review available devices for intracranial stenting. 

4) Review devices in development for intracranial stenting. 


The current status of stent-assisted angioplasty for treatment 

of intracranial atherosclerotic stenosis will be reviewed and 

appropriate patient selection criteria will be discussed. 

Angioplasty alone of intracranial arteries has proven to be 

less durable and efficacious than desired. Stent-assisted 

angioplasty of peripheral and coronary arteries has proven 

superior to angioplasty alone. The use of stent-assisted 

angioplasty for intracranial arteries has been restricted by the 

physical limitations of available stents and delivery systems. 

Technical improvements in coronary stent systems have 

allowed for progressively safer use and more distal placement 

in the neurovasculature. However, there are still many 

intracranial lesions that cannot be treated safely using coronary 

stents. A balloon expandable stent system designed 

specifically for the neurovasculature was developed and 

approved by the FDA in 2002 (1). Unfortunately, economic 

considerations lead the manufacturer to cease distribution of 

this device. Another stent system designed for treatment of 

69 

intracranial atherosclerotic stenosis has been developed, 

with FDA approval pending at this time. At present, there is 

no available FDA-approved stent designed for neurovascular 

atherosclerosis. 

REFERENCES 

1. SSYLVIA Study Investigators. Stenting of symptomatic atherosclerotic 

lesions in the vertebral or intracranial arteries 

(SSYLVIA): study results. Stroke 2004;35(6):1388-1392 



8:00 AM - 9:30 AM 

Theatre 

(23) Facial Imaging (ASHNR) 

(136) Mandibular Imaging 

— James J. Abrahams, MD 

(137) Parotid and Submandibular Gland Pathology 

— Edward E. Kassel, MD 

(138) Buccal, Masticator, and Parapharyngeal 

Spaces 

— Deborah L. Reede, MD 

Moderators: Edward E. Kassel, MD 


Mandibular Imaging 




able to: 

1) Explain imaging of the jaw, using dental CT software programs 

(i.e., DentaScan). 

2) Discuss dental implants. 

3) Discuss imaging of dental infection. 

4) Illustrate the various types of lesions and abnormalities of 

the jaw, which are seen using this imaging technique. 


The following talk will focus on four areas: 1. Imaging of the 

jaw, using dental CT software programs (i.e., DentaScan). 

Tuesday

Tuesday 

2. Dental implants. 3. Imaging of dental infection. 4. Various 

types of jaw lesions and abnormalities, seen using this imaging 

technique. What are dental CT software programs and 

why were they developed? Basically, these are reformatting 

programs that use the data from thin 1 mm axial images to 

reformat multiple cross-sectional images along the curve of 

the maxilla or mandible and multiple reformatted panoramic 

views. The reformatted images avoid dental streak artifact, 

which typically is seen on direct coronal images. These 

images provide an outstanding view of the mandible and 

maxilla and allow one to “unravel” the curve of the jaw. 

These images are designed for accentuating detail of the 

bone, and soft tissue is best evaluated with standard CT or 

MR imaging. Why were these programs developed? These 

programs were developed originally to evaluate patients 

prior to undergoing surgery for dental implants. Dental 

implants are metallic cylinders that are surgically imbedded 

in the mandible and maxilla to support a dental prosthesis 

(fixed denture). This technique allows patients who are 

edentulous (missing teeth) to have a more natural prosthesis 

that is fixed rather than the standard removable dentures. 

This provides much better function and improves self-image. 

Unfortunately, the amount of bone in the mandibular and 

maxillary alveolar ridge (portion supporting teeth) varies 

and therefore dentists and oral surgeons need this imaging 

technique to determine if sufficient bone is available to allow 

dental implantation and to avoid violating the mandibular 

canal (neurovascular canal) and maxillary sinuses. Why does 

the amount of bone vary? Basically, if one is missing their 

teeth, there is no longer vertical stress on the bone and disuse 

atrophy occurs with gradual resorption of the bone. The 

CT dental programs were developed therefore for evaluating 

patients planning implant surgery. Since then many other 

uses for them have been described. They have been instrumental 

in evaluating lesions and tumors of the jaw by providing 

the precise relationship of the lesion to the neural vascular 

bundle (inferior alveolar canal and incisive foramen) 

and to the maxillary sinuses and root apicies. Patients with 

dental infections have been evaluated as well as ones with 

genioplasties using osteotomies or silastic implants. Other 

uses are foreign body localization, determining root resorption, 

evaluating oral cancer patients for bone invasion, etc. 

These all will be illustrated. 

REFERENCES 

1. Abrahams JJ. Dental CT imaging: A look at the jaw. 

Radiology 2001;219:334-345 

2. Abrahams JJ. Anatomy of the jaw revisited with a dental CT 

software program: Pictorial essay. AJNR Am J Neuroradiol 

1993;14:979-990 

3. Abrahams JJ. The role of diagnostic imaging in dental 

implantology. Radiol Clin North Am 1993;31(1):163-180 

4. Abrahams JJ. Dental implants and multiplanar imaging of 

the jaw. In: Som P, ed. Imaging of the Head and Neck. St. 

Louis, Mosby 1995;350-374 

5. Abrahams JJ. Dental implants and related pathology. In: Som 

P, ed. Imaging of the Head and Neck. St. Louis, Mosby 

2003:919-929 

70 

Parotid and Submandibular Gland Pathology 




able to: 

1) Differentiate benign from malignant appearing salivary 

neoplasms. 

2) Identify the more common parotid (solitary or multiple) 

solid or cystic lesions. 

3) Develop a list of specific imaging features to be included 

in their imaging reports as pertain to parotid masses. 


A variety of disease processes may involve salivary gland 

tissues including developmental, inflammatory/infectious, 

obstructive, systemic (autoimmune) traumatic or neoplastic. 

This presentation will emphasize neoplastic lesions as focal 

mass lesions within the salivary tissues. The more common 

benign and malignant salivary lesions will be presented, 

emphasizing limitations in differentiating benign from lowgrade 

malignant neoplasm. Benign and malignant nonepithelial 

salivary tumors (neurogenic, lymphoma, sarcoma) 

occur much less frequently in the adult population and may 

need to be differentiated from primary salivary tumors as 

may solitary metastatic lesions. Multiple parotid mass 

lesions offer a different likelihood of pathologic entities 

from that of solitary mass lesions or that of cystic salivary 

lesions. Occasional nonneoplastic lesions (obstructive, 

inflammatory, autoimmune - BLEL, Sjogren’s, ARPC) will 

be presented as entities also to be considered in the diagnostic 

consideration of salivary mass lesions. 

REFERENCES 

1. Som PM, Brandwein MS. Salivary glands: Anatomy and 

pathology. In: Som PM, Curtin HD, eds. Head and Neck 

Imaging, 4th edition. St. Louis, Mosby 2003;39:2005-2133 

2. Eisele DW, Johns ME. Salivary gland neoplasms. In: Bailey 

BJ, ed. Head and Neck Surgery-Otolaryngology, 3rd edition. 

Philadelphia, Lippincott-Williams and Williams 

2001;2(107):1279-1297 

3. Alvi A, Myers EN, Carrau RL. Malignant tumors of the salivary 

glands. In: Myers EN, Suen JY, eds. Cancer of the Head 

and Neck, 3rd edition. Philadelphia, W.B. Saunders 

1996;26:525-561 

Buccal, Masticator, and Parapharyngeal Spaces 

Deborah L. Reede, MD 


Upon completion of this presentation. participants will be 

able to: 

1) Learn the anatomy of the buccal, masticator, and parapharyngeal 

spaces. 

2) Review the imaging characteristics of lesions in these 

spaces. 

3) Learn clinical findings that may help in the diagnosis of 

specific lesions. 

4) Learn key anatomical landmarks that are important for 

treatment planning.


Buccal Space 

Located superficial to the buccinator muscle. 

Boundaries of the Buccal Space 

Medially Buccinator Muscle 

Anterior and Lateral Lesser and Greater Zygomatic Muscle, 

Risorous Muscle and their investing fascia 

Posterior Mandible, Masseter and Pterygoid Muscles 

Superiorly - no distinct anatomical separation from the 

infratemporal fossa Inferiorly - no distinct anatomical separation 

from the submandibular and 

masticator spaces 

Contents of the Buccal Space 

• Fat 

• Parotid Duct 

• Minor Salivary Glands 

§ Accessory Parotid Gland (largest of the minor salivary 

glands) 

Nerves 

§ Buccal Division of the Facial Nerve 

§ Buccal Division of the Mandibular Nerve 

• Arteries 

§ Facial Artery 

Buccal Artery 

• Facial Vein 

• Lymph Node 

§ Anterior And Posterior Buccal Nodes 

Buccal Space Lesions 

• Salivary Glands 

§ Minor Salivary Gland Tumors 

§ Accessory Parotid Tumor 

• Parotid Duct 

§ Dilated Duct Secondary to Stone or Stricture 

• Lymph Nodes 

§ Lymphoma (can be extranodal) 

§ Metastatic Disease 

• Soft Tissues 

§ Sarcoma 

§ Hemangiomas 

§ Lipomas 

§ Abscess 

§ Cellulitis 

Masticator Space 

The boundaries are formed by the superficial layer of the 

deep cervical fascia, which splits to encase the muscles of 

mastication. This space is closed inferiorly because the fascial 

layers meet and fuse at the inferior border of the 

mandible. The medial and lateral superior boundaries differ. 

Superomedially the skull base forms the upper margin of this 

space. The foramen ovale, which is located in this portion of 

the skull base, facilitates the spread of tumor and infection 

intracranially from the masticator space along the course of 

V3 to the cavenous sinus. Superolaterally the upper margin 

of the masticator space is the point of attachment of the temporalis 

muscle to the outer table of the skull. This region is 

sometimes referred to as the temporal fossa or the suprazygomatic 

masticator space. 

Contents of the Masticator Space 

• Muscle 

§ Masseter Muscle 

§ Pterygoid Muscles (medial and lateral) 

§ Temporalis Muscle (lower portion) 

• Bone 

§ Ascending Ramus of the Mandible 

• Nerves 

71 

§ Mandibular Division of the V CN (V3) 

Pathology 

Findings on cross-sectional imaging that suggest the presence 

of a masticator space lesion are: 

1. Enlargement of the muscles of mastication 

2. Abnormality of the ascending ramus of the mandible 

3. Posteromedial displacement of the fat in the parapharyngeal 

space 

4. Enlargement of the foramen ovale or involvement of V3 

Masticator Space Lesions 

• Muscles 

§ Sarcoma 

§ Fibrous Histiocytoma 

§ Masseteric Muscle Hypertrophy 

§ Metastases 

§ Denervation Atrophy 

§ Myositis 

• Nerves 

§ Neural Tumors (Schwanomas) 

• Connective Tissue 

§ Hemangiomas 

§ Lymphangiomas 

§ Abscess/cellulitis 

• Mandible 

§ Osseous Lesions 

Parapharyngeal Space 

This space is located between the pharynx and masticator 

space. 

Boundaries of the parapharyngeal space 

• Medially - Fascia of the pharynx (nasopharynx and 

oropharynx) 

• Laterally - Fascia superficial layer of the deep cervical fascia 

(medial border of the masticator space). 

• Superiorly - Skull Base 

• Inferiorly - Hyoid Bone 

Parapharyngeal Space Lesions 

Lesions in this region are classified as pre and poststyloid (or 

as carotid or parapharyngeal space lesions) and can originate 

within the space or from adjacent structures (pharynx, masticator 

space, and parotid gland). 

Prestyloid Parapharyngeal Space 

§ Salivary gland tumors from the deep lobe of the parotid 

gland or parapharyngeal rests. 

§ Located anterior to the carotid sheath structures 

§ Displace the paraphayngeal space fat medially 

• Post Styloid Parapharyngeal Space 

§ Paraganglioma, nerve sheath tumor, neural tumors from 

the sympathetic chain. 

§ Displace the carotid sheath structures anteriorly 

§ Sympathetic chain lesions may displace the carotid sheath 

structured laterally 

• Masticator Space Lesions 

§ Posteromedial displacement of parapharyngeal fat 

• Pharyngeal Lesions 

§ Displace parapharyngeal fat laterally 

BIBLIOGRAPHY 

1. Hollingshead WH. Textbook of Anatomy. 3rd Ed. 

Hagerstown, Harper & Row:1974;821-823 

2. Som PM, Curtin HD. Parapharyngeal and Masticator Space 

Lesions. Ch. 38. In: Som PM, Curtin HD (eds). Head and Neck 

Imaging, Vol 2, 4th ed. St. Louis: Mosby Year Book, 

2003;1954-2004 

Tuesday

Tuesday 

3. Som PM, Curtin HD. Fasciae and Spaces. Ch. 34. In: Som PM, 

Curtin HD (eds). Head and Neck Imaging, Vol 1, 4th Ed. St. 

Louis: Mosby Year Book, 2003;1805-1828 

4. Tart RP, Kotzur IM, Mancusco AA, et al. CT and MR imaging 

of the buccal space and buccal space masses. Radiographics 

1995;15:531-550 


8:00 AM - 9:30 AM 

Room 107 

(24) Updates and Controversies on MR 

Safety: Can Patients with Implantable 

Cardiac Rhythm Devices Safely 

Undergo MR Imaging? (General) 

(ARS)* 

— J. Rod Gimbel, MD, FACC and 

Emanuel Kanal, MD, FACR 

Moderator: Emanuel Kanal, MD, FACR 

*Audience Response System 

Updates and Controversies on MR Safety: Can Patients 

with Implantable Cardiac Rhythm Devices Safely 

Undergo MR Imaging? 

J. Rod Gimbel, MD, FACC and Emanuel Kanal, MD, FACR 

Dr. Gimbel is Board Certified in Cardiology and Cardiac 

Electrophysiology. After completing his training in Internal 

Medicine at New York University-Bellevue Hospital Center, 

Dr. Gimbel completed fellowships in Cardiology and 

Electrophysiology at the Cleveland Clinic Foundation. Since 

completion of his training, he has been in private practice in 

Knoxville, TN. He is the author of several peer-reviewed 

manuscripts concerning cardiac rhythm device-MR imaging 

interaction. 



able to: 

1) Understand the growing indications and use of 

implantable cardiac rhythm devices. 

2) Understand the safety concerns and risks associated with 

scanning patients with implantable cardiac rhythm devices. 

3) Understand the strategies that have been used to allow 

patients with implantable cardiac rhythm devices undergo 

MR imaging. 

72 


Magnetic resonance (MR) imaging is the imaging modality 

of choice for a wide variety of conditions. Implantable cardiac 

pacemakers (PM) and cardioverter-defibrillators (ICD) 

are used to treat an expanding list of cardiac conditions. 

Traditionally, patients with electronically sensitive devices 

such as PMs and ICDs are prohibited from undergoing MR 

imaging because of numerous safety concerns. Neither manufacturers 

or the Food and Drug Administration (FDA) recognize 

currently marketed implantable PMs or ICDs as MR 

imaging safe. Despite this, because of clinical necessity, 

nearly 300 pacemaker and ICD patients have reportedly 

undergone safe MR imaging. While this experience is 

encouraging, significant concern remains about the safety of 

imaging patients with implantable cardiac rhythm devices. 

Among these concerns are transient and permanent device 

dysfunction, rapid cardiac pacing, pacemaker inhibition, and 

local thermogenic tissue destruction. And albeit under poorly 

characterized circumstances, several device patient deaths 

have undeniably occurred. Appropriately labeled FDA 

approved MR imaging compatible systems are under development. 

Available data suggests that currently marketed 

pacemakers and ICDs might undergo MR imaging in selected 

patients with an acceptable risk benefit profile. 

Recommendations have been made to facilitate “safe” scanning 

in device patients when clinical circumstances warrant. 

These include appropriate patient and device selection, 

thoughtful pre-MR imaging programming of the device, 

extensive monitoring and supervision during MR imaging, 

and appropriate follow up. Within the context of the rapidly 

expanding use of MR imaging and cardiac rhythm devices, 

we review the available data on the safety and feasibility of 

performing MR imaging on patients with pacemakers and 

ICDs. 


8:00 AM - 9:30 AM 

Room 103 


PowerPoint 

— John L. Go, MD 

— David S. Martin, MD


10:00 AM - 11:36 AM 

Room 105/106 

(26a) INTERVENTIONAL: 

Intracranial Aneurysms 



(26b) INTERVENTIONAL: Stroke Therapy and 

Stenting 

(26c) INTERVENTIONAL: Angioplasty, Stenting, and 

New Techniques 

(26d) ADULT BRAIN: Neoplasms and New 

Techniques 

Moderators: John C. Chaloupka, MD 

Aquila S. Turk, DO 


Abciximab in the First-Line Treatment and Prevention of 

Thromboembolism in Patients with Ruptured 

Aneurysms 

Aviv, R. I. 1 · Richard, O. 2 · Patel, M. 2 · Collqhoun, I. 2 

1 2 Sunnybrook Hospital, Toronto, ON, CANADA, Charing 

Cross Hospital, London, UNITED KINGDOM 

PURPOSE 

Abciximab (Reopro, Eli Lilly, Indianapolis, IN) is well 

described as a rescue treatment for thrombus development in 

the context of unruptured aneurysms. Although no well controlled 

studies have evaluated the risk of hemorrhage in the 

neurovascular interventional setting, no significant increase 

hemorrhagic risk was seen in the Abciximab in Ischemic 

Stroke Trial or the EPIC study. Reports of increased systemic 

hemorrhage-related complications in cardiac patients have 

tempered its application to the treatment of procedure-related 

thrombosis in the context of ruptured aneurysms. To date, 

1 case report and 1 series, constituting 5 patients, have 

described the use of Abciximab for thrombotic complications 

arising during the treatment of an acutely ruptured 

aneurysm. In all cases, aneurysms were secured initially 

prior to Abciximab administration. In one patient the use of 

Abciximab was a last resort after the use of further heparin, 

intravenous nonsteroidal (Ketorolac) and urokinase.We 

describe the first-line use of Abciximab in 13 patients with 

ruptured aneurysms. 


A retrospective single-center review of ruptured aneurysm 

patients who were treated with Reopro was performed. 

Demographic data, time from ictus, CT scan, and angiogram 

to treatment were recorded. Number of coils inserted, com- 

73 

plication, and percentage aneurysm occlusion at time of 

Reopro administration was documented. Subsequent patient 

course and outcome is recorded. 

RESULTS 

Thirteen patients received Reopro, 8 female, 5 male, median 

age was 56 years (range 28-79 years). Seven patients were 

WFNS grade 1, 4 grade 2 and 2 grade 5 (one of which 

improved to grade 2 prior to procedure). The median time 

from ictus to intervention was 4 days (range 4-23), but from 

initial diagnostic procedure to treatment 1 day. Aneurysms 

were within the ACOM (5/13), PCOM (5/13), and 1/13 each 

PICA, MCA and basilar. Mild to moderate spasm was present 

in 4/13 patients at time of treatment. The median number 

of coils placed in the aneurysm prior to Reopro administration 

was 5 (range 0-13) with all ruptured aneurysm secure, 

apart from 2 cases. Both were partially treated aneurysms (1 

basilar tip, 1 ACOM). All patients received iv Reopro bolus 

without infusion as a first-line treatment apart from one 

patient who was first treated with iv aspirin. The dose ranged 

from 5 mg (6/13) to 20 mg (1/13). Recanalization was complete 

in 8 (61.5%), partial in 3/13 (23%), and failed in 1. One 

patient was given Reopro prophylactically because of coil 

prolapse without evidence of thrombus formation. Eleven 

patients (85%) had no/transient weakness postprocedure. 

Three had symptomatic infarcts. One patient had coil-related 

rupture and died. One patient who had a parenchymal 

hematoma following ruptured aneurysm required hematoma 

evacuation and shunt 2 days after coiling following a decline 

in GCS due to perihematoma edema. There was no increase 

in hematoma size following Reopro administration. 

CONCLUSION 

We demonstrate successful first line use of Reopro in the 

treatment of thromboembolic complications in patients with 

ruptured aneuysms. Eighty-five percent of patients with documented 

thrombotic complications were asymptomatic at 

discharge. Hemorrhage following aneurysm rupture after 

Reopro administration, although uncommon, may be more 

severe, proving fatal in our single case (8%). 

KEY WORDS: Abciximab, ruptured, aneurysm 


Endovascular Treatment of Aneurysms: Comparative 

Gene Expression of Growth Factors, MMPs, and TIMPs 

of Neointimal Cells Recruited on the Embolic Agent and 

Evolution with Flow Patterns, Time and with Recurrence 

in an Experimental Model 

Raymond, J. 1 · Ogoudikpe, C. 1 · Salazkin, I. 1 · Metcalfe, A. 1 · 

Gevry, G. 1 · Chagnon, M. 2 · Robledo, O. 1 

1Centre Hospitalier Universitaire Montréal, Notre Dame 

Hospital, Montreal, PQ, CANADA, 2University of Montreal, 


PURPOSE 

Molecular events after embolization of aneurysms remain 

unknown. We attempted to identify genes associated with 

healing or recurrence in a model in which neointima at the 

neck varies according to flow zones. 

Tuesday

Tuesday 


Bilateral carotid venous pouch aneurysms were constructed 

in 36 dogs and embolized with gelatin sponges. Angiography 

(n = 22) and pathologic studies (n = 17) were performed at 

T 0 and 3 weeks and results were scored using a qualitative 

index applied to the distal (inflow) and proximal (outflow) 

zones of the neck. In 14 animals, mRNA expression at the 

proximal or distal segment of the sponge was analyzed by 

RT-PCR. 

RESULTS 

The model recurs at 3 weeks as shown by significantly worse 

angiographic scores as compared to T 0 (P < .01). Neointimal 

scores differed, with a more complete neointima at the proximal 

aspect of the sponge (P = 0.027). Embolization was followed 

by migration of CD31 + , CD14 + , smooth muscle αactin 

+ (SMA + ) cells that progressively expressed metalloproteinases 

(MMP-9,-12,-14), but stable or lesser, retarded 

expression of inhibitors (TIMP1-4). Growth factors (PDGF- 

BB, TGF-β1, TNF-α, MCP-1 and Ang-1) were expressed at 

increasing levels, maximal at 7-14 days. Differences 

between distal and proximal zones were limited to slightly 

increased expression of MMP-2 proximally (P < .035). 

CONCLUSION 

Gene expression after embolization is compatible with patterns 

associated with neointima formation. The identification 

of key factors involved in recurrence needs additional studies. 

KEY WORDS: Aneurysms, experimental model, growth factors 

and metalloproteases 


Characteristics of Intracranial Aneurysms Associated 

with Rupture Complicating Endovascular Treatments 

Summers, P. E. · Padmanabhan, R. · Corkill, R. · Kueker, W. 

· Molyneux, A. · Byrne, J. V. 

University of Oxford 

Oxford, UNITED KINGDOM 

PURPOSE 

Meta-analyses of published reports have shown that rupture 

occurs more frequently when endosaccular coil embolization 

is performed in small aneurysms acutely after SAH, but no 

correlations have been made with other aneurysm attributes 

because of small case numbers (1, 2). A single center study 

of prospectively collected data was performed to assess the 

relevance of aneurysm characteristics to this complication. 


Over a 12-year period 1477 EVTs were performed on 1619 

aneurysms. Complicating aneurysm perforations occurred in 

36 (2.22%). Comparisons of patient (age, sex, delay since 

last SAH, WFNS grade), and aneurysm (location, size, prior 

rupture, CT Fisher grade, vasospasm) attributes were made. 

A standardized endosaccular coil embolization technique 

was used and the study period stratified into three 4-year 

periods. For each attribute, rupture rates were compared 

between subgroups using χ 2 tests (α < 0.05). 

74 

RESULTS 

The distribution of procedural rupture events is summarized 

in Table 1. Rupture rates were similar in the patient subgroups 

for gender and age. No increased frequency was 

identified in sessions in multiple aneurysm treatment sessions 

(1.44% vs 2.39% for single aneurysm sessions), or 

when the experience was stratified over the period of data 

collection or by delay since last SAH (> 15 days vs > 15 

days). The aneurysm attributes associated with significant 

differences in rupture incidence were: previous rupture status, 

size, vasospasm at time of the interventional procedure, 

WFNS scores of 1 vs 0, and any elevated Fisher grade. Our 

data are suggestive of differences between vascular sites in 

the frequency of rupture, though only at the level of α < 0.1. 

Amongst the sites with more than 200 treated aneurysms, the 

PComm and BA termination and AComm showed similar 

rupture rates (2.16%, 2.2%, and 2.93% respectively), which 

closely approximated the population average, while the 

MCA trifurcation showed a significant, 5-fold lower incidence 

(0.44%). The remaining locations all had less than 90 

treatment cases. Compared to the pooled data for locations 

with more than 200 cases, the PICA showed a significantly 

greater (8.16%) rupture incidence. 

Summary of Aneurysm Rupture Rates 

Sex Freq Age Freq Prev. Freq Size* Freq WFNS Freq Fisher* Freq 

(%) (%) Rupt* (%) (%) (1 v 0)* (%) (all v 0, (%) 

1,2 v 3,4 

F 2.54


Selective Coiling of Aneurysm and Parent Artery for the 

Treatment of Unruptured or Recurrent Wide-Neck 

Proximal Anterior Cerebral Artery Aneurysms 

Silvaggio, J. A. · Guilbert, F. · Roy, D. · Raymond, J. · Iancu- 

Gontard, D. · Weill, A. 

Hôpital Notre-Dame 

Montréal, PQ, CANADA 

PURPOSE 

Aneurysms of the proximal segment of the anterior cerebral 

artery (A1) are uncommon. Anatomical features of these 

aneurysms present unique surgical challenges in visualization 

of the aneurysm neck and preservation of related anterior 

basal perforating arteries. Surgical trapping and proximal 

artery occlusion have been described as alternative strategies 

for treating aneurysms deemed to be unclippable. 

Endovascular coil embolization of these aneurysms also has 

been reported, however wide neck aneurysms are known to 

be prone to recurrence following endovascular treatment. We 

describe the radiographic and clinical results of an alternate 

strategy of coil embolization of the aneurysm and parent 

artery occlusion for treatment of these uncommon lesions. 


Between January 2000 and November 2004, 624 cerebral 

aneurysms were treated endovascularly at our institution. 

Review of our prospective database revealed five cases of 

A1 aneurysms treated by coil embolizaton of the aneurysm 

and intentional parent artery occlusion after confirmation of 

adequate collateral circulation. Further detailed evaluation of 

these patients was undertaken by reviewing medical charts 

and follow-up angiography. Pre and postembolization crosssectional 

images also were reviewed, looking for procedurerelated 

radiographic complications. Demographic characteristics, 

presentation, angiographic characteristics, immediate 

treatment success, procedure-related clinical and radiographic 

complications, and recurrence rate were collected and analyzed. 

RESULTS 

Proximal anterior cerebral artery aneurysms accounted for 

0.8 percent of all cerebral aneurysms treated by endovascular 

methods at our institution during the study period. Mean 

age at presentation was 61 years (range 55 to 74 years). Four 

of the five patients were female. Three patients initially presented 

with subarachnoid hemorrhage, one with visual 

symptoms attributable to aneurysm mass affect, and one 

incidentally. Three patients had undergone previous procedures 

directed towards the aneurysm: one patient underwent 

selective aneurysm coiling following a late asyptomatic 

recurrence after surgical clipping; two patients had progressive 

recurrence after previous aneurysm coiling. The two 

patients with unruptured aneurysms had coil embolization of 

the aneurysm and parent artery occlusion as the primary 

treatment. Average aneurysm size was 14 mm (range 8 to 20 

mm). The aneurysm neck was considered wide (> 4 mm) in 

all cases. Multiple aneurysms were present in one patient. 

Aneurysm embolization and parent artery occlusion with 

platinum coils was initially successful in all cases. There 

were no symptomatic procedure-related complications. One 

patient was found to have a clinically silent left caudate head 

75 

infarct on routine postembolization CT scan. Recanalization 

of the aneurysm neck and the previously occluded A1 segment 

was identified in one patient on follow-up angiography 

CONCLUSION 

Parent artery occlusion for the treatment of unruptured or 

recurrent wide-neck A1 aneurysms appears to be safe in the 

presence of adequate collateral circulation. Further evaluation 

of this approach is required to determine the impact on 

recurrence and perforator-related ischemic complications. 

KEY WORDS: Aneurysm, coiling, A1 


Orbit Complex Coils for Intracranial Aneurysms: 

Midterm Angiographic Results with Focus on Packing 

Density 

Wakhloo, A. K. · Linfante, I. · Sandhu, J. S. · Perlow, A. · 

Gounis, M. J. · Sadasivan, C. · Lieber, B. B. 


Miami, FL 

PURPOSE 

In vivo and in vitro data support that complex-shaped 

Orbit coils achieve a higher aneurysm packing density 

than any currently available platinum coils. This may reduce 

the rate of recanalization associated with aneurysm coiling. 

We present midterm angiographic follow up in patients treated 

with the Orbit system. 


Between May 2002 and July 2004 97 patients with a total of 

115 aneurysms were treated using the Orbit system. The 

mean age of patients was 59 years, with a range of 11-86 

years (ratio female to male 2:1). Thirty-seven patients (32%) 

presented with SAH after aneurysm rupture. Eighty-seven 

aneurysms (75%) were located in the anterior circulation, 

with 14 MCA bifurcation aneurysms and 21 PComA 

aneurysms. Twenty aneurysms (17%) were located at the 

basilar tip. Eighteen aneurysms were < 4 mm, while 77 and 

20 aneurysms were 5-10 mm and > 10 mm respectively. 

Orbit system alone was used in 73 aneurysms and in combination 

with Neuroform stent in 8 aneurysms. In remaining 

34 aneurysms Orbit was combined with Matrix, GDC, 

Sapphire, or hydrocoils. Aneurysms treated with Orbit alone 

had an average volume of 893.8 ± 212.2 mm 3 corresponding 

to an average size of 11 mm. The average aneurysm neck 

size was 4.36 ± 1.96 mm (range 2-12 mm). Computational 

study showed an average coil packing density of 34.4 ± 19%. 

RESULTS 

Complete or near-complete packing could be achieved in all 

aneurysms, but required balloon-assistance in 22% of the 

aneurysms. Follow-up angiography was available in 32 

aneurysms (11 ruptured) treated with Orbit alone. Average 

angiographic follow up was 8.5 months (range 6-21). We did 

not observe any rerupture in the follow-up interval. 

Recanalization was seen in 5/32 aneurysms (15.6%) including 

3 patients with basilar tip aneurysms. Retreatment was 

required in 3 aneurysms (9.3%). Device-related permanent 

deficit was seen in 2 patients (2.7%). One patient required 

surgical removal of a coil that fractured. The patient experienced 

an embolic event that resulted in progressively 

Tuesday

Tuesday 

improving mild expressive aphasia. Another patient reruptured 

her aneurysm during coiling with delayed neurologic 

recovery. 

CONCLUSION 

Orbit provides a user-friendly, safe, effective, and innovative 

modality for the treatment of aneurysms. Complex coils 

allow a high packing density. Midterm data show a low 

recanalization rate as compared with other platinum coils, or 

bioactive and coated coils. Basilar bifurcation aneurysms 

remain a challenge. 

KEY WORDS: Intracranial aneurysms, endovascular treatment, 

complex coils 


Can Aneurysm Neck Size Elastase-Induced Aneurysms 

Be Controlled? A Prospective Study 


Kadirvel, R. · Mandrekar, J. N. · Cloft, H. J. · Kallmes, D. F. 

Mayo Clinic 

Rochester, MN 

PURPOSE 

We report this prospective study in order to confirm our prior 

work, which indicated that the neck size of elastase-induced 

aneurysm model in rabbits could be controlled by adjusting 

the position of inflated balloon. 


Ninety elastase-induced, saccular aneurysm models in rabbits 

were analyzed prospectively. We defined the study sample 

based on height of the balloon relative to the origin of the 

right common carotid artery (RCCA). Group 1 (n = 62) comprised 

cases where the occlusion balloon resided low, completely 

within the brachiocephalic/subclavian arteries. Group 

2 (n = 28) included cases with a high balloon position, where 

the balloon resided within both the RCCA and the brachiocephalic/subclavian 

arteries. Follow-up digital subtraction 

angiography (DSA) was performed in all cases. The necks of 

aneurysms were characterized as narrow (diameter, < or = 4 

mm) or wide (diameter, > 4 mm). The aneurysm dimensions, 

including neck diameter, dome:neck ratio, and aneurysm 

width and height, were measured and compared between 

groups. Also, the proportions of narrow- and wide-necked 

aneurysms were compared between groups. Continuous 

variables were compared using the Student’s t test. 

Proportions were compared using the Chi-Square test. 

RESULTS 

The mean aneurysm neck diameter of Group 1 was significantly 

larger than that of Group 2 (Fig. 1) (3.4 ± 1.2 mm versus 

2.3 ± 0.9 mm, respectively, p < .01). The proportion of 

wide-necked aneurysms in Group 1 was significantly larger 

than that in Group 2 (38% versus 4%, p < .05). Mean 

dome/neck ratios were 1.2 ± 0.4, and 1.7 ± 0.7 for Groups 1 

and 2, respectively (p < .01). Mean aneurysm width in Group 

1 was larger than that in Group 2 (3.8 ± 1.0 mm versus 3.2 ± 

0.9 mm, p < .05). There was no significant difference in 

aneurysm height between Groups 1 and 2 (8.0 ± 1.7 mm 

verus 7.5 ± 2.2 mm, p > .05). 

76 

Fig. 1 A, Group 1. Intraarterial digital subtraction angiogram 

(DSA), right anterior oblique view, demonstrating a widenecked 

aneurysm. 

Fig. 1 B, Group 2. Intraarterial digital subtraction angiogram 

(DSA), right anterior oblique view, demonstrating a narrownecked 

aneurysm. 

CONCLUSION 

The neck size of elastase-induced aneurysm models in rabbits 

can be controlled by adjusting the position of the inflated 

balloon, with balloon positioning that bridges from the 

CCA to the subclavian/brachiocephalic arteries yielding 

wider-necked aneurysms. 

KEY WORDS: Aneurysm model, rabbit 


Evaluation of Healing Response of Matrix® Bioactivity 

in a Canine Sidewall Aneurysm Model 

Turk, A. S. 1 · Grum, K. 1 · Consigny, D. 1 · Rappe, A. 1 · Grinde, 

J. 2 · Polyakov, I. 2 · Carr-Brindle, V. 2 

1 2 University of Wisconsin, Madison, WI, Boston Scientific 

Corporation, Fremont, CA 

PURPOSE 

Endovascular treatment is an accepted alternative to open 

surgery for many aneurysms but often is subject to aneurysm 

recurrences. Development of bioactive coils that accelerate 

the healing response may offer benefits in reducing this 

recurrence rate. 


Lateral wide-necked sidewall morphology aneurysms were 

surgically created in the common carotid artery of 15 

canines. Ten of the aneurysms were treated with Matrix® 

detachable coils and 5 were treated with standard Guglielmi 

detachable coils (GDC®), using standard embolization techniques. 

All aneurysms were evaluated angiographically and 

explanted 14 days after embolization. The explanted specimens 

were sent for histopathologic evaluation. 

RESULTS 

Packing densities were 22% for Matrix® and 26% for 

GDC®-treated aneurysms. Angiographic assessment of 

Matrix-treated aneurysms indicated 6/10 completely occluded 

at embolization and 10/10 at explant (2 weeks later). 

Angiographically, all GDC® aneurysms were occluded 

completely at embolization (4/4) and at explant. One of the

GDC®-treated aneurysms was atypically small and wide 

necked which resulted in coil extrusion into the parent artery 

and was therefore not included in further analysis. 

Histopathology showed recanalization in half (2 of 4) of the 

GDC® and 20% (2 of 10) of Matrix®-treated aneurysms. 

Three of four aneurysms treated with GDC® demonstrated 

mostly unorganized thrombus within the coil mass compared 

to 3/10 aneurysms treated with Matrix®. The Matrix®-treated 

aneurysms demonstrated increased fibrocellular tissue 

and inflammation, with less vascular spaces, relative to 

GDC® treated aneurysms. 

CONCLUSION 

Canine experimental wide-necked aneurysms treated with 

Matrix® demonstrated advanced healing at 14 days as compared 

to standard GDCs®, similar to that previously reported 

in swine. 

KEY WORDS: Aneurysm, coil, canine 


Long-Term Patency of Elastase-Induced Aneurysm 

Model in Rabbits 


Kadirvel, R. · Mandrekar, J. N. · Cloft, H. J. · Kallmes, D. F. 

Mayo Clinic 

Rochester, MN 

PURPOSE 

Long-term patency in untreated experimental aneurysms 

represents a critical attribute of any system proposed for the 

testing of aneurysm occlusion devices. Our purpose was to 

evaluate the long-term patency in elastase-induced saccular 

aneurysm models in rabbits. 


Serial intravenous digital subtractive angiography (IV DSA) 

was done in 20 elastase-induced saccular aneurysm models 

in rabbits 1, 3, 6, 9, and 12 months after creation. Aneurysm 

dimensions, including neck diameter, and aneurysm width 

and height, were measured and calculated from IV DSA 

images. Comparison of the aneurysm sizes across time were 

performed using the Wilcoxon paired signed rank test and 

Friedman test. 

RESULTS 

None of the 20 aneurysms showed spontaneous thrombosis 

at any time-point. Mean dimensions did not change over 

time for any parameter. Mean aneurysm widths at 1, 3, 6, 9, 

and 12 months were 3.8 ± 0.8 mm, 3.9 ± 0.9 mm, 4.1 ± 1 

mm, 4.0 ± 1.1 mm, and 4.1 ± 1.0 mm, respectively, p = .356. 

Mean aneurysm heights at 1, 3, 6, 9, and 12 months were 8.7 

± 1.7 mm, 8.4 ± 1.5 mm, 8.7 ± 2.0 mm, 8.8 ± 1.7 mm, and 

8.6 ± 1.6 mm, respectively, p = .471. Mean neck diameters 

at 1, 3, 5, 9, and 12 months were 3.5 ± 0.7 mm, 3.4 ± 1.0 mm, 

3.5 ± 0.9 mm, 3.5 ± 0.9 mm, and 3.6 ± 0.9 mm, respectively, 

p = .117). 

77 

Figure A and B. Rabbit number 7: IV DSA 1 month and 1 

year after creation. The aneurysm remains patent and stable 

radiographically. 

CONCLUSION 

Long-term patency in elastase-induced saccular aneurysm 

models in rabbits is excellent. Aneurysm dimensions remain 

stable for up to 1 year following creation. 

KEY WORDS: Aneurysm model, patency, rabbit 

Acknowledgment: This project was supported by NIH grant 

NS42646. 


Aneurysm Embolization with Matrix Detachable Coils: 

Assessment of Durability at 6-Month Follow Up 

Fiorella, D. 1 · Albuquerque, F. C. 2 · McDougall, C. G. 2 

1 2 Cleveland Clinic Foundation, Cleveland, OH, Barrow 

Neurological Institute, Phoenix, AZ 

PURPOSE 

Matrix detachable coils (MDC) are platinum coils coated 

with a bioabsorbable polymeric material (BPM) - polyglocolic-polylactic 

acid (PGPLA) - engineered to create a cellular 

reaction to promote scar tissue formation after aneurysm 

embolization. Theoretically, this would result in a more 

durable occlusion of aneurysms then bare platinum coils. 

The current study was undertaken to assess the durability of 

aneurysm occlusion after embolization with Matrix coils. 


Patients were enrolled prospectively in a database. Coils 

used for embolization were recorded in the operative notes 

for the procedure. Only aneurysms embolized with 50% or 

greater volume of MDC were included. All patients with 

Neuroform stents were excluded from the study. Patients 

were followed with conventional angiography (CA) and MR 

angiography (MRA). Patients with suspected recanalization 

on MRA underwent CA. 

RESULTS 

Over an 18-month period, 123 patients with 125 aneurysms 

underwent MDC aneurysm embolization (without 

Neuroform stent support). Follow-up data (average 5.9 

months; range 1-17 months) were available for 68 

aneurysms (48 with CA, 20 with MRA only). Aneurysm 

sizes were 52 small aneurysms with small necks (SASN), 4 

Tuesday

Tuesday 

small aneurysms with wide necks (SAWN), 11 large (L) 

aneurysms, and 1 giant (G) aneurysm. Overall, there were 22 

recanalizations (32%), 13 of which required retreatment 

(19%); 46 aneurysms were unchanged or demonstrated progressive 

thrombosis (68%). Progressive thrombosis was 

demonstrated for 18 of 48 aneurysms with CA follow up 

(37.5%). The recanlization rate for SASNs was 23% (12/52) 

with a 13.5% retreatment rate (7/52). The recanalization rate 

for large aneurysms was 72.7% (8/11) with a 54.5% retreatment 

rate (6/11). 

CONCLUSION 

In the absence of Neuroform stent support, aneurysms 

embolized with the MDC system demonstrated relatively 

high rates of recanalization — particularly large aneurysms. 

Many of the recanalizations were of sufficient size to warrant 

retreatment. A high rate of progressive thrombosis also 

was observed. The current data suggest the possibility of two 

patient subgroups — Matrix “responders” and “nonresponders.” 

In patients who do not “respond” to Matrix, the 

absorbtion of the polymer coating would be expected to 

result in the loss of approximately 70% of the initial coil 

packing volume, thus accounting for the high rates of significant 

recanalization. 

KEY WORDS: Aneurysm, matrix, embolization 


Surveillance of Intracranial Aneurysms Treated with 

Detachable Coils: A Comparison of MR Angiographic 

Techniques 

Willinsky, R. A. 1 · Farb, R. I. 1 · Nag, S. 1 · Scott, J. N. 2 · 

Marotta, T. 3 · Montanera, W. J. 3 · Tomlinson, G. 4 · terBrugge, 

K. G. 1 


2 3 Foothills Medical Centre, Calgary, AB, CANADA, St. 

Michael’s Hospital, Toronto, ON, CANADA, 4University Health Network, Toronto, ON, CANADA 

PURPOSE 

To evaluate and compare the performance of two MR angiographic 

techniques for the follow-up of coiled intracranial 

aneurysms. 


Over a period of 18 months 29 coiled aneurysms (in 28 

patients) underwent follow-up evaluation. Each follow-up 

evaluation included a 3D time-of-flight MR angiogram 

(TOF MRA), an autotriggered elliptic-centric-ordered threedimensional 

gadolinium-enhanced MR angiogram (ATECO 

MRA) as well as a selective intraarterial digital subtraction 

catheter angiogram (DSA) which served as the “gold standard.” 

RESULTS 

Of the 36 follow-up assessments, visualization to allow confident 

interpretation was experienced in 36 (100%) of the 

ATECO MRAs. Adequate visualization was seen in 32 

(89%) of the TOF MRAs yielding 4 cases of imaging failure 

with TOF MRA. Of 36 follow-up assessments, 11 residual 

aneurysm components (RACs) greater than 2 mm were 

described on DSA (31% of treated aneurysms). Of these, 9 

were seen on ATECO MRA (sensitivity of 81% and speci- 

78 

ficity of 88%) and 4 were seen on TOF MRA (sensitivity of 

40% and specificity of 90%) excluding one case where TOF 

MRA registered as an “imaging failure.” The two RACs not 

seen on ATECO MRA both measured 3 mm. The six RACs 

not seen on TOF MRA measured 3, 4, and 5 mm. Digital 

subtraction angiography recommended retreatment in 2 

cases. Retreatment was additionally recommended by 

ATECO and TOF MRA in another case and by ATECO alone 

in a further case. 

CONCLUSION 

ATECO MRA provides a noninvasive reliable method of 

angiography for the surveillance of coiled aneurysms and is 

superior to TOF MRA for this purpose. 

KEY WORDS: Cerebral aneurysm, MR imaging, angiography 


Superior Cerebellar Aneurysms: Vertebral Artery 

Dominance and Basilar Artery Tortuosity 

Corrigan, K. · Ringer, A. · Tomsick, T. A. 

University of Cincinnati Medical Center 

Cincinnati, OH 

PURPOSE 

Superior cerebellar artery (SCA) aneurysms represent 

approximately 1-2% of all intracranial aneurysms and more 

commonly occur left sided. Vertebrobasilar anatomical features 

have been suggested to be the etiologic factors in their 

occurrence and location. Our purpose is to investigate the 

role of vertebral artery dominance and basilar artery tortuosity 

and curving in the formation of these aneurysms. Our 

hypothesis is that vertebral dominance results in basilar 

artery (BA) curvature secondary to unequal hemodynamic 

force entering the basilar from a dominant vertebral artery 

(VA), thereby directing a force vector at the origin of the 

SCA resulting in aneurysm formation. Furthermore, as leftsided 

aneurysms are more common, it should follow that a 

rightward convexity of the basilar also should be more common. 


We reviewed 24 SCA origin aneurysms, 15 from patients 

managed at the University of Cincinnati University Hospital, 

9 collected from a literature review. For each patient, the 

angiogram was reviewed and the side of the aneurysm, VA 

dominance, direction of BA curve, presence or absence of 

P1, and SCA single/multiple branch anatomy was noted. 

RESULTS 

Of our 24 aneurysms, 18 (75%) were left-sided aneurysms. 

Of 17 with adequate image documentation, 7 had dominant 

left VA anatomy, 8 had codominant VA anatomy, and 3 had 

dominant right VA anatomy. Of the 7 cases with dominant 

left VA anatomy, 3 had a convex right BA curve proximal to 

the aneurysm. In addition, 6 of the 8 cases where there was 

no VA dominance also were associated with a BA curve, suggesting 

that VA dominance alone is not the cause of basilar 

curvature. No significant correlations of dominant VA, or 

basilar tortuosity were discovered with left-sided aneurysms. 

Vertebral artery dominance and BA curvature in the 6 right 

SCA aneurysms were not correlative.

CONCLUSION 

Therefore, while our analysis again confirms that SCA 

aneurysms are more commonly left sided, it fails to confirm 

the hypothesis that flow dynamics attributed to VA dominance 

and BA curvature are the primary factor contributing 

to their formation. 

KEY WORDS: Aneurysms, superior cerebellar 


Intracranial Traumatic Arterial Aneurysms: 

Pathophysiology, Natural History, and Treatment 

Strategies: Experiences in Ramathibodi Hospital, 

Thailand 

Pongpech, S. · Jiarakongmun, P. 

Ramathibodi Hospital 


PURPOSE 

In Thailand, traumatic aneurysms frequently are related with 

high velocity vehicle accidents, especially motorcycle accidents. 

We have more than 5 million motorcycles spreading 

all over the country. Traumatic intracranial aneurysms are 

considered rare conditions, about 0.15-0.4% of all intracranial 

aneurysms, and about 8% of posttraumatic subarachnoid 

hemorrhage (SAH) have traumatic aneurysms. This condition 

carries predictable high mortality and morbidity, ranging 

between 27-50% and even higher in ruptured traumatic 

aneurysms. Awareness and properly selected investigation 

and treatment strategy can prevent and reduce incidences of 

such high morbidity and mortality. 


We retrospectively reviewed all patients with traumatic intra 

and extracranial vascular lesions in Ramathibodi hospital 

between June 1995 and May 2005, and describe in terms of 

incidence, pathophysiology, natural history, treatment strategy, 

and results. 

RESULTS 

Among 320 cases of traumatic intra and extracranial vascular 

lesions (during 1995-2005), we found 285 carotid cavernous 

fistulae (TCCF), 16 external carotid (ECA) fistulae, 

41 traumatic aneurysms of internal carotid arteries (ICA) 

(19/41 isolated aneurysms), 17 isolated arterial dissection 

with 9 associated aneurysms, 10 ECA aneurysms, 5 vertebrovertebral 

fistulae (VVF), 3 vertebral aneurysms, 2 anterior 

cerebral artery (ACA) aneurysms and 1 middle cerebral 

artery (MCA) aneurysm. In our series 33/52 of traumatic 

aneurysms are related to traumatic AVF. 

CONCLUSION 

In Thailand, most of intracranial aneurysms result from high 

velocity vehicle accidents, and contain high risk of mortility 

and morbidity. Sixty-one percent of cases (17/28) have associated 

fractured skull and traumatic AVF, as well as a warning 

criteria for higher risk traumatic aneurysms. Symptoms 

and risk of traumatic aneurysms depend on their locations; 

extradural lesion has lower risk except for false aneurysm in 

sphenoid sinus (25%) that could present with life-threatening 

epistaxis (11.5%). Supraclinoid and cortical aneurysms 

have higher risk of recurrent subarachnoid and intracranial 

hemorrhage frequently (42%) and recently within 2-3 week 

79 

intervals (80%). Location of hematoma other than subarachnoid 

blood is suggestive of traumatic aneurysm ( i.e., interhemispheric 

hematoma is predilection to the distal ACA 

aneurysm). Base of skull lesions (i.e., vertebral aneurysms) 

contain lower risk of intracranial hemorrhage and have higher 

potential to heal themselves. In our institute, interventional 

neuroradiology is treatment of choice for the patients suffering 

from traumatic aneurysms, which need senior well 

experienced staff to evaluate urgent intervention to prevent 

deteriorated irreversible condition. Detachable balloon 

embolizations are the treatment of choice for high-flow 

carotid cavernous fistulae, as well as other embolic material 

(NBCA/coils) in arteriovenous fistulae and traumatic 

aneurysm. Surgical treatments are feasible for removal of the 

hematoma and mass effect plus curative aim for traumatic 

cortical aneurysms. 

REFERENCES 

1. Schievink WI. Intracranial aneurysms. N Engl J Med 

1997;336(1):28-40 

2. Pia HW, Langmaid C, Zeirski J. Cerebral Aneurysms. 

Advances in Diagnosis and Therapy. Springer Verlag, 1979 

3. Weir B. Aneurysms Affecting The Nervous System. Williams 

and Wilkins:1987;209-261 

4. Lasjaunias P. Segmental identity and vulnerability in cerebral 

arteries. Interven Neuroradiol 2000;6:113-124 

KEY WORDS: Traumatic aneurysm 


10:00 AM - 11:33 AM 

Theatre 

(26b) INTERVENTIONAL: Stroke 

Therapy and Stenting 



(26a) INTERVENTIONAL: Intracranial Aneurysms 




Techniques 

Moderators: Christopher F. Dowd, MD 

Allan J. Fox, MD, FRCPC, FACR 

Tuesday

Tuesday 


Role of CT Angiography in Selecting Patients with Acute 

Ischemic Stroke for Intraarterial Thrombolysis 

Xavier, A. R. 1 · Dash, S. 2 · Gupta, G. 1 · Prestigiacomo, C. J. 1 

· Farkas, J. 3 

1New Jersey Medical School/University of Medicine & 

Dentistry of New Jersey, Newark, NJ, 2Long Island College 

Hospital, Brooklyn, NY, 3Maimonides Medical Center, 

Brooklyn, NY 

PURPOSE 

CT angiography is being used increasingly in the evaluation 

of cerebrovascular diseases. We evaluated the safety and 

accuracy of CT angiography in selecting patients with acute 

ischemic stroke presenting within 6 hours of symptom onset 

for intraarterial thrombolytic therapy. 


The medical records of patients with acute ischemic stroke 

with attempted intraarterial thrombolysis at our institution 

between October 1999 and March 2003 were reviewed. As 

part of our institutional protocol, all acute stroke patients 

underwent a noncontrast CT head followed by CT angiography 

during the initial evaluation. Patients within 6 hours of 

symptom onset who had occlusion of a major intracranial 

blood vessel on CT angiography underwent conventional 

angiography followed by intraarterial thrombolysis with a 

combination of pharmacologic and mechanical tools. The 

CT and conventional angiographic images were assessed 

independently to document the initial site of arterial occlusion. 

In addition, the charts were reviewed to assess for renal 

and fluid-electrolyte abnormalities on admission and during 

the hospital course. 

RESULTS 

A total of 40 patients presenting within 6 hours of symptom 

onset (age 65 ± 17 years; range: 21-94; 22 were men) were 

identified to have large vessel occlusions on CT angiography 

over a 40-month period. Twenty-four patients (60%) had 

arterial occlusions in the middle cerebral artery (MCA); 11 

(28%) had occlusions of the internal carotid artery (ICA), 4 

(10%) had occlusions of the basilar artery (BA), and one 

(2%) had occlusion of the anterior cerebral artery (ACA). 

Subsequent conventional angiography confirmed the site of 

arterial occlusion in every patient. Every patient received 

intraarterial thrombolysis except one patient who had an 

MCA occlusion on CT angiogram and spontaneously 

recanalized during the angiogram. There was a 100% correlation 

between the sites of arterial occlusion identified by CT 

angiography and subsequent conventional angiography in 

this patient cohort. One of the patients who had baseline 

renal insufficiency developed transient azotemia following 

the procedure; however, none of the patients had permanent 

worsening of renal function or required hemodialysis during 

the hospital stay. 

CONCLUSION 

CT angiography appears to be an accurate and safe screening 

tool in patients with acute ischemic stroke, particularly when 

evaluating for arterial occlusion of a major cerebral blood 

vessel that might be amenable to intraarterial thrombolytic 

therapy. Whether this strategy of preselecting acute stroke 

patients for aggressive intraarterial therapy leads to an 

80 

improved risk-benefit profile for acute stroke interventions 

needs to be evaluated prospectively in a larger group of 

patients. 

KEY WORDS: CT angiography, acute ischemic stroke, 

thrombolysis 


Catheter Navigation, Thrombolysis, and Stenting of 

Acute Symptomatic Cervical Internal Carotid Artery 

Occlusion 

Lum, C. · Srinivasan, A. · Stys, P. · Hogan, M. · Miller, W. · 

Nguyen, T. · Sharma, M. · Goyal, M. 

University of Ottawa 

Ottawa, ON, CANADA 

PURPOSE 

The treatment of acute anterior circulation stroke distal to an 

occluded cervical internal carotid artery at the common 

carotid artery bifurcation (cICA) presents a challenge. There 

is evidence suggesting treatment only with iv-tpa results in 

poor outcomes (1). Case series have described some success 

with endovascular treatment (2). However, published experience 

is lacking regarding residual stenoses and clot at the 

cICA postacute thrombolysis. Since 2001, our protocol for 

treatment for acute stroke presenting under 6 hours included 

an option for direct intraarterial thrombolysis of large vessel 

(ICA,M1) clot. No patient with large vessel occlusion is 

denied standard iv tPA if presenting under 3 hours. Our goal 

was to evaluate the outcomes of patients who presented with 

symptoms related to cICA occlusion who underwent 

endovascular treatment during this period and to describe the 

rationale and technique for acute cICA stenting of residual 

clot/stenosis postthrombolysis. 


All patients presented < 6 hours from stroke onset. CT 

angiography and perfusion studies were performed to confirm 

location of clot and to estimate the degree of penumbra. 

When possible, iv tPA was given prior to endovascular treatment. 

The techniques for vessel recanalization were analyzed. 

Postprocedure CT scans were reviewed for hemorrhage. 

Outcomes were assessed using the modified Rankin 

scale (mRS). Good outcomes were assigned an mRS < = 2. 

RESULTS 

Over a 2-year 3-month period, a total of 24 patients underwent 

endovascular treatment for acute stroke under this protocol. 

Six patients had cICA clot. Five of 6 had preprocedure 

noninvasive imaging studies confirming the occluded ICA 

and perfusion studies suggesting a large penumbra and significant 

neurologic symptoms related to the affected hemisphere. 

Five of 6 had clot in the MCA distal to an occluded 

ICA, one patient occluded at the origin of the ICA postendarterectomy. 

All were treated with intraarterial tPA through 

a catheter advanced through the occluded ICA. Three of 6 

underwent acute ICA stenting for flow across residual cICA 

clot/stenosis. Two of 6 had angioplasty of MCA clot. Four of 

5 patients in which TIMI 3 flow was achieved had good outcomes, 

one patient had mRS = 3, one patient mRS = 5. There 

were no deaths. One patient who did not undergo stenting 

had an asymptomatic intracranial bleed.

CONCLUSION 

Thrombolysis of intracranial clot through an occluded ICA is 

feasible, with a low morbidity and may be associated with 

good outcomes. In this series, acute cICA stenting was not 

associated with any adverse events. 

REFERENCES 

1. Christou I, Felberg RA, Demchuk AM, et al. Intravenous tissue 

plasminogen activator and flow improvement in acute 

ischemic stroke patients with internal carotid artery occlusion. 

J Neuroimag 2002;12:199-123 

2. Wang H, Lanzino G, Fraser K, et al. Urgent endovascular 

treatment of acute symptomatic occlusion of the cervical 

internal carotid artery. J Neurosurg 2003;99:972-977 

KEY WORDS: Stroke, thrombolysis, stenting 


MR Imaging-Guided Thrombolysis of Acute Stroke in a 

Canine Model 

Khawar, S. · Shaibani, A. · Schirf, B. · Parkinson, R. J. · Shin, 

W. · Cashen, T. · Carroll, T. J. 


Chicago, IL 

PURPOSE 

We have developed a canine model of embolic stroke by the 

catheter-directed injection of autologous blood clots into the 

ICA that is reversible via IA rt-PA administration. We have 

tested the hypothesis that MR imaging can be used successfully 

to monitor and guide thrombolysis in real-time. 

Furthermore we have validated a method for measuring 

quantitative CBF with MR imaging (qCBF), in a setting of 

acute stroke by comparing qCBF with fluorescent microsphere 

CBF measurements. 


A series of adult dogs were instrumented with 6 F femoral 

arterial sheaths placed bilaterally. Under X-ray guidance, 

one catheter was placed in the ICA, the other in the left atrium. 

Autologous emboli, prepared by mixing arterial blood 

and thrombin, were injected into the ICA via the microcatheter. 

X-ray DSA confirmed vessel occlusion. 

Flourescent microspheres were injected before and after clot 

injection and post rt-PA infusion to validate perfusion 

changes. MR imaging was performed with IA injections of 

Gd-DTPA, on a whole body 3 T MR system. Quantitative 

CBF (qCBF), diffusion-weighted imaging, and dynamic 

MRA (3D TRICKS) acquisitions were acquired before, during, 

and after rt-PA administration to document stroke evolution 

and clot lysis. Quantitative CBF measurements were 

acquired with a novel imaging technique which utilizes T1 

changes resulting from the contrast injection. A 2 mg bolus 

injection of rt-PA preceded a 6 mg infusion (45 min). Stroke 

resolution was confirmed with MRA and MR perfusion 

images. Fluorescent microsphere measurements served as a 

reference standard for validation of MR imaging derived 

qCBF values. 

RESULTS 

Strokes were produced in all animals embolized. 

Quantitative CBF measurements, diffusion-weighted imaging, 

microspheres and TTC staining confirmed the presence 

81 

of the stroke. IA rt-PA successfully reestablished blood flow 

(Fig ). The average CBF measured with MR imaging (53.83 

+/-25.23 ml/100 gm-min) was in agreement with microsphere 

CBF (50.59 +/-23.02 ml/100 gm-min). In embolized 

animals, the MR imaging qCBF difference between and 

infarcted and contralateral hemispheres was significant (p < 

0.05) and confirmed by microspheres (Table). 

Cerebral Blood Flow (ml/100g-min) 

Normal Hemisphere Infarcted Hemisphere 

qCBF (MRI) 64.08 +/- 16.12 36.96 +/- 21.52 

Microspheres (reference) 67.50 +/- 19.01 39.94 +/- 23.32 

CONCLUSION 

Our initial results indicate we can reliably create and lyse 

embolic strokes in conjunction with MR imaging. Average 

MR imaging qCBF values correlate well with gold standard 

microsphere flow values. 

KEY WORDS: Stroke thrombolysis, animal model, quantitative 

MR perfusion 


Recanalization Versus Perfusion in Acute Stroke 

Revascularization Therapy 

Neff, J. · Khatri, P. · Khoury, J. · Tomsick, T. A. · The IMS 

Study Group 

University of Cincinnati Medical Center 


PURPOSE 

Previous intracranial revascularization reports and studies 

have used “recanalization” and “reperfusion” interchangeably. 

However, recanalization of the primary arterial occlusive 

lesion (AOL) may not reflect the level of global reperfusion 

achieved. Reperfusion may not fully describe the 

local effect of treatment, or risk of reocclusion. Drugs, 

devices, or combinations, may differ in ability to achieve primary 

recanalization and distal reperfusion. We hypothesize 

that evaluating both recanalization and reperfusion is important 

in evaluating drug/device flow restoration. 


We reviewed angiograms from the Interventional 

Management Study (IMS I) IV/IA rtPA group (n = 61). AOL 

recanalization scores and the TIMI reperfusion scores were 

determined for each patient. The AOL Recanalization Score 

was defined as: 0 = no recanalization of the primary occlusion, 

1 = incomplete or partial recanalization of the primary 

Tuesday

Tuesday 

occlusion with no distal flow, 2 = incomplete or partial 

recanalization of the primary occlusion with distal flow, 3 = 

complete recanalization of the primary occlusion with distal 

flow. The TIMI Global Reperfusion Score was defined as: 0 

= no perfusion, 1 = perfusion past the initial occlusion, but 

no distal branch filling, 2 = perfusion and incomplete or slow 

distal branch filling, and 3 = full perfusion with filling of all 

distal branches. We compared the two scores to one another 

to evaluate score agreement. We also compared the scores to 

clinical outcome. We defined good outcome as modified 

Rankin Score (mRS) 0-2. 

RESULTS 

Arterial occlusive lesion and TIMI scores are listed by score 

grade and good outcome in Table 1. Only 22% of patients 

with complete AOL recanalization had complete reperfusion. 

Arterial occlusive lesion scores and TIMI scores 

showed only modest agreement (Kappa = 0.30, 95% CI 

0.16-0.44). Both TIMI perfusion scores of 2-3 (p = 0.02) and 

AOL Recanalization Scores of 2-3 (p = 0.05) predicted good 

outcome by chi-square analysis. 

Table 1. 

AOL TIMI 

Median (25th-75th % ile) 3 (1-3) 2 (1-2) 

Mean 

Grade (total) 

2.0 1.5 

0 12 (19.7%) 12 (19.7%) 

1 6 (9.8%) 16 (26.2%) 

2 11 (18%) 26 (42.6%) 

3 

Grade (with MRS 0-2) 

32 (52.5%) 7 (11.5%) 

0 3 (12%) 3 (12%) 

1 1 (4%) 4 (16%) 

2 6 (24%) 14 (56%) 

3 15 (60%) 4 (16%) 

CONCLUSION 

Both the AOL Recanalization Score and TIMI Reperfusion 

Score can independently predict good clinical outcome with 

the IV/IA rtPA treatment. However, the TIMI Reperfusion 

Score is more predictive. Other treatment paradigms may 

show different relationships between recanalization and 

reperfusion. Applying both techniques may be useful in 

comparing different intracranial revascularization therapies. 

KEY WORDS: Thrombolysis, stroke, rTPA 


Utility of Transcranial Doppler Monitoring during Wada 

Testing and Potential Applications to Therapeutic 

Endovascular Procedures 

Morris, P. · Tegeler, C. · O’Donovan, C. A. · Meads, D. · 

Kim, J. · Whitlow, C. 



PURPOSE 

To evaluate the utility of transcranial Doppler (TCD) monitoring 

during Wada testing in the examination of effects on 

cerebrovascular physiology of intraarterial amobarbital 

injections. 

82 


Simultaneous real-time continuous TCD monitoring was 

conducted in the middle cerebral arteries (MCA) of 23 

patients undergoing bilateral internal carotid (ICA) angiography 

and intraarterial amytal injections for Wada testing. 

Realtime audio feedback via speakers was relayed to the 

operators in the angiography suite. Continuous streaming of 

data from bilateral 2 mHz transducers held in place by a head 

frame was recorded digitally and correlated with clinical 

events during the procedures. 

RESULTS 

Following intraarterial amobarbital injection into the ICA, 

TCD demonstrated a consistent and substantial fall in peak 

mean velocities in MCA flow enduring for approximately 2 

minutes. Gradual reversal to preinjection baseline then 

occurred over 2-5 minutes more, correlating with clinical 

recovery of the patient. Moreover, the MCA tracings demonstrated 

a substantial degree of possibly microembolic events 

in the intracranial circulation during routine angiographic 

maneuvers such as catheter flushing, wire manipulation, and 

contrast injection. Transcranial Doppler tracings were sensitive 

to even small changes in regulation of the continuous 

flush system used routinely, suggesting that even small 

changes in blood viscosity were detectable on TCD. 

CONCLUSION 

1. The transient clinical deficit evident in patients following 

intraarterial amytal correlated with a marked reduction in ipsilateral 

hemispheric blood flow demonstrated by TCD. 2. 

Presumed microembolic events that are clinically silent may be 

common during angiographic procedures. 3. Application of the 

observation of transient intracranial effects of amobarbital has 

applications to endovascular procedures where temporary suppression 

of the brain’s physiologic requirements for blood flow 

might be desirable. Three interventional cases in which this 

desired effects was utilized will be discussed. 4. Transcranial 

Doppler has potential to be a powerful adjunctive tool in the 

execution of safe endovascular procedures. 

KEY WORDS: Wada, transcranial Doppler, blood flow 


Angioplasty of Intracranial Stenosis: Long-Term Clinical 

Outcome 

Marks, M. P. 1 · Wojak, J. C. 2 · Al-Ali, F. 3 · Jayaraman, M. V. 1 

· Marcellus, M. L. 1 · Connors, J. J. 4 · Do, H. M. 1 

1 2 Stanford University Medical Center, Stanford, CA, Our 

Lady of Lourdes Regional Medical Center, Lafayette, LA, 

3 4 Borgess Medical Center, Kalamazoo, MI, St. Joseph’s 

Hospital, Tampa, FL 

PURPOSE 

A recent randomized trial for symptomatic intracranial 

stenosis has suggested there is a high stroke rate with medical 

therapy alone (1). This multicenter study reports the 

long-term clinical results of intracranial angioplasty in the 

treatment of this disease. 


One hundred twenty patients with fixed symptomatic intracranial 

stenoses had 137 intracranial stenoses treated by primary 

angioplasty. There were 84 male and 36 female patients.

Patient’s age ranged from 31-82 years (mean 62.3 + 12.2). All 

patients had stenoses of > 50% attributable to neurologic 

symptoms (stroke and/or TIA) while on antiplatelet or anticoagulant 

medication. Angiograms were evaluated pre and 

postangioplasty for the degree of stenosis and the presence of 

an intimal flap indicating dissection. Patients were assigned a 

pretreatment Rankin score and Rankin score for the last clinical 

follow up. The 137 lesions were seen at the following locations: 

internal carotid artery 45 (38%), MCA 25 (21%), vertebral 

43 (36%), basilar 22 (l8%), and PCA 2 (2%). 

RESULTS 

Pretreatment stenoses varied from 50-99% (mean 81.5 + 

10.5). Posttreatment stenoses varied from 0-95 % (mean 

36.06 + 20.3.) There were 3 (2.5%) periprocedural (within 

30 days of the procedure) strokes and 3 (2.5%) periprocedural 

deaths giving a combined stroke and death rate of 5%. 

An intimal flap was seen in 25 patients (20.8%) and at follow-up 

imaging 20 (80%) had healed spontaneously, 1 (4%) 

was stable, and 2 (8%) showed occlusion of the treated vessel. 

One hundred seventeen patients were available for follow 

up which varied from 2 months to 120 months (mean 

41.5 months). During this period of time 7 patients had a 

stroke in the territory of treatment and 5 additional patients 

had strokes in territories other than the territory of treatment. 

Including the periprocedural stroke and death rate this yielded 

a stroke rate of 3.19/100 patient-years in the territory of 

treatment and a total stroke rate of 4.41/l00 patient-years. In 

the study cohort there were 95 patients available with greater 

than 12 months follow up (mean 50.1 months). In this group 

there was a stroke rate of 3.2/100 patient-years in the territory 

appropriate to the area of treatment. Eighty-five patients 

(71%) had Rankin scores of 0-2 pretreatment. At the time of 

last follow up 103 patients (89%) had Rankin scores of 0-2. 

Seven patients (5.8%) showed worsening in their Rankin 

score. Fifty-four patients (45.0%) were unchanged and 59 

patients (49.2%) were improved. 

CONCLUSION 

Intracranial angioplasty can be performed with a high degree 

of technical success and a low risk of complications. Longterm 

follow up following intracranial angioplasty suggests 

there is a significant reduction in the risk of future strokes 

when compared with available data for patients receiving 

medical therapy. 

REFERENCES 

1. Chimowitz M, Lynn M, Howlett-Smith H, et al. Warfarinaspirin 

symptomatic intracranial disease (WASID) trial: 

Final results. Stroke 2004;35:235 

KEY WORDS: Angioplasty, atherosclerosis, intracranial 

stenosis 


Protected Stent-Assisted Carotid Angioplasty in 

NASCET-ACAS Noneligible Patients 

Gomori, J. M. · Cohen, J. E. 

Hadassah Hebrew University Medical Center 

Jerusalem, ISRAEL 

PURPOSE 

The most feared complication of carotid angioplasty is distal 

83 

emboli of dislodged plaque debris. Our purpose is to assess 

the technical feasibility and clinical results of stent-assisted 

carotid angioplasty using cerebral protection devices in high 

risk patients. 


The patient population was composed of 36 patients; 22 men 

and 14 women, with ages ranging from 60-85 years (mean 

73). Selection criteria were NASCET-ACAS exclusion criteria 

and high risk for CEA under general anesthesia. All the 

patients presented at least 2 vascular risk factors. Seventyeight 

percent of the lesions were considered symptomatic. 

Before the procedure, cranial CT/MR imaging, cervical US 

and DSA of supraaortic vessels and cerebral parenchymography 

were performed. Stenoses were evaluated according to 

NASCET criteria: Mean stenosis: 79%. Calcific plaque: 

65%. Ulcerated plaque: 60%. Associated lesions in supraaortic 

vessels: 45%. Mean lesion length: 21 mm. 

RESULTS 

The balloon/filter guidewire was able to successfully cross 

the lesion in all the cases with no procedural neurologic 

complication. In all the cases the immediate angiographic 

result demonstrated residual stenosis of less than 20%. 

Operative morbidity: bradychardia, 7 cases (19%). All 

patients were followed clinically for more than 6 months. 

Postoperative morbidity: reperfusion hemorrhage, 1 case; 

myocardial infarction (4 months after) 1 case. Postoperative 

mortality: 0%. 

CONCLUSION 

Cerebral protection either with a filter device or a balloon is 

technically feasible in most of the cases. Clinical results 

compare favorably with unprotected angioplasty series. 

Further investigations are necessary to determine whether 

the use of the cerebral protection device will improve the 

results of carotid angioplasty. 

KEY WORDS: Carotid, stent, protection device 


Deployment of a Self-Expanding Stent Alone May Be 

Sufficient to Treat Most Patients with Severe Carotid 

Bifurcation Stenosis 

Bussiere, M. 1 · Lownie, S. P. 1 · Khan, V. 1 · Lee, D. 1 · Gulka, 

I. 1 · Kalapos, P. 1 · Men, S. 2 · Pelz, D. M. 1 

1University of Western Ontario, London, ON, CANADA, 

2Hacettepe University, Ankara, TURKEY 

PURPOSE 

Cerebral embolism is a major potential complication of 

carotid angioplasty and stenting. Techniques that limit the 

number of manipulations and procedure time should 

decrease this procedural risk. We previously have demonstrated 

that deployment of a self-expanding stent alone could 

gradually dilate a severely stenosed carotid artery without 

deliberate use of balloon angioplasty. We wished to determine 

how often this “stent only” approach could be used to 

successfully treat severe carotid stenosis and what factors 

would preclude such an approach and require concurrent 

angioplasty. 

Tuesday

Tuesday 


Over 5 years, 60 consecutive patients with symptomatic 

severe carotid bifurcation stenosis, considered high risk for 

carotid endarterectomy, were treated with the initial intention 

of using a “stent only” approach. Preoperative and 

immediate postoperative NASCET stenosis were assessed at 

the time of angiography. Duplex ultrasonography at baseline 

and regular intervals was employed to follow changes in 

peak systolic velocity (PSV) and ICA/CCA ratio. 

RESULTS 

Seventy-four percent of patients were male with an average 

age of 72 years (50-85 years). Patients were followed for a 

mean of 1 year (2 days to 44 months). Forty-nine arteries 

were treated with stent alone, 11 required balloon predilatation 

and poststenting angioplasty was required in 4 patients. 

Cerebral protection devices were not used. Mean preoperative 

NASCET stenosis was 79% (50-99) in the stent only 

group and 90% (75-99) in the combined treatment group. 

Two patients (4%) had near occlusions in the stent-alone 

treatment group, compared to 5 patients (42%) in the group 

requiring both angioplasty and stenting. Overall, the average 

NASCET stenosis was reduced to 42% (0-73). Preoperative 

mean PSV was 457 cm/s (167-1038) with an ICA/CCA ratio 

of 8 (3.6-18.7). At longest follow up, mean PSV fell to 163 

cm/s (0-623) with an ICA/CCA ratio of 2 (0-9). Two perioperative 

ipsilateral ischemic strokes occurred in the stent-only 

group. Seven deaths occurred during follow up, all unrelated 

to treatment of the carotid stenosis. Restenosis occurred in 7 

patients, with 3 arteries progressing to occlusion. Five of 

these patients had received stents alone and 2 combined 

angioplasty and stenting. Factors preventing successful use 

of the “stent only” approach were: severity of the stenosis 

and degree of calcification of the surrounding plaque (11 

patients), intraluminal thrombosis (2 patients) and restenosis 

(5 patients). Further results from ongoing data collection and 

analysis will be presented. 

CONCLUSION 

Deployment of a self-expanding stent alone was possible in 

80% (49/60) of severe carotid artery stenoses with successful 

short- and long-term treatment achieved in 86% (42/49). 

Carotid arteries with very severe stenoses and heavy calcification 

may require balloon angioplasty along with stent 

placement. 

KEY WORDS: Carotid stenosis, stent, angioplasty 


Stenting of the Vertebral Artery Origin Stenosis: 

Diffusion-Weighted Imaging Findings 

Arat, A. · Canyigit, M. · Cil, B. · Saatci, I. · Cekirge, S. · 

Balkanci, F. 



PURPOSE 

Although angioplasty with/without stenting of extracranial 

vertebral artery (VA) stenoses can be performed with a high 

technical success rate and a low complication rate (1), the 

benefit of this treatment remains unproven in the absence of 

data regarding the natural history of atherosclerotic VA 

stenosis (2), the etiology and prevention of posterior circula- 

84 

tion ischemia (3) and the risk of distal embolization during 

VA stenting (4). We retrospectively evaluated our experience 

with VA stenting in an effort to determine the risk of distal 

embolization associated with the procedure. 


Between June 2000 and November 2004, 29 patients with 

stenting of the origin of VA for atherosclerotic disease in our 

institution were identified. There were no permanent neurologic 

sequela in any of these patients related to VA stenting. 

Eleven of these patients had undergone MR imaging of the 

brain with diffusion-weighted imaging immediately before 

and after (+/- 2 days) of the stenting procedure. 

RESULTS 

Diffusion-weighted imaging demonstrated clinically silent 

punctate emboli in the posterior circulation in 4 of the 11 

patients (12 vertebral stents) with a 33% risk of clinically 

silent distal emboli. One of these patients who had simultaneous 

placement of bilateral carotid stents with distal protection 

had additional bilateral minute hemispheric emboli. 

Another patient with simultaneous stenting of the right common 

carotid artery had a left hemispheric punctate lesion. 

CONCLUSION 

Stenting of vertebral artery origin stenosis may be associated 

with a significant risk of distal embolization. 

Angiography, placement of the guiding catheter, crossing of 

the lesion or inflation of the stent balloon may singly or in 

combination be the cause of distal embolization. Use of distal 

protection devices (5) may be considered for VA stenting 

in patients with clear indications for VA stenting. 

REFERENCES 

1. Wehman JC, Hanel RA, Guidot CA, Guterman LR, Hopkins 

LN. Atherosclerotic occlusive extracranial vertebral artery 

disease: indications for intervention, endovascular techniques, 

short-term and long-term results. J Interv Cardiol 

2004;17:219-232 

2. Janssens E, Leclerc X, Gautier C, Godefroy O, Koussa M, 

Henon H, Lucas C, Leys D, Pruvo JP. Percutaneous transluminal 

angioplasty of proximal vertebral artery stenosis: 

long-term clinical follow-up of 16 consecutive patients. 

Neuroradiology 2004;46:81-84 

3. Jenkins JS, Subramanian R. Endovascular treatment for vertebrobasilar 

insufficiency. Curr Treat Options Cardiovasc 

Med 2002;4:385-391 

4. Jaeger HJ, Mathias KD, Drescher R, Hauth E, Bockisch G, 

Demirel E, Gissler HM. Diffusion-weighted MR imaging 

after angioplasty or angioplasty plus stenting of arteries 

supplying the brain. AJNR Am J Neuroradiol 2001;22:1251- 

1259 

5. Mintz EP, Gruberg L, Kouperberg E, Beyar R. Vertebral 

artery stenting using distal emboli protection and transcranial 

Doppler. Catheter Cardiovasc Interv 2004;61:12-15 

KEY WORDS: Vertebral artery, stents, diffusion-weighted 

imaging


Embolectomy by the Combination of Retrieval and Filter 

Devices: A Novel Technique for Hyperacute Embolic 

Stroke: Experimental Modeling Results 

Suzuki, Y. · Fujitsuka, M. · Chaloupka, J. C. 

University of Iowa Hospitals and Clinics 

Iowa City, IA 

PURPOSE 

Recanalization rates of chemical thrombolysis for acute 

embolic stroke can be relatively low. Supplemental techniques 

and technologies have been tried also with variable 

and inconsistent results. This motivated us to develop techniques 

that may definitively extract an embolized clot using 

already commercially available mechanical devices. 


Life-sized silicone models of the carotid system were created 

by the lost wax technique. A blood clot, laden with tantalum 

powder was used as the embolic material and delivered 

at carotid terminus. All techniques were perfomed under Xray 

fluoroscopy, DSA, and pulsatile flow conditions. A second 

“hybrid” model, consisting of silicone vasculature 

inlayed with ex vivo harvested swine arteries was used also. 

Withdrawn 

The endovascular system consisted of a retrieval device (In- 

Time), a filter device (Spider), and a flow control guiding 

catheter (PATLIVE). The retrieval procedure is illustrated by 

a serious of figures (Figs 1-7). The balloon-tipped guiding 

catheter (PATLIVE) was introduced to the common carotid 

artery and inflated; subsequently an In-Time retrieval device 

was introduced and placed into the clot. The In-Time tip was 

shaped like a spindle basket and pulled gently to proximal; 

after closing basket, In-Time is extracted. The capture procedure 

consisted of placement of a folded Spider coaxially 

through the guiding catheter while also aspirating. The capture 

wire was delivered and the balloon deflated, allowing 

flow to resume and trapping the captured clots into the filter. 

The Spider was then pulled into PATLIVE and the entire unit 

extracted. 

RESULTS 

The study with silicone models was performed eight times 

(four by radiolucent clot, four by radiopaque clot) and all of 

the clots were extracted successfully. In two cases, it was 

identified by fluoroscopy or direct view that the clot came 

free from basket on the way and they were retrieved again. 

The study with hybrid models was successful in 6/8 cases. 

The two failed cases were due to fragmentation of radiolucent 

clots that could not be identified easily. In-Time locked 

owing to the excessive expansion of basket in two studies 

but it did not extract vascular intima. The capacity of Spider 

was sufficient and it could capture all clots. 

85 

CONCLUSION 

The above-described maneuver shows promise to effectively 

accomplish such an objective, and may change the strategy 

for endovascular treatment of acute thromboembolic 

stroke. Further refinements and new developments in microcatheter 

retrieval devices will be needed to translate this 

approach into effective clinical practice. 

KEY WORDS: Embolectomy, stroke, device 


In Vivo Thrombogenicity of Embolic Protection System 

for Angioplasty and Stenting 

Guilbert, F. · Metcalfe, A. · Leblanc, P. · Weill, A. · Roy, D. · 

Raymond, J. 

Centre Hospitalier Universitaire Montréal, Notre Dame 

Hospital 


PURPOSE 

Despite the increasing use of embolic protection systems for 

carotid stenting, the intrinsic in vivo thrombogenicity of 

such devices remains unknown. In order to quantify what 

may represent a potential hazard, we conducted a comparison 

study of different devices commercially available. 


Pools of platelet and fibrinogen from 4 pigs were labelled 

respectively with In-111 and I-125. Animals were anticoagulated 

with 10 000U of unfractioned heparin but no 

antiplatelet regiment was administered. The experiment was 

conducted in 2 stages. The first part concerned validation of 

radioactive labelling using stainless steel, hydrophilic, and 

damaged hydrophilic-coated guide wires. These were either 

intact, damaged, or covered with polaxamer and were compared 

pairwise after introduction for 1 minute in the abdominal 

aorta from a bifemoral approach (n = 32 in 2 animals). 

Radioactivity then was counted. The second stage was 

undertaken to evaluate the thrombogenicity of 3 different 

types of embolic protection systems (n = 24; 12 nitinol mesh 

nets, 6 fabric membrane nets, 6 fabric membrane umbrellas). 

Tuesday

Tuesday 

In a paired fashion, devices were deployed for 1 minute into 

the carotid artery segment dedicated to each type of system. 

The experiment was repeated 3 times under normal conditions 

and 3 times under flow arrest by temporary carotid 

occlusion. Platelet and fibrin deposition on the devices as 

well as the carotid segment harvested after animal sacrifice, 

were counted and reported as a ratio to a coated stainless 

steel guide wire standard measurement. Devices also were 

photographed macroscopically and the amount of clot deposition 

scored using a qualitative scale. 

RESULTS 

1) Relative platelet adhesion and fibrin deposition were 9 

times and 18 times more abundant on stainless steel wires 

than coated wires respectively (P = .05 and P = .005 respectively). 

Damaged coated wires showed more platelets and 

fibrin than intact counterparts (P = .07 and P = .08). 

Polaxamer coating of stainless steel wires led to a reduction 

in platelet and fibrin deposition (P = .06; P = .10). Thus the 

experimental model provides outcome measures that seem 

biologically meaningful. 2) Embolic protection system of 

fabric membrane nets, under normal flow conditions, were 

5-6 and 15-16 times more thrombogenic (for both platelet (P 

= .04) and fibrin (P = .007)) than nitinol mesh nets. 3) 

Testing under flow arrest conditions led to a marked 

increased in the thrombogenicity of nitinol mesh nets 

(platelets P = .05; fibrin P = .03), but unchanged thrombogenicity 

of fabric membrane nets or umbrellas. 4) Clot deposition, 

as expressed by qualitative scores of stereophotographs, 

are more abundant in flow arrest as compared to 

normal flow conditions (P = .018 for all devices). 5) Fibrin 

and platelet deposition on assigned carotid artery segments 

were measured in decreasing amounts, from the presence of 

fabric membrane nets, to fabric membrane umbrellas, to nitinol 

mesh nets. 

CONCLUSION 

Protection devices have an intrinsic thrombogenicity which 

varies with materials and designs. Fabric-manufactured systems 

appeared more thrombogenic than metallic devices in 

normal flow conditions in this model. 

KEY WORDS: Thrombogenicity, devices, experimental models 


Variations in Hounsfield Values of the Normal Basilar 

Artery: The Difficult Assessment of an Hyperdense 

Basilar Artery 

Wallace, M. A. · Branstetter, B. F. 



PURPOSE 

To assess whether the CT criteria utilized in evaluating a 

hyperdense middle cerebral artery sign (HMCAS) applies to 

the basilar artery. 


The Hounsfield units (HU) of the basilar artery in 100 

patients who underwent an unenhanced CT head study were 

determined by 2 mm region of interest (ROI) measurements. 

Patients were selected retrospectively on criteria that includ- 

86 

ed an unenhanced CT of the head being obtained on a patient 

less than 50 years old where basilar insufficiency was not 

clinically suspected. Furthermore, all patients had an angiographic 

study by CT (CTA), MR or conventional intraarterial 

digital subtraction technique (DSA) on the same day as the 

CT study documenting the absence of thrombus in the basilar 

artery. 

RESULTS 

In our sample of 100 patients with an average age of 33 

years, we demonstrated the average basilar HU to be 38.7. 

Twenty-two percent of the patients had basilar HUs above 

previously published criteria indicating thrombus in the middle 

cerebral artery. There was no correlation between patient 

age and the measured HU as determined by regression analysis. 

The measured values showed relatively large variance 

with a standard deviation of 6.2 HU. 

CONCLUSION 

The objective and subjective criteria used to determine 

intraarterial thrombus as described by the HMCAS cannot be 

applied to the basilar artery. Our study indicates that normal 

basilar arteries demonstrate great variations of attenuation 

values making a hyperdense artery sign less useful in this 

region of cerebral blood flow. 

KEY WORDS: Hounsfield, basilar, thrombosis


10:00 AM - 11:44 AM 

Room 107 

(26c) INTERVENTIONAL: 

Angioplasty, Stenting, and New 

Techniques 





Stenting 


Techniques 

Moderators: Jacques E. Dion, MD 

Donald W. Larsen, MD 


Are Protection Devices Necessary during Carotid 

Stenting? 

Joseph, G. J. · Bell, T. 

Presbyterian Hospital 

Charlotte, NC 

Embolization of plaque debris is a potential complicating 

factor in treating carotid vascular disease with angioplasty 

and stent placement. A host of devices have been developed 

to prevent this occurrence. They are costly and add time and 

additional potential for plaque disturbance with their deployment. 

Do they provide enough protection of the distal cerebral 

vasculature to warrant their use? We have performed 

over 400 carotid artery angioplasty/stent procedures in 7 

years at two institutes. The vast majority have been done 

without the use of distal protection devices. The complication 

rate without these devices compares favorably to and in 

many cases is less than reported rates in trials using the 

devices. We will describe our experience both with and without 

these devices and review our experience and outcomes. 

A review of the devices and available literature on their benefit/utility 

will be covered. 

KEY WORDS: Embolization, angioplasty, carotid artery stent 

This paper was selected by the Southeastern Neuroradiology 

Society (SENRS) to receive the Best Paper Award at its 28th 

Annual Meeting held October 20-24, 2004. 

87 


Procedural Complications of Carotid Angioplasty and 

Stenting without Cerebral Protection Devices 

Kadkhodayan, Y. · Derdeyn, C. P. · Cross, D. T. · Moran, C. J. 

Washington University School of Medicine 

St. Louis, MO 

PURPOSE 

Embolic stroke is the primary complication of carotid angioplasty 

and stenting. In an effort to reduce this complication, 

cerebral protection devices have been developed. Based primarily 

on the results of the SAPPHIRE (1) trial performed 

without a medical treatment cohort or an unprotected angioplasty 

group, some have recommended that carotid angioplasty 

and stenting only be performed with protection 

devices. Whether the risks inherent in the use of these 

devices outweigh their benefits has not been established. The 

goal of the present study is to report our experience with 

carotid angioplasty and stenting without the use of cerebral 

protection devices. 


From March 1996 to December 2003, 167 carotid angioplasty 

and/or stenting procedures were performed without 

cerebral protection devices in 152 patients (57 women, 95 

men, mean age 64 years, range 19 to 92 years). Seven of 

these patients had angioplasty alone. Eighty-nine of the 

patients presented with focal neurologic symptoms. The 

study population was heterogeneous with reasons for referral 

(based on NASCET exclusion criteria) including surgically 

inaccessible lesions (n = 62), medical comorbidities (n 

= 24), prior ipsilateral endarterectomy (n = 72), radiation 

therapy to the neck (n = 16), elderly patients (age > 79; n = 

19), or nonatherosclerotic carotid disease (e.g., dissections, 

pseudoaneurysms; n = 22). The patients’ medical records 

were reviewed for clinical characteristics, techniques used, 

and resulting procedural and 30-day complication rates. 

RESULTS 

The procedural stroke rate was 4 of 167 (2.4%). This included 

3 hemispheric strokes and 1 retinal embolus. All events 

occurred in patients with symptomatic atherosclerotic stenosis. 

These patients are compared with atherosclerotic stenosis 

patients without procedural ischemic complications in the 

table. The procedural transient ischemic attack rate was 8 of 

167 (4.8%). The rate of procedural asymptomatic angiographic 

abnormalities was 7 of 167 (4.2%). The rate of nonneurologic 

complications was 6 of 167 (3.6%). At 30 days 

postprocedure, the symptomatic ischemic complication rate 

was 5 of 160 (3.1% of available follow up, 3 transient, 2 permanent). 

The rate of asymptomatic angiographic abnormalities 

was 0.6% (1 treated vessel occluded, asymptomatic). 

The rate of nonneurologicl complications was 2.5%. This 

included 2 myocardial infarctions (1 death). 

Tuesday

Tuesday 

Comparison of atherosclerotic stenosis patients with and 

without procedural stroke 

Clinical characteristic Patients with Patients without ischemic 

stroke, n=4 (%) complications, n=122 (%) 

Symptomatic on presentation 4 (100.0) 69 (56.6) 

Surgically inaccessible lesion 0 46 (37.7) 

—High cervical lesion 36 

—Lesion at common carotid origin 5 

—Failed CEA 5 

Medical comorbidity 0 24 (19.7) 

Prior ipsilateral CEA 1 (25.0) 66 (54.1) 

— Symptomatic on presentation 1 25 

— Asymptomatic 0 41 

Radiation therapy to the neck 2 (50.0) 14 (11.5) 

Elderly (age > 79) 1 (25.0) 16 (13.1) 

CONCLUSION 

In the peri-procedural period (up to 30 days), the cumulative 

incidence of stroke, myocardial infarction, or death was 

4.8% in those who were randomized to receive a stent in 

SAPPHIRE. The limitations of a retrospective study 

notwithstanding, this rate was comparable in the present 

series: 8 of 160 (5.0%). None of the asymptomatic patients 

had a stroke. Unprotected angioplasty and stenting has been 

safe and effective in this heterogeneous patient population, 

particularly in patients without symptomatic stenosis. 

REFERENCES 

1. Yadav JS, Wholey MH, Kuntz RE, et al. Protected carotid 

artery stenting versus endarterectomy in high-risk patients. 

N Engl J Med 2004;351:1493-1501 

KEY WORDS: Carotid artery stenosis, angioplasty, stents 


Cognitive Changes after Carotid Artery Stenting 

Grunwald, I. Q. 1 · Supprian, T. 2 · Struffert, T. 1 · Falkai, P. 2 · 

Krick, C. 1 · Reith, W. 1 

1Clinic for Diagnostic and Interventional Neuroradiology, 

2 Homburg, GERMANY, Clinic for Psychiatry and 

Psychotherapy, Homburg, GERMANY 

PURPOSE 

The aim was to test for changes in cognitive performance 

after carotid artery stenting. 


Ten patients were neuropsychologically tested at least 24 

hours before and 48 hours after carotid artery stenting. To 

diminish thromboembolic events we used a proximal protection 

device. The following neuropsychologic tests were 

selected: MMSE, symbol digit test and subtests of the 

CERAD battery (verbal fluency test, constructional practice, 

word list memory, delayed recall). Affective state was determined 

by the BECK Depression Inventory. 

RESULTS 

None of the patients suffered from depression (BDI < 1) or 

dementia (MMSE 29.9 +/- 1.5). Nine of 10 patients (p = 

0.12) showed increased speed in ZVT (corresponding to trail 

making test). Most patients showed better or at least similar 

results concerning delayed recall (p = 0.31). No change was 

observed concerning digit symbol test, word list memory, 

88 

and verbal fluency and constructional practice. Discussion: 

Better results concerning ZVT and delayed recall after 

carotid stenting might be due to improved brain perfusion. 

CONCLUSION 

After carotid stenting cognitive and memory performance 

seems to improve. As this is, to our knowledge, the first 

study comparing neuropsychological performance after 

stenting further studies with different time intervals and 

more refined testing are needed. 

KEY WORDS: Carotid artery, stenting, neuropsychologic 


Drug-Eluting Stent Therapy for Intracranial and 

Vertebral Origin Atheromatous Stenosis 

Nesbit, G. · Petersen, B. · Lutsep, H. · Clark, W. · Barnwell, S. 

Oregon Health & Science University 

Portland, OR 

PURPOSE 

Intracranial (IC) angioplasty with or without supportive stent 

therapy has been performed with appropriate safety and 

early efficacy; however, a significant restenosis rate of 35% 

may hamper long-term results (1). The restenosis rate in the 

coronary vessels has been similar, but has been reduced significantly 

with the use of drug-eluting stents (DES). We 

evaluated our initial experience and mid-term results of 

intracranial and vertebral angioplasty for atherosclerosis 

using DES. 


Eight patients with symptomatic IC and vertebral stenosis of 

70% or greater, not responsive to medical therapy (anticoagulation 

or dual antiplatelet), have been treated with intracranial 

angioplasty and stent support using DES. Four Cypher 

(Cordis Corp., Miami, FL) and 4 Taxus (Boston Scientific 

Corp., Natick, MA) stents, approved for coronary use were 

used in the treatment of these eight patients. The patients 

were monitored clinically during their hospitalization for the 

procedure and in follow-up evaluation at 1, 3, and 6 months 

by an independent neurologist (HL, WC). Follow-up angiography 

is performed at 6 months irrespective of symptoms to 

determine asymptomatic restenosis. 

RESULTS 

Although these devices are stiffer than bare (non-DES) 

stents, technical success was achieved in all patients with a 

reduction of the preprocedural mean stenosis of 85% to 5%. 

This may be due in part to patient selection. There were no 

peri-procedural complications, including TIA, stroke, or 

hemorrhage. One patient had a mild stroke at 5 weeks poststenting 

with good recovery due to subacute stent thrombosis;however 

she was only on Plavix because of an aspirin 

allergy. All other patients have remained asymptomatic at 2 

to12 (mean 8) month follow up, currently. Follow-up 

angiography has been obtained in 6 patients with no stenosis 

in 4, 10% restenosis in one and the symptomatic occlusion in 

one.

CONCLUSION 

These initial clinical results are very encouraging with similar 

safety and clinical efficacy results compared to the 

SSYLVIA trial (1). The restenosis rate appears to be 

decreased; however larger series or trials will be necessary. 

Strict antiplatelet therapy is necessary to prevent subacute 

complications similar to the larger coronary series. 

REFERENCES 

1. Lutsep H. for the SSYLVIA Investigators. Stenting of symptomatic 

atherosclerotic lesions in the vertebral or intracranial 

arteries (SSYLVIA): study results. Stroke 2004;35(6):1388- 

1392 

KEY WORDS: Drug-eluting stent, intracranial stenosis, 

angioplasty 


Initial Experience with Neurovascular Placement of 

Drug-Eluting Coronary Stents 

Barr, J. D. · Dixit, S. · Morris, D. · Broadbent, P. · Singel, S. 

A. · Arthur, A. 

Baptist Memorial Hospitals 

Memphis, TN 

PURPOSE 

Eight patients with atherosclerotic stenoses of the cervical or 

intracranial arteries were treated by placement of drug-eluting 

stents. 

Materials & Methods 

Drug-eluting coronary stents were used to treat eight patients 

with atherosclerotic stenoses of the vertebral artery origin (N 

= 4), intracranial vertebral artery (N = 2), and the basilar 

artery (N = 2). Seven patients had symptoms referable to the 

lesions treated. One patient was asymptomatic. Six patients 

were treated with Taxus TM Express 2TM (Boston Scientific, 

Natick, MA) stents, and two patients with Cypher TM (Cordis, 

Miami, FL) stents. 

RESULTS 

The stents were placed successfully in all patients without 

technical or clinical complications. Less than 25% residual 

stenosis was achieved in all cases. Clinical follow up (mean 

5.5 months, range 0-14 months) has revealed no new neurologic 

symptoms referable to the posterior circulation in any 

patient. No clinical findings suggestive of an adverse reaction 

to either sirolimus (Cypher TM ) or paclitaxel (Taxus TM 

Express 2TM ) were noted. Six-month posttreatment angiography 

is planned for all patients. Two patients have completed 

angiographic follow up. Follow-up angiographic evaluation 

is pending for the remaining patients. 

CONCLUSION 

Our preliminary results suggest that placement of drug-eluting 

stents within the neurovasculature is safe. Restenosis of 

vertebral artery origin lesions treated by angioplasty or stenting 

and restenosis of intracranial arteries treated by angioplasty 

occurs frequently. Restenosis of intracranial arteries 

treated by bare-metal stents is also problematic. Drug-eluting 

stents have proven highly effective at reducing restenosis 

within the coronary arteries. Angiographic follow up 

89 

pending for our patients will be necessary to determine if 

these stents have similar benefits when used in the neurovasculature. 

KEY WORDS: Stent, angioplasty 


Treatment of Intracranial Atherosclerotic Stenoses with 

Balloon Dilatation and Subsequent Self-Expanding Stent 

Deployment 

Henkes, H. 1,2 · Reinartz, J. 2 · Miloslavski, E. 2 · Lowens, S. 2 · 

Liebig, T. 1 · Kuehne, D. 1 

1 2 Alfried Krupp Krankenhaus, Essen, GERMANY, Robert 

Janker Klinik, Bonn, GERMANY 

PURPOSE 

The endovascular treatment of atherosclerotic intracranial 

arterial stenoses has been based previously on balloon dilatation 

or the deployment of a balloon expandable stent. Both 

methods have advantages (balloon: flexibility; balloon 

expandable stent: high radial force) and drawbacks (balloon: 

risk of elastic recoil and dissection; balloon expandable 

stent: limited flexibility, risk of injury to the vessel due to 

excessive straightening, overexpansion at ends of stent). In 

this study, a new combination of balloon dilatation, followed 

by the deployment of a self-expanding microstent has been 

used. 


A total of 21 patients with atherosclerotic intracranial arterial 

stenoses, symptomatic despite medical treatment, were 

enrolled. Balloon angioplasty (Maverick PTA catheter) was 

followed by the deployment of a Neuroform (6 patients) or a 

Wingspan (15 patients) nitinol stent. 

RESULTS 

An anatomically and clinically adequate result was achieved 

in all patients. The initial degree of stenosis was 72% 

(mean). Balloon dilatation resulted in an average residual 

stenosis of 54% (mean), reduced further to a mean of 38% 

after stent deployment. Arterial dissection, occlusion of the 

target artery or symptomatic distal emboli was not encountered. 

In one patient, a side branch occlusion occurred after 

dilatation of a M1 stenosis, with complete neurologic recovery. 

All patients were either stable or improved 4 weeks after 

the treatment. Recurrent TIA did not occur in any patient. 

Angiographic follow-up results are pending. 

CONCLUSION 

Balloon dilatation and subsequent deployment of a selfexpandable 

stent for the treatment of symptomatic intracranial 

arterial stenoses combines the advantages of both techniques 

and allows a rapid, clinically effective, and technically 

safe treatment of these frequently challenging lesions. 

KEY WORDS: Intracranial stenting, intracranial angioplasty, 

self-expandable stent 

Tuesday

Tuesday 


PTFE Stent-Graft Therapy for Intracranial Dissecting 

Pseudoaneurysms Using the Jomed Balloon-Expandable 

Coronary Stent Graft 

Nesbit, G. · Petersen, B. · Roberts, W. · Barnwell, S. 


Portland, OR 

PURPOSE 

Intracranial dissecting pseudoaneurysms carry a very high 

risk for subsequent hemorrhage and mortality, and also carry 

significant risk for surgical or endovascular therapy. Stentgraft 

therapy may provide the ideal therapy in select patients; 

however most devices cannot be navigated into the cerebral 

circulation. We evaluated our initial results and midterm 

results of intracranial stent-graft treatment of intracranial 

pseudoaneurysms using the coronary Jomed stent-graft 

device. 


Eight patients with symptomatic intracranial pseudoaneurysms 

were treated using Jomed coronary stent grafts 

(Abbott Laboratories, Abbott Park, IL), approved under a 

humanitarian device exemption for coronary perforations, 

using compassionate-use criteria. Patients were selected for 

this therapy based on the risk of surgical or endovascular 

therapy and favorable anatomy for navigation of the device. 

Patients were treated with intraprocedural anticoagulation 

and procedural and follow-up dual antiplatelet therapy, and 

clinical follow-up evaluation during the hospitalization with 

follow-up angiography in select patients. 

RESULTS 

The device was deployed successfully in all patients with no 

peri-procedural complications, including TIA or rehemorrhage. 

Occlusion of the pseudoaneurysm was obtained within 

24 hours in all patients. One patient had mild worsening 

symptoms at 24 hours due to SAH-associated vasospasm in 

the middle cerebral artery distal to the stent, which responded 

to angioplasty. One patient, initially presenting with TIA, 

developed asymptomatic stent thrombosis, with false lumen 

recanalization, and remains asymptomatic. All other patients 

have remained asymptomatic, with 2 to 30 (mean 23) month 

follow up. Follow-up angiography, beyond 6 months has 

been obtained in 4 patients with one occlusion with false 

lumen filling mentioned above, no stenosis or recanalization 

of the pseudoaneurysm in the other three. 

CONCLUSION 

These initial clinical results are very encouraging with 

appropriate safety and clinical and angiographic efficacy 

results. This current device is now under review for an 

intracranial aneurysm humanitarian device exemption indication, 

but it carries some limitations due to its relatively 

stiff design. Given these encouraging results, devices 

designed for intracranial navigation should be persued and 

may provide even better clinical results in this select group 

of patients. 

KEY WORDS: Pseudoaneurysm, stent-graft, intracranial 

90 


Stent Neuroform TM : Two Years of Experience in 54 

Patients with 58 Aneurysms 

Branca, V. · Isalberti, M. · Nuzzi, N. P. · Cinnante, C. · Sina, 

C. · Avignone, S. · Costa, A. 

Ospedale Maggiore Policlinico Milano 

Milan, ITALY 

PURPOSE 

To report our experience with self-expanding stent 

Neuroform TM since October 2002 until November 2004. 


We treated 55 patients, with 2 failures, using 65 stents 

Neuroform TM (37 of those were Neuroform 2 TM ). In 42 

patients with singular wide-neck aneurysm we released one 

stent for each patient. After stent deployment we immediately 

performed the aneurysm embolization with coils (in 17 

cases we used Matrix TM detachable coils), except for two 

patients with in-stent thrombotic phenomena. In two patients 

with basilar tip aneurysm we used two stents intersected to 

reconstruct the anatomy of the vessels and Guglielmi detachable 

coils (GDC TM ) for the aneurysmal embolization. In six 

patients with dissecting aneurysm we reconstructed the 

artery with Neuroform TM ; two of them were treated also with 

GDC TM . Five patients had giant aneurysms: one is still in 

treatment; in one we posed four stents partially superimposed, 

in the others we employed stents and GDC TM . All 

patients underwent aspirin therapy, 300 mg/day. One week 

after the endovascular procedure we performed an angiographic 

control in 47 patients. Later follow up, after a period 

included from 2 months to 2 years, was done in 36 patients. 

RESULTS 

One stent was misplaced. In one patient, with dissecting 

aneurysm, we occluded the vessel because the coils became 

entangled in the stent. Two patients had in-stent thrombotic 

phenomena, resolved with urokinase, immediately after 

stenting. In another patient, with giant aneurysm, thrombotic 

phenomena occurred the day after the procedure due to 

extension of the intraaneurysmal thrombosis. Three patients 

had a little subarachnoid hemorrhage, without neurologic 

complications, during micro guide-wire exchange. No early 

or late neurologic events were observed. Aneurysmal recurrence 

was observed in 5 patients (two patients with unorganized 

intraaneurysmal thrombus, two with dissected 

aneurysm, and one patient treated with Onyx TM ). We reached 

steady occlusion in 31 patients. 

CONCLUSION 

We consider stent Neuroform TM a good device in the treatment 

of intracranial aneurysms with large or without neck. 

We observed much progress in the delivery from 

Neuroform TM to Neuroform 2 TM , with less difficulty in 

deployment. 

KEY WORDS: Vincenzo, branca, neuroform


Endovascular Reconstruction for Treatment of 

Intracranial Aneurysms by a Novel Self-Expanding LEO 

Stent 

Lylyk, P. · Haas, L. · Ferrario, A. · Miranda, C. · Musacchio, 

A. · Pabon, B. 

ENERI 

Buenos Aires, ARGENTINA 

PURPOSE 

Endovascular treatment for cerebral aneurysms has become 

widespread. Many medical devices have been used and original 

devices are being introduced constantly. We herein 

report our initial experience in treatment of wide-necked 

intracranial aneurysms using a new LEO nitinol self-expanding 

stent (BALT extrusion, Montmorency, France). 


Twenty-six patients, aged 15 to 80 years, with complex widenecked 

intracranial aneurysms were selected to be treated with 

a combined approach that consisted of delivery of a flexible 

new self-expanding neurovascular stent through a Vasco (BALT 

extrusion, Montmorency, France) microcatheter to cover the 

neck of the aneurysm and subsequently fill the lesion with coils 

through the stent mesh in some cases. Aneurysms were located 

within the internal carotid artery (15), basilar artery (5), vertebral 

artery (4), and media cerebral artery (1). Patient records 

were analyzed for clinical presentation, location, size, type of 

vascular defect, preprocedural regimen of antiplatelet agents, 

devices used, procedure-related complications, and adverse 

events. In all patients, 72 hours before the procedure, the administration 

of ASA and clopidogrel was started. The strategies consisted 

of: stent deployment to cover the neck of the aneurysm 

without posterior coil embolization, combined approach with 

subsequent filling of the sac with coils through the stent and, 

coil embolization with further stent placement, and stent in 

stent. Clinical and radiographic outcomes were registered. 

RESULTS 

As regards clinical presentation, there was mass effect in 13 

cases, SAH in 8 cases and incidental in 4. Segmental defect 

was documented in all of them. Stent placement covering the 

neck was feasible in 96.1%, so that a total of 25 LEO stents 

were implanted. Procedure-related morbimortality was 12% 

(thrombo-embolic events) with better angiography and clinical 

results with administered glycoprotein IIb-IIIa inhibitors. 

CONCLUSION 

In our experience, LEO self-expanding stent is an extremely 

flexible device and technically is easy to deploy so that it can 

be maneuvered readily and safely through severely tortuous 

vessels, thus enabling the treatment of intracranial side-wall 

wide-necked aneurysms. The combination of endovascular 

reconstruction of the parent vessel by means of a stent followed 

by coil embolization offers a promising therapeutic 

alternative for lesions not able to coil embolization alone. Is 

important in all patients preprocedural regimen of 

antiplatelet agents 72 hours before procedure. Although 

immediate angiographic results are promising, long-term 

angiographic and clinical follow up is required. 

KEY WORDS: Segmental defect, intracranial aneurysms, 

self-expanding stent 

91 


Focal Blood-Brain Barrier Disruption by 260 KHz 

Ultrasound: A Method for Targeted Drug Delivery 

McDannold, N. · Vykhodtseva, N. · Raymond, S. · Jolesz, F. 

· Sheikov, N. · Hynynen, K. 

Harvard Medical School and Brigham and Women’s 

Hospital 

Boston, MA 

PURPOSE 

Previously, we showed that the blood-brain barrier (BBB) 

can be locally disrupted by pulsed focused ultrasound sonication 

(frequencies 0.7 and 1.6 MHz) at low pressure amplitude 

levels if an ultrasound contrast agent containing preformed 

gas bubbles is injected beforehand. Here, we hypothesize 

that the BBB disruption also can be induced at lower 

frequencies that penetrate the skull better, do not suffer significant 

wave distortion while propagating through the skull, 

and produce a larger focal spot that is often desirable for targeted 

drug delivery or molecular imaging. 


We used a focused ultrasound transducer with a frequency of 

0.26 MHz (diameter/radius of curvature 10/8 cm) to sonicate 

rabbit brains after injecting Optison intravenously. Both sonications 

through the intact skull and through a craniotomy 

were performed. The sonications were done under MR 

image (1.5 T or 3 T) guidance. Burst lengths of 1 and 10 ms 

were investigated with a repetition rate of 1Hz and duration 

of sonication of 20 s. The pressure amplitude was varied 

between 0.14 and 0.9 MPa (acoustic power during the burst 

0.09-4.3 W). Blood-brain barrier opening was measured by 

following the MR image signal intensity over time after 

injecting an MR imaging contrast agent that does not normally 

penetrate through the BBB. The animals (34 rabbits) 

were sacrificed at 4-6 hours or 28 days after sonication, and 

the brain was harvested for histology. An additional set of 

rabbits (4) was sacrificed at 1-60 minutes after sonication to 

perform an electron microscopy (EM) evaluation. The 

results showed that focal BBB disruption occurred at a rarefactional 

pressure amplitude of 0.2 MPa (estimated in the 

brain) in some cases (20%). At 0.3 MPa, about 90% of the 

locations were disrupted with 100% of the locations disrupted 

at 0.4 MPa. The histology did not show brain necrosis or 

damage in the power range tested. However, there were 

small extravasations of erythrocytes in 60% of the locations 

sonicated at the pressure amplitudes of 0.4 MPa, but none at 

lower pressures. The EM findings indicated that the ultrasound 

caused transport through the tight junctions and also 

induced transport through vacuoles. The survival experiments 

(9 animals) showed that BBB repair began shortly 

after sonication. It was reduced but open after 3 hours and 

completely closed at 5 hours. The animals sacrificed at 28 

days did not display any noticeable effects. In these cases, 

the average enhancement at the sonicated locations in imaging 

at 28 days was in control locations in the contralateral 

side of the brain. 

RESULTS 

These results demonstrate that the low frequency sonications 

could induce focal BBB disruption at pressure amplitudes and 

power values lower than what was found at higher frequencies. 

Similarly, there was less tissue damage at the higher 

Tuesday

Tuesday 

power values, and the exposure conditions could be selected 

such that the disruption was induced without even the extravasation 

of erythrocytes that were shown to be present at higher 

frequencies. 

CONCLUSION 

This is an important result for focal drug delivery into the 

brain. 

KEY WORDS: Ultrasound, blood-brain barrier, MR imaging 


3D Roadmap in Neuroangiography: Technique and 

Clinical Interest 

Söderman, M. 1 · Babic, D. 2 · Andersson, T. 1 

1Karolinska University Hospital, Stockholm, SWEDEN, 

2Philips Medical Systems, Best, THE NETHERLANDS 

PURPOSE 

We present the first clinical results with a novel technique, 

“3D roadmap,” based on rotational angiography with the fluoroscopic 

image as an overlay. 


The 3D-roadmap technique is basically conventional unsubtracted 

digital fluoroscopy with an overlay of image data 

extracted from a previous rotational angiography. There is 

feedback within the imaging system so that images from the 

rotational angiography accurately reproduces the actual 

positions of the C-arm, the X-ray tube, and the image intensifier 

as well as the field-of-view. Consequently, the principal 

advantage of 3D roadmapping is that it allows an unlimited 

number of changes in the position of the X-ray equipment 

and the magnification, using only one injection of contrast 

medium. Since the 3D roadmap is created in a separate 

workstation, it does not interfere with conventional fluoroscopy, 

roadmapping, or image acquisition. The reconstructed 

volume image is displayed on an in-room monitor 

where it is manipulated with an optical Bluetooth mouse 

wrapped in sterile transparent plastic. Brightness, contrast, 

and opacity can be adjusted individually for the fluoroscopy 

and the overlay. The head of the patient is immobilized during 

the entire procedure by a specially made head support. 

The 3D roadmap has been used now during 30 interventional 

procedures, mostly endovascular treatments of intracranial 

aneurysms. 

RESULTS 

The quality of the fluoroscopy has not been quite as good as 

in the angio system itself, but sufficient enough to clearly 

visualize a microguidewire. In several patients with tortuous 

vascular anatomy the 3D roadmap allowed for multiple 

changes in incidence and magnification during catheterization, 

using only the inital rotational angiography as a 

roadmap. This saved time and contrast media. For patient 

with straightforward anatomy the gain was less evident. An 

observation was that for aneurysm treatment monoplane was 

in almost all cases sufficient, and the procedures were performed 

with the lateral plane parked. 

92 

CONCLUSION 

In a clinical setting the 3D roadmap often facilitates intravascular 

neuronavigation, with reduction of procedure time and 

contrast media usage. A potential problem might be that 

because of the limited number of contrast injections a thromboembolic 

complication may go undetected for a significant 

time. 

KEY WORDS: Cerebral angiography, roadmap, interventional 

neuroradiology 


High-Resolution Brain Imaging Using an 

Intraparenchymal Loopless RF Antenna 

Phillips, M. D. 1 · Karmarker, P. 2 · Baker, K. B. 1 · Viohl, I. 3 · 

Lowe, M. J. 1 · Steiner, C. 1 · Nyenhuis, J. 4 · Rezai, A. R. 1 · 

Bottomley, P. A. 2 

1 2 The Cleveland Clinic Foundation, Cleveland, OH, The 

Johns Hopkins University, Baltimore, MD, 3Rentendo, Milwaukee, WI, 4Purdue University, West Lafayette, IN 

PURPOSE 

To study the safety and efficacy of MR imaging-guided 

placement of an RF loopless antenna into brain parenchyma 

using the subthalamic nucleus as a target to produce highresolution 

imaging of deep brain structures. The overall goal 

is to develop a guidance and delivery system for MR imaging 

interactively guided minimally invasive neurosurgical 

procedures. 


A cannula and microelectrode system made out of nitinol 

were arranged to form a loopless RF antenna. The design 

was patterned after previously developed intravascular 

catheter-based MR imaging coils. The cannula and coil are 

tested for imaging and safety performance on Siemens 

Allegra® 3 T scanner. Safety testing was performed using 

fiber-optic temperature probes and a previously described 

gel phantom. Imaging protocols were deemed safe if they 

produced less than 2°C of heating. MR imaging-guided 

placement procedures and high-resolution imaging capabilities 

of the cannula microelectrode RF loopless antenna system 

were studied using a macaque animal model. Using a 

frameless stereotactic system mounted to the anesthetized 

macaques calvarium the loopless RF system was advanced 

via a burr hole into the brain to the superior aspect of the subthalamic 

nucleus under interactive MR imaging guidance. 

Images were acquired using the transmit/receive head coil as 

well as high-resolution images using the loopless RF antenna 

system. Images from the loopless antenna were intensity 

corrected to adjust for local intensity differences. 

RESULTS 

Heating tests demonstrated less than 1°C of heating for all 

image sequences tested. Interactive MR imaging-guided 

methodology allowed for placement of the loopless RF 

antenna system at the superior aspect of the subthalamic 

nucleus in a single pass. Figure 1 demonstrates an image 

acquired from both the head coil system (1A) and an intensity 

corrected image from the intraparenchymal loopless RF 

antenna system (1B). The distal aspect of the RF antenna is 

that the the superior aspect of the STN (white arrow).

CONCLUSION 

The results demonstrate the first images ever acquired from 

an intraparenchymal MR imaging coil system. Further, they 

demonstrate that the systems can be placed safely and efficiently 

into small deep brain targets in an MR imaging interactively 

guided fashion. We believe that this methodology 

will have a broad range of applications for minimally invasive 

neurosurgical procedures. 

KEY WORDS: Loopless rf antenna 


Endoscopic Visualization of the Lateral and Third 

Ventricles via Intraspinal Navigation 

Giles, B. 1 · Fujimoto, T. 2 · Replogle, R. 1 · Fujimoto, H. 3 · 

Miller, S. 1 · Purdy, P. 1 

1University of Texas Southwestern Medical Center, Dallas, 

TX, 2Kobe University Medical School, Kobe, JAPAN, 

3Okayama University Medical School, Okayama, JAPAN 

PURPOSE 

Percutaneous intraspinal navigation (PIN) is a percutaneous 

approach to the cerebral subarachnoid space from a lumbar 

puncture (1) and can be used to access multiple spinal and 

cerebral structures. This study compares fiberscopic (fibersoptic) 

versus videoscopic (CCD chip on tip of scope) visualization 

of lateral and third ventricular structures. 


An unembalmed male cadaver was obtained and transported 

to our research angiography suite and placed in the prone 

position. Access to the lumbar subarachnoid space was 

achieved via lumbar puncture and a 2.8 mm outer diameter 

(OD) tip deflecting fiberscope with a 1.2 mm working channel 

was introduced into the subarachnoid space through a 9 

French sheath. The endoscope was advanced cranially under 

endoscopic and X-ray fluoroscopic guidance. The fiberscope 

was removed and a 3.8 mm OD tip-deflecting videoscope 

with a 1.2 mm working channel then was introduced into the 

subarachnoid space via a partial second lumbar laminectomy 

secondary to the inability to place an adequately sized sheath 

percutaneously and advanced cranially to the interpeduncular 

cistern. Endoscopic images were stored using a direct 

digital capture apparatus. The third ventricle was entered via 

direct puncture in the interpeduncular cistern. The lateral 

ventricle was entered from the third ventricle via the foramen 

of Munro. 

93 

RESULTS 

The floor of the third ventricle was visualized, and both 

endoscopes were able to advance through the floor of the 

third ventricle. Endoscopic position was confirmed with Xray 

fluoroscopy. The foramen of Monro was visualized and 

both scopes were directed through the foramen of Monro 

into the lateral ventricle and endoscopic position was confirmed 

again with fluoroscopy. Inside the lateral ventricle 

choroid plexus and surface features of the lateral ventricle 

were identified (Fig.). 

CONCLUSION 

Accessing ventricular structures via an intraspinal approach 

is technically feasible under endoscopic guidance. The 

videoscope provides a greater degree of detail of the surface 

anatomy and a higher degree of spatial resolution than compared 

to traditional fiberscopes. This approach may provide 

an alternative to traditional surgical approaches to the lateral 

and third ventricles. Reduction of the diameter of the videoscope 

should allow it to be placed percutaneously. The safety 

and efficacy of PIN should be explored. 

REFERENCES 





Neuroradiol 2003;24:361-365 

KEY WORDS: Endoscopy, ventricles, percutaneous 

intraspinal navigation 

Tuesday

Tuesday 


10:00 AM - 11:36 AM 

Room 205 

(26d) ADULT BRAIN: Neoplasms and 






Stenting 



Moderators: James G. Smirniotopoulos, MD 



Grading of Primary Intraaxial Brain Tumors Using 

Perfusion CT 

Schramm, P. 1 · Klotz, E. 2 · Tronnier, V. 1 · Hartmann, M. 1 

1 University of Heidelberg Medical School, Heidelberg, 

GERMANY, 2 Siemens Medical Solutions, Forchheim, 

GERMANY 

PURPOSE 

Perfusion CT allows to quantitatively assess hemodynamic 

characteristics of brain tumors. In this study we investigated 

if different tumor types can be distinguished from each other 

using Patlak analysis of dynamic CT data. We studied to 

which extent cerebral blood volume (CBV) and permeability 

values calculated from this approach allow a reliable grading 

of 4 types of primary intraaxial brain tumors. 


Thirty-five patients with a total of 36 tumors were examined 

with a 40 second dynamic scan after injection of 50 ml of 

iodinated contrast media. Seven patients had low-grade 

glioma/astrocytoma (WHO II), 8 patients had anaplastic 

glioma (WHO III), 13 patients had glioblastoma (WHO IV), 

and 6 patients had intracerebral lymphoma. One patient had 

both low-grade astrocytoma (WHO II) and glioblastoma 

(WHO IV) in the contralateral hemisphere. All tumor types 

were confirmed by histopathology. The CT data were analyzed 

with a commercial implementation of the Patlak 

method (Perfusion CT, Siemens, Erlangen, Germany). 

Tumor regions of interest (ROIs) were drawn in a morphologic 

image and automatically transferred to maps of CBV 

and permeability. As control the contralateral cortex was 

evaluated also after excluding major vessels. Regions of 

interest mean values were calculated and group differences 

were tested for significance using the Wilcoxon test. 

94 

RESULTS 

Normal brain tissue had a CBV of 3.0 ± 0.9 ml/100 ml and a 

permeability (P) of 0.6 ± 0.8 ml/100 ml/min. The corresponding 

values for the tumors were (CBV in ml/100 ml, P in 

ml/100 ml/min): glioma WHO II (2.9 ± 1.8, 1.4 ± 1.3), glioma 

WHO III (4.8 ± 2.0, 4.6 ± 3.2), glioma WHO IV (4.3 ± 2.1, 4.5 

± 3.4), lymphoma (3.2 ± 1.1, 8.4 ± 5.2). Permeability of lowgrade 

glioma was not significantly different from normal brain 

tissue, for all other tumors it was significantly increased (p < 

0.0001). Cerebral blood volume of high-grade glioma was significantly 

higher (p < 0.02) than that of normal tissue, but not 

elevated in low-grade glioma and lymphoma. Permeability in 

high-grade glioma and lymphoma was significantly higher (p 

< 0.01) than in low-grade glioma. Cerebral blood volume in 

high-grade glioma was significantly higher (p < 0.05) than in 

low-grade glioma and in lymphoma. 

CONCLUSION 

Cerebral blood volume and permeability in low-grade 

glioma do not differ from normal brain tissue. High-grade 

glioma can be differentiated from low-grade glioma by a significant 

increase in both CBV and permeability. Lymphoma 

on the other hand can be distinguished from glioma by an 

increased permeability while CBV is normal. Perfusion CT 

allows to reliably classify glioma and lymphoma based on 

quantitative measurements of CBV and permeability using 

perfusion CT. 

KEY WORDS: Perfusion CT, brain tumors, grading 


Evaluation of Quantitative Diffusion-Weighted MR 

Imaging in Anaplastic Gliomas with or without 

Oligodendroglial Components 

Kitajima, M. 1 · Hirai, T. 1 · Yamura, M. 1 · Hayashida, Y. 1 · 

Yamasahita, Y. 1 · Korogi, Y. 2 · Nakamura, H. 1 · Kuratsu, J. 1 

1 2 Kumamoto University, Kumamoto, JAPAN, University of 

Occupational and Environmental Health, School of 

Medicine, Kitakyushu, JAPAN 

PURPOSE 

Prognosis of anaplastic gliomas is usually poor; however, it 

has been noted that existence of the oligodendroglial components 

leads to relatively better response to chemotherapy, 

and anaplastic astrocytomas with oligodendroglial component 

having a better prognosis than pure anaplastic astrocytomas. 

Although characteristic imaging findings of oligodendroglial 

tumors have been reported, there is no precise 

evaluation with diffusion-weighted MR imaging. The purpose 

of this study was to evaluate the ADC values of 

anaplastic gliomas with oligodendrogial components including 

anaplastic oligodendroglioma (AOD) and anaplastic 

oligoastrocytoma (AOA) compared to those of anaplastic 

astrocytoma (AA). 


Thirty-two consecutive patients (19 men and 13 women, 

ranging in age from 19 to 77 years; mean 47.5 years) with 

histologically proven anaplastic gliomas [AOD (n =10), 

AOA (n =14), AA (n = 8)] underwent conventional and echoplanar 

imaging (EPI) diffusion-weighted MR imaging. Five 

regions of interest (ROI) were chosen from each lesion on an 

ADC map. Mean ADC values (ADCmean) of five ROIs

were calculated in each case. Minimum ADC (ADCmin) and 

maximum ADC (ADCmax) values also were chosen among 

five chosen ROIs. Resultant values were normalized to those 

of contralateral uninvolved white matter (nADCmean, 

nADCmin, nADCmax). These normalized ADC (nADC) 

values were correlated with each histologic type. Statistical 

analyses were performed using one factor ANOVA and post 

hoc test. Histologic characteristics in each type, particularly 

in the ratio of nucleus to cytoplasm, were evaluated. 

RESULTS 

The nADCmean and nADCmin of AOD and AOA were 

lower than those of AA; the difference was statistically significant 

(p < 0.05). The nADCmean and nADCmin of AOA 

was between those of AA and AOD. The ratio of nucleus to 

cytoplasm was higher in AOD than that in AA. 

CONCLUSION 

Anaplastic gliomas with oligodendrogial components 

showed lower mean and minimum ADC values than those 

without oligodendrogial components. This information is not 

obtained with conventional MR imaging and is useful for the 

differentiation between these two types of high grade 

gliomas. 

KEY WORDS: Anaplastic glioma, diffusion-weighted MR 

imaging, oligodendroglioma 


Can Proton MR Spectroscopic Imaging Differentiate 

between Neoplastic and Nonneoplastic Brain Lesions in 

Adults? 

Horska, A. 1 · Hourani, R. 1 · Brant, L. J. 2 · Weingart, J. 1 · 

Barker, P. 1 

1 2 The Johns Hopkins University, Baltimore, MD, National 

Institutes of Health/National Institute on Aging/Gerontology 

Research Center, Baltimore, MD 

PURPOSE 

Noninvasive differentiation between neoplastic and nonneoplastic 

brain lesions may sometimes be difficult using conventional 

imaging techniques. The goal of this study was to 

95 

investigate whether proton MR spectroscopic imaging ( 1 H 

MRSI) can aid in differentiating between brain tumors and 

nonneoplastic lesions in adults. 


We retrospectively examined 69 patients with untreated primary 

brain lesions (age = 47.9 ± 14.7 years, 35 men). Thirtysix 

patients had a neuropathologically confirmed brain 

tumor. Other patients were diagnosed with stroke (N = 4), 

demyelination (N = 10), surgically confirmed benign lesion 

(N = 10), and nonpathologically confirmed stable lesions (N 

= 9; 11 months to 10 years follow up, with no evidence of 

lesion progression and clinical deterioration). MR imaging 

and high-resolution, multislice MRSI (1) were performed at 

1.5 T. Spectra were evaluated in the lesion and in the contralateral 

normal-appearing parenchyma. Concentrations of 

N-acetyl aspartate (NAA), creatine (Cr), and choline (Cho) 

were expressed as ratios NAA/Cho, NAA/Cr and Cho/Cr, 

calculated from areas under the respective signals. 

Normalized ratios Cho norm , Cr norm , and NAA norm , 

expressed to the contralateral hemisphere were evaluated 

also. Discriminant function analysis was used to identify 

metabolite ratios that can predict inclusion in the neoplastic 

and nonneoplastic lesion groups. An F-test was performed to 

assess the significance of the discriminant function. Data 

(Fig. 1) are presented as means ± standard deviations. 

RESULTS 

The discriminant function included ratios NAA/Cho, 

Cho norm, NAA norm , and NAA/Cr (presented in the 

decreasing magnitude of correlation with the discriminant 

function, Fig. 1). The discriminant function analysis correctly 

classified 84.2% of original grouped cases (p < 0.0001). 

Five tumor cases were misclassified as nononcologic lesions 

and six nononcologic lesions were classified as tumors. 

CONCLUSION 

The results presented here suggest a promising role for proton 

MR spectroscopy in distinguishing between brain tumors 

and nononcologic lesions. The use of high-resolution MR 

spectroscopic imaging allows the examination of multiple 

regions in the lesions and subsequent selection of a region 

with the most abnormal spectroscopic appearance. In contrast, 

previous studies used single voxel MR spectroscopy 

with a larger voxel size (2-4). However, nononcologic 

demyelinating lesions may sometimes show similar spectra 

to high-grade brain tumors (high Cho, decreased NAA). In 

Tuesday

Tuesday 

these cases, it remains to be determined if combination of 

MRSI with other techniques, such as perfusion MR imaging, 

will further improve diagnostic specificity. 

REFERENCES 

1. Duyn JH, et al. Radiology 1993;188:277-282 

2. Rand SD, et al. AJNR Am J Neuroradiol 1997;18:1695-1704 

3. Bautzen J, et al. AJNR Am J Neuroradiol 2000;21:1213-1219 

4. Moller-Hartmann W, et al. Neuroradiology 2002;44:371-381 

KEY WORDS: Brain lesions, MR spectroscopy 

Dr. Hourani will be the presenter. 


Comparing Perfusion Parameters Obtained from a 

Single Compartment and Pharmacokinetic Modeling 

Methods Obtained from Dynamic Susceptibility 

Contrast-Enhanced Perfusion MR Imaging with Glioma 

Grade 

Law, M. · Young, R. · Sasaki, T. · Rad, M. · Babb, J. · 

Johnson, G. 


New York, NY 

PURPOSE 

Numerous methods for measuring perfusion parameters and 

numerous different parameters can be found in the literature 

for the grading of gliomas. To compare perfusion parameters 

obtained from two different models: 1) Indicator Dilution 

Model which assumes contrast material remains intravascular 

in single compartment; 2) Pharmacokinetic Model (PK), 

a two-compartment model in the grading of primary glial 

neoplasms. Each of these parameters were correlated with 

histopathologically confirmed grade in primary glial neoplasms. 


Seventy-three patients with primary glial neoplasms underwent 

conventional MR imaging, and dynamic, susceptibility 

contrast-enhanced MR imaging (DSC MRI). Relative CBV 

measurements were obtained from regions of maximal 

abnormality relative to the contralateral normal white matter 

as determined from rCBV color maps of each lesion. Vp and 

K trans measurements derived from a pharmacokinetic modeling 

algorithm were obtained in all tumors. Absolute measurements 

of CBF, CBV, and MTT were made using standard 

algorithms and an automated method for obtaining the arterial 

input function. The data from the three methodologies 

then was compared to histopathologic grades (based on a 

three-tiered Ringertz system) determined from specimens 

obtained by volumetric resection or stereotactic biopsy. 

RESULTS 

For each measure, Tukey’s honestly significant difference 

(HSD) procedure was applied to the ranks in order to make 

all pairwise comparisons among the tumor grades while 

maintaining the familywise type I error rate for the set of 

comparisons at or below the 5% level. Assuming that the 

tumor grades are ordered as Low grade glioma < Low Grade 

ODG < Ana Astro < GBM, the following table shows the 

Spearman rank correlation of rCBV, VP, K trans , CBF, CBV, 

96 

and MTT with tumor grade. Tumor grade was significantly 

positively correlated with each of rCBV, VP, CBF and CBV, 

but not with K trans or MTT. Thus, higher grade tumors tended 

to be associated with higher values of rCBV, VP, CBF and 

CBV while neither K trans nor MTT exhibited a significant linear 

association with tumor grade. 

CONCLUSION 

Cerebral blood volume measurements taken relative to contralateral 

white matter demonstrated the best correlation/prediction 

of glioma grade and type. This may relate to rCBV 

partially correcting for recirculation and leakage as well as 

reducing noise compared with perfusion parameters taken 

relative to an arterial input function. VP measurements taken 

from a PK model also showed good correlation with glioma 

grade. 

KEY WORDS: Perfusion, glioma, MR imaging 


Imaging Molecular Markers of Hypoxia in Glioblastoma 

Multiforme with MR Spectroscopic Imaging and 

Dynamic Susceptibility Contrast-Perfusion MR Imaging 

Law, M. · Oh, S. · Livshiz, J. · Johnson, G. · Zagzag, D. 


New York, NY 

PURPOSE 

Hypoxia-inducible factor 1 α (HIF-1α) has long been associated 

with tumor hypoxia. More recently carbonic anhydrase-9 

(CA-9) has been discovered to be a more accurate 

marker for tissue hypoxia and potential glioma infiltration 

than HIF-1α. The purpose of this study is to correlate relative 

cerebral blood volume measurements and lactate/lipid 

measurements from dynamic susceptibility contrast MR 

imaging and high-resolution MR spectroscopy respectively 

with the spatial distribution of these molecular markers of 

hypoxia. 


Twenty patients with a glioblastoma multiforme underwent 

conventional MR imaging, DSC T2*-weighted MR imaging, 

and proton MRS. Relative CBV measurements were 

obtained from regions of maximum perfusion as determined 

from rCBV color maps within the tumoral and the leading 

edge-peritumoral regions of the glioma. Metabolite ratios 

(Cho/Cr, Cho/NAA, lactate/Cr and lipid/Cr ratios) were 

measured using high-resolution 2D spectroscopic imaging at 

an echo time of 144 ms. This then was spatially compared to 

the expression of HIF-1α and CA-9 in the tumoral and peritumoral 

region of the gliomas. 

RESULTS 

Statistically significant differences were demonstrated 

between the tumoral and leading edge-peritumoral regions 

for rCBV, Cho/Cr, Cho/NAA, Lactate/Cr and Lipid/Cr ratios 

(p < 0.0005, < 0.0005, < 0. 005, < 0.0005, < 0.0005 respectively 

). Molecular studies of human gliomas demonstrates 

elevation of HIF-1α and HIF-1β within necrotic regions and 

also within the peritumoral, so called “leading edge” of the 

glioma. Carbonic anhydrase-9 has been shown to be a more 

specific marker for tumor hypoxia but is not expressed in the 

leading edge-peritunoral region. There is increase in rCBV

in the regions of HIF-1α expression in the tumoral, leading 

edge and peritumoral regions. However, lactate/Cr and 

lipid/Cr is elevated only in the tumoral region but not in the 

peritumoral region. 

CONCLUSION 

Relative CBV, lactate and lipid levels were demonstrated to 

be elevated in the tumoral regions of GBMs where there is 

increased expression of HIF-1α and CA-9. However, within 

the peritumoral region, where there is expression of HIF-1α 

but not CA-9, there is elevation of rCBV without elevation 

of lactate and lipids. Relative cerebral blood volume is found 

to correlate with HIF-1α expression whereas lactate/lipid 

levels correlate more accurately with CA-9 expression in 

GBMs. 

KEY WORDS: Human glioma, molecular, perfusion, spectroscopy 


Initial Timing of MR Imaging and Assessment of MR 

Angiography Using Intravenous Superparamagnetic 

Carbohydrate-Coated Iron Oxide Particles in Primary 

High-Grade Brain Tumors and/or Cerebral Metastases 

Neuwelt, E. A. · Manninger, S. P. · Solymosi, D. · Nesbit, G. 

· Jerosch-Herold, M. · Stevens, A. · Hoffman, J. M. 


Portland, OR 

PURPOSE 

MR imaging and angiography with ferumoxytol (carbohydrate-coated 

iron oxide particles) was compared to standard 

gadolinium-enhanced MR imaging to assess the preliminary 

safety and efficacy of ferumoxytol (Code 7228, Advanced 

Magnetics, Cambridge, MA) imaging. This study also is 

designed to define the appropriate imaging time postinjection 

of the agent, the time course of lesion signal intensity, 

the differences in lesion enhancement with the two contrast 

agents, the utility of bolus infusion of ferumoxytol to produce 

better quality MRA and perfusion images, and the optimum 

imaging sequence parameters and concentration of ferumoxytol 

at 1.5 and 3 T. 


Six of the 12 proposed patients done to date with primary 

high-grade glial tumors or brain metastases were randomized 

to either 1.5 T or 3 T magnet for their initial imaging. 

Each patient had baseline scan for comparison without/with 

gadolinium (including MRA and a perfusion study). Twentyfour 

hours later they underwent MRA and perfusion imaging 

using multiple bolus ferumoxytol infusions (a total dose of 4 

mg Fe/kg was administered in different stages and dilutions 

over 1 hour). MR scans then were obtained at 4-6, 16-20, 24- 

28 and 72 hours postinfusion on both 1.5 and 3 T, if possible. 

Some patients had surgery and had intraoperative MR 

imaging (0.1 T magnet) as well. 

RESULTS 

The lesions were seen on both 1.5 T and 3 T images, and 

informative images on intraoperative low field magnets (0.1 

T) were obtained 48 hours after ferumoxytol injection. The 

delayed enhancement pattern was different from that seen 

with gadolinium and was predominantly in the peripheral 

97 

aspects of the tumor. The signal intensity of enhancement 

was maximum at around the 24-28 hours time-point postferumoxytol 

injection then decreased over time, while the volume 

of enhancement slowly increased. Dynamic bolus MRA 

images with ferumoxytol were of good quality. This initial 

experience suggests a lower concentration (1:8 dilution) 

decreases susceptibility artifacts. Because of the long plasma 

half-life and initially intravascular localization of ferumoxytol, 

the enhanced high-resolution TOF MRA resulted in good 

visualization of the tumor vasculature. The postferumoxytol 

perfusion studies were comparable to those made with 

gadolinium. The tumors usually showed decreased mean 

transit time (MTT) compared to gray matter and to the 

healthy white matter but their cerebral blood volume (CBV) 

was larger. 

CONCLUSION 

So far ferumoxytol proved to be safe for bolus administration 

in patients with primary high-grade brain tumors and/or 

cerebral metastases, and no adverse events occurred. MR 

imaging showed a different delayed enhancement pattern, 

mainly peripheral, in comparison to gadolinium, which gradually 

enlarged over 72 hours. This contrast agent appears to 

be very useful in dynamic bolus MR angiography and perfusion 

imaging, and its long plasma half-life can facilitate 

high-resolution MRA images. Additional studies are warranted 

to better define its role in CNS imaging. 

KEY WORDS: Iron oxide contrast agent, MR imaging, brain 

tumors 


Imaging of Hypoxia within the F98 Glioma Model on 

High-Resolution Ultra-High Field Gradient-Echo MR 

Imaging 

Christoforidis, G. A. · Yang, M. · Yange, W. · Barth, R. F. · 

Heverhagen, J. · Schmalbrock, P. · Knopp, M. 

The Ohio State University 

Columbus, OH 

PURPOSE 

To assess the ability to visualize hypoxic regions within the 

F98 rodent glioma model using ultra-high field high-resolution 

MR imaging using a gradient-echo imaging. 


Tumor cells derived from the F98 undifferentiated glioma 

cell line were implanted into the right caudate nucleus in 4 

Fischer rats. The tumor cells were allowed to grow in vivo 

until the animal displayed premorbid signs (weight loss, 

ataxia, and periorbital hemorrhage). At this time the animals 

were anesthetized and ventilated. They then were imaged at 

8 T on a Bruker AVANCE (Bruker, Billerica, MA) interfaced 

with Techron (Crown International, Elkhart, IN) gradient 

amplifiers and Manex gradients (Magnex Scientific, 

Abingdon, England) using a custom built radio-frequency 

front end. A custom made 4 cm diameter birdcage coil was 

tuned to the head of the rat at 340 MHz while the rat was in 

the prone position. Imaging included a gradient recalled 

echo (GRE) sequence (TR/TE 700/16 msec, flip angle 450, 

NEX = 2, 512 x 512 matrix and 78 µm in-plane resolution) 

before and after USPIO (2 mg Fe/kg) administration via the 

femoral vein. The regions within the tumor which displayed 

Tuesday

Tuesday 

hypointense microvasculature were outlined. Two minutes 

prior to sacrifice, 10 mM EF5 (dissolved in 5% glucose) was 

administered with a volume of 1% of the animal mass. 

Following sacrifice, the brain of rat was removed and frozen 

for immunohistochemical (IHC) and histopathologic analyses. 

Immunohistochemical specimens and histopathologic 

specimens were compared to MR images for corresponding 

regions of hypoxia. 

RESULTS 

There was a close relationship between regions identified as 

hypoxic on the IHC stained specimens for hypoxia and 

regions of hypointense microvascularity on 8 T GRE MR 

imaging on all four rodents images. Fig.1a is an 8 T GRE 

MR image of a rodent bearing the F98 glioma. On Fig. 1a the 

arrowheads point to areas of signal loss corresponding to 

hypointense microvasculature. Because signal loss is 

thought to result from higher concentrations of deoxyhemoglobin, 

it is believed that these regions may be hypoxic. The 

IHC stain for hypoxia is superimposed on the 8 T GRE 

image in Fig. 1b and corresponds to the foci of hypointensity 

on 8 T GRE MR imaging. 

CONCLUSION 

Gradient recalled echo 8 T high-resolution MR imaging has 

potential for identifying regions of hypoxia within the the 

F98 glioma model. 

KEY WORDS: MR imaging, glioma, hypoxia 


Modulation of Language Activation by Brain Tumors: 

Results from Standardized Preoperative Functional MR 

Imaging in 57 Consecutive Patients 

Stippich, C. · Rapps, N. · Konrad, F. · Tronnier, V. · Sartor, K. 

University of Heidelberg 

Heidelberg, GERMANY 

PURPOSE 

To assess the effects of brain tumors on language activation 

in Broca’s and Wernicke’s areas and on language lateralization 

using standardized preoperative functional MR imaging 

(fMRI). 


Fifthy-seven strictly right-handed patients with brain tumors 

of the left hemisphere that affected Broca’s area (n = 19) or 

Wernicke’s area (n = 38) had standardized preoperative 

block-designed BOLD fMRI of language function at 1.5 T 

using a covert word generation task (WG) and a covert sentence 

generation task (SG). In each individual a regional lateralization 

index was calculated for Broca’s and Wernicke’s 

area using the formula LI = (LH - RH)/(LH + RH) with L 

and R representing the number of activated voxels in the left 

98 

and right hemisphere at the same statistical threshold. The 

effects of the brain tumors on language-associated brain activation 

were assessed posthoc by statistically correlating the 

LIs of the patients to normative data obtained from 14 righthanded 

controls. FMRI data analysis: BrainVoyager® ; 

Posthoc statistics: Mann-Whitney U-test (SPSS V.11). 

RESULTS 

In Broca’s area WG revealed a significant effect of the brain 

tumors on language lateralization (p = 0.01) with a median 

LI = -0.22 in the patient group compared to a median LI = 

0.97 in the control group. In Wernicke’s area SG indicated 

significant changes (p = 0.007) in the patient group with a 

median LI = 0.5 (median LI = 0.96 in controls). 

CONCLUSION 

Our data support the hypothesis that unilateral brain tumors 

modulate complex language networks in both hemispheres. 

Besides changes in the activation of the directly affected language 

areas (Broca and Wernicke), the activation in the 

anatomical homolog areas of the right hemisphere increased. 

Therefore the right hemisphere seemed to contribute substantially 

to the compensation mechanisms for language 

function in patients with brain tumors. As the measurable 

effect depends also on the applied stimulation task, a standardization 

of preoperative fMRI protocols is warranted. 

KEY WORDS: Functional MR imaging, language, plasticity 


Differentiation of Benign Subtypes of Meningiomas by 

Diffusion Tensor Imaging-Based Calculation of Tensor 

Shape 

Tropine, A. 

Johannes Gutenberg-Universität Mainz 

Mainz, GERMANY 

PURPOSE 

Presurgical knowledge of the elasticity of tumors facilitates 

neurosurgical treatment planning. Fibroblastic meningiomas 

usually are considered to be of a harder consistency during 

operation than other noncalcified benign subtypes of meningiomas. 

The purpose of this study was to assess the potential 

of diffusion tensor imaging (DTI) for the histopathologic differentiation 

of meningiomas by preoperative MR imaging. 


In 17 patients with benign noncalcified meningiomas, DTI 

(b = 900 or b = 1000 s/mm²) was performed, and after calculation 

of diffusion tensors, mean diffusivity (MD) and 

fractional anisotropy (FA) values were calculated and the 

geometrical shape of the tensor was computed as well. All 

examinations were performed on a 1.5 T scanner. 

RESULTS 

Histologic analysis revealed 8 meningothelial, 3 mixed, and 

6 fibroblastic meningiomas. Whereas meningothelial meningiomas 

were represented by spherical tensors corresponding 

to isotropic diffusion, all fibroblastic meningiomas predominantly 

showed disk-shaped tensors indicating in-plane diffusion. 

This may be caused by the fascicular arrangement of 

long, spindle-shaped tumor cells and the high content of 

intra and interfascicular fibers as shown by histology. In

addition, a capsule-like rim of in-plane diffusion surrounded 

most meningiomas irrespective of their histologic type. The 

other DTI parameters did not show significant differences 

between the subtypes. 

CONCLUSION 

Diffusion tensor imaging-based analysis demonstrates a oneto-one 

correlation of the shape of the diffusion tensor and the 

histopathologic type of benign meningiomas. It is a promising 

method to differentiate between fibroblastic and other 

subtypes of benign meningiomas in order to obtain information 

about their “hard” or “soft” consistency prior to 

removal. 

KEY WORDS: Diffusion tensor imaging, tensor shape, 

meningiomas 


Unusual Imaging Appearance and Location of an 

Intracranial Chondroma 

Lin, Q. · Xu, W. 

The Affiliated Hospital of Qingdao Medical College 

Qingdao University 

Qingdao, CHINA 

PURPOSE 

To present a case of surgically resected chondroma, a rare 

tumor in the left parieto-occipital region, and to describe 

unusual imaging appearances, clinical symptoms, and operative 

finds. 


A 25-year-old woman presented to neurosurgical department 

with 4 months of headache and visual defect. Those symptoms 

had become more serious 2 weeks earlier. Physical 

examination revealed papilledema with eyes and defect of 

visual field on the right side of the eyes. Head CT and MR 

imaging were performed. 

RESULTS 

CT imaging revealed an ovalshaped, well defined mass in 

the left parieto-occipital region. The mass was heterogeneous 

with a thick hyperdense peripheral zone, an isodense 

to hypodense central core, and a small part of peripheral 

zone of slight enhancement. MR imaging showed a peripheral 

zone of hypointensity relative to brain on T1- and T2weighted 

imaging, corresponding to hyperdense peripheral 

zone on the CT, and more central core of hypointensity on 

T1-weighted imaging and of hyperintensity on T2-weighted 

imaging. The signal intensity of the central core was more 

similar to CSF on T1- and T2-weighted imaging, but no 

99 

attenuation on FLAIR imaging. At surgery, a gray-pink, 

firm, avascular and extraaxial mass was seen. It loosely 

adhered to the convexity of the overlying brain and compactly 

attached to dura in a part of the mass. The mass was 

resected completely. 

CONCLUSION 

Chondroma is an extraaxial, avascular, circumscribed tumor, 

and rare in intracranial locations except skull base region. 

The tumor was heterogeneous and slight enhancement on CT 

and MR imaging. There are unusual imaging appearances in 

the presented case. 

KEY WORDS: Intracranial chondroma, heterogeneous, imaging 

appearances 


Longitudinal MR Imaging Features in Glioblastoma 

Multiforme Treated with Brachytherapy versus External 

Beam Radiation Therapy 

Aiken, A. H. · Larson, D. · Chang, S. · Cha, S. 



PURPOSE 

MR imaging is the modality of choice for imaging surveillance 

in patients with glioblastoma multiforme (GBM) 

undergoing radiation therapy (RT). Although standard RT 

has been external beam radiation (EBRT), more patients now 

are being treated with higher dose, focal radiation such as 

brachytherapy (brachy). MR imaging patterns of recurrent 

contrast enhancement (CE), representing recurrent tumor or 

radiation necrosis, have not been described in patients treated 

with brachy. The purpose of this retrospective study was 

to compare temporal patterns of recurrent contrast enhancement 

in patients treated with brachy versus EBRT. Several 

distinct patterns of CE were identified and correlated with 

histopathology showing either recurrent tumor or radiation 

injury at reoperation. 


Serial MR imaging scans over at least the first year after 

gross total resection (GTR) and RT were available for review 

in 15 patients treated with brachy (permanent low activity I- 

125 seeds placed at the time of GTR) followed by EBRT and 

20 patients treated with only standard EBRT. The characteristic 

of CE (linear, nodular, feathery, or solid), serial progression 

and location of CE relative to the resection cavity 

were described for each patient. Radiologic recurrence of CE 

was correlated with histopathologic analysis of tissue at the 

time of reoperation. 

Tuesday

Tuesday 

RESULTS 

Fourteen of 20 EBRT patients demonstrated focal nodular 

CE along the resection cavity within 4 months. Twelve of 14 

had presumed clinical and radiographic tumor recurrence, 

which was proven pathologically in 9/12. The remainder of 

the EBRT patients (6/20), who did not progress within 1 year 

follow up, developed either transient linear enhancement or 

no abnormal enhancement. Seven of 15 brachy patients initially 

developed linear rim enhancement within 4 months 

that progressed to feathery CE over the course of 1-2 years. 

While this was interpreted initially as an ominous radiographic 

finding, histopathology in all 7 patients showed 

only radiation necrosis at reoperation. Eight of 15 eventually 

developed more focal nodular contrast enhancement (on a 

background of linear CE in 5/8) along the resection cavity. 

Tumor recurrence was confirmed by pathology in 3 patients, 

PET in 1 patient, and clinical and radiographic progression 

in the remaining 4 patients. 

CONCLUSION 

Our preliminary data suggest that longitudinal MR imaging 

features differ between GBM patients treated with EBRT 

versus brachy. In both groups, nodular enhancement strongly 

suggested tumor recurrence. Feathery enhancement, 

which progressed from linear, rim enhancement around the 

cavity, seen only in brachy patients, strongly indicated radiation 

necrosis. Our data also show that the pattern of radiation 

injury differs after treatment with brachy versus EBRT 

alone. Previous reports of radiation injury after EBRT 

describe lesions remote from the primary tumor or resection 

cavity. All pathologically proven radiation necrosis in our 

brachy patients occurred immediately adjacent to the seedlined 

resection cavity. Linear rim enhancement around the 

resection cavity seen in almost all brachy patients likely represents 

a pattern of radiation injury unique to brachytherapy 

and should not be confused with recurrent tumor. This study 

suggests that close observation of early patterns of contrast 

enhancement may help distinguish radiation necrosis from 

recurrent tumor on subsequent follow-up imaging. 

KEY WORDS: Brachytherapy, glioblastoma multiforme, radiation 

necrosis 


Differences between the Right and Left Corticospinal 

Tracts in Normal Adults Revealed by Quantitative MR 

Imaging 

Reich, D. S. 1 · Wakana, S. 1,2 · Smith, S. A. 1,2 · Jones, C. K. 1,2 · 

Calabresi, P. A. 1 · Mori, S. 1,2 

1 2 The Johns Hopkins University, Baltimore, MD, Kennedy 

Krieger Institute, Baltimore, MD 

PURPOSE 

To quantify the normal variation between the right and left 

corticospinal tracts and to seek a structural correlate of handedness. 

Functional and structural asymmetry in the brain is 

well established. Post-mortem analysis has revealed differences 

between the right and left corticospinal tracts, both in 

the brain and spinal cord. In the medulla, for example, the 

corticospinal tract originating from the left precentral gyrus 

usually decussates at a more rostral level than its counterpart 

on the right (Kertesz and Geschwind 1971). Morphometric 

evidence (Rademacher, et al. 2001) suggests that the corti- 

100 

cospinal tract in the cerebral hemispheres is larger on the 

left, but no definitive correlation of tract size with handedness 

has been demonstrated. 


We imaged the brains of 40 normal adult subjects with diffusion 

tensor MR imaging (DTI) at 1.5 T and 3 T. We also 

obtained T2-weighted images and magnetization-prepared 

rapid gradient-echo images, which were coregistered with 

the DTI images. From the results, we reconstructed the corticospinal 

tracts using the method of fiber association by 

continuous tractography. We also used the DTI color map to 

perform a region-of-interest (ROI) analysis on voxels known 

to contain the pontine and medullary corticospinal tracts. 

Along the reconstructed tracts and within the selected ROIs, 

we calculated tract volume; transverse magnetization relaxation 

time constant (T2); fractional anisotropy (the degree to 

which nearby structures are oriented in parallel); mean diffusivity 

or trace; and diffusion-weighted signal. We quantified 

the differences between the corticospinal tracts using a 

normalized asymmetry measure that could be compared 

across individuals. 

RESULTS 

Normal variation between the right and left corticospinal 

tracts can be demonstrated readily with quantitative DTI and 

fiber tractography. Median asymmetries across subjects were 

on the order of 1% to 4%, with inter-quartile ranges of about 

the same magnitude, for all parameters except tract volume. 

For some individuals and some parameters, particularly trace 

and T2, asymmetries could range up to 40%, but this was 

distinctly uncommon. Asymmetries in tract volume were 

substantially higher (median about 15%), but the accurate 

measurement of tract volume is subject to substantial artifactual 

variation relating to fiber tracking techniques. 

Volume asymmetries were substantially lower in the ROI 

analysis (median 3%). We found no consistent relationship 

between absolute laterality (right vs left) and any of the 

parameters listed above, nor could we demonstrate a statistically 

significant correlation with handedness (p > 0.05 in all 

cases, Fisher’s exact test). 

CONCLUSION 

Structural asymmetries between the right and left corticospinal 

tracts in normal adult subjects are modest for most 

parameters examined, and we could not demonstrate a structural 

basis for handedness in the corticospinal tracts. Our 

results are important because they constitute a normative 

data set that can be used to establish the significance of 

changes in many disease states, including multiple sclerosis 

and stroke, that can affect the corticospinal tracts. 

KEY WORDS: Corticospinal tract, asymmetry, fiber tractography


10:15 AM - 11:45 AM 

Room 103 


PowerPoint 

Tuesday Afternoon 

1:00 PM - 2:30 PM 

Room 107 



(28) Functional MR Imaging (ASFNR) 

(194) Operative Risk Assessment for Neurosurgical 

Patients 

— John L. Ulmer, MD 

(195) Functional MR Imaging Data Analysis for the 

Radiologist: Facts and Artifacts 

— Joseph A. Maldjian, MD 

(196) How Lesions Affect Brain Function: Imaging 

of Cortical Plasticity 

— Andrei I. Holodny, MD 

Moderator: Andrei I. Holodny, MD 

Operative Risk Assessment for Neurosurgical Patients 




able to: 

1) To understand the role of fMRI and DTI in preoperative 

planning. 

2) To appreciate the complimentary nature of fMRI and DTI 

in assessing the risk of surgically induced neurologic 

deficits. 

101 

3) To appreciate the potential for preoperative mapping to 

reduce neurologic complications in neurosurgery patients. 

4) To understand how to integrate imaging data with clinical 

and intraoperative mapping data to achieve an optimal lesion 

resection, while avoiding neurologic deficits. 


Resections of primary brain tumors and other lesions can 

improve the functional performance of neurosurgery 

patients, enhance the effectiveness of adjuvant therapies and 

increase the duration of patient survival, provided that surgically 

induced neurologic deficits can be avoided. In order to 

achieve maximal cytoreductive excision of tumors and successful 

resection of nonlethal lesions while preserving neurologic 

function and the quality of life, a critical balance 

between resection extent and the risk of a postoperative 

deficit must be sought. Preoperative risk assessments, 

including that afforded by blood oxygen level dependent 

(BOLD) functional magnetic resonance imaging (fMRI) of 

eloquent cortex and diffusion tensor imaging (DTI) of white 

matter tracts and fasciculi, can provide important determinants 

of therapeutic planning. However, the information 

afforded by these advanced imaging techniques also must be 

integrated with clinical data and other functional assessments 

to yield an optimal surgical result. This requires 

expertise on the part of the functional neuroradiologist 

beyond a basic understanding of the physical and physiologic 

mechanisms underlying fMRI and DTI. A broader understanding 

of functional anatomy, the functional consequence 

of lesion location, intraoperative functional mapping techniques, 

and the clinical rationale of neurosurgery is required 

to adequately utilize fMRI and DTI in clinical preoperative 

planning scenarios. This session will present strategies for 

integrating functional data, including that derived from 

fMRI and DTI, for successful operative planning and neurologic 

risk assessments for neurosurgical patients. 

Functional MR Imaging Data Analysis for the 

Radiologist: Facts and Artifacts 

Joseph A. Maldjian, MD 



able to: 

1) Review postprocessing steps in functional MR imaging. 

2) Identify the common artifacts encountered in functional 

MR imaging. 

3) Describe the causes of common artifacts in functional MR 

imaging (fMRI). 


Unlike conventional MR imaging in which a displayed voxel 

in an image typically is related directly to the signal detected 

in that voxel, functional MR imaging introduces additional 

postprocessing steps that can influence the appearance 

of the final image. The postprocessing steps involved in generating 

functional MR imaging activation overlaid on 

anatomical images will be reviewed. Each step will be related 

to potential artifacts that can be generated. These steps 

typically include data reconstruction, motion correction, statistical 

inference, and fusion with anatomical images. 

Artifacts also can arise prior to any postprocessing steps. 

Tuesday

Tuesday 

Sources of distortions in raw data images will be reviewed. 

These image distortions can result in errors in spatial localization 

of activation foci, as well as the absence of activation 

in the data sets. Careful evaluation of the preprocessed raw 

images and the final activation maps will be emphasized. 

Sources of artifacts related to normalization, global signal, 

and paradigm delivery/performance also will be described. 

How Lesions Affect Brain Function: Imaging of Cortical 

Plasticity 




able to: 

1) Learn how to incorporate functional MR imaging (fMRI) 

into the clinical setting. 

2) Appreciate the radiologic factors that may limit fMRI. 

3) Understand that brain tumors may lead to cortical reorganization 

and learn how to evaluate this cortical plasticity 

with fMRI. 


Functional MR imaging is an extremely powerful tool which 

has revolutionized the study of the human brain. Aside from 

the basic sciences, fMRI has clear applications to clinical 

medicine. Recent advances will make fMRI studies accessible 

for clinical radiologists, including those in private practice. 

This will bring this commanding technology to the aid 

of our patients. Until recently, it was thought that if a lesion 

destroyed a part of the brain responsible for certain function, 

this function was lost forever. However, recent work in 

patients with strokes, tumors, and arteriovenous malformations 

have shown that the brain actually can reorganize in 

response to a lesion, and that another part of the brain can, to 

some extent, take over the function of the part of the brain 

affected by the lesion. Functional MR imaging has a number 

of known limitations some of which can lead to “pseudoreorganization." 

However, the radiologist, understanding the 

physics of MR imaging and having the capacity to interpret 

the routine MR images that come with the fMRI data, is in a 

unique position to appreciate these limitations and avoid the 

associated pitfalls. In this session, I will point out strategies 

to optimize the performance and interpretation of the fMRI 

exam to separate out true cortical reorganization from artifacts. 

102 


1:00 PM - 2:30 PM 

Room 105/106 

(29) Fundamentals of Medical 

Reimbursement for Diagnostic and 

Therapeutic Neuroradiology (ASITN) 

(197) Mechanism of Reimbursement for Hospitals 

and Physicians 

— J. Arliss Pollock, MD 

(198) Importance and Methodology of the RUC 

Survey Process 

— Andrew Ku, MD 

(199) Update on Neurovascular Procedure 

Reimbursement Initiatives 

— John J. Connors, III, MD 

Discussion 

Moderator: John J. Connors, III, MD 

Mechanism of Reimbursement for Hospitals and 

Physicians 




able to: 

1) Understand the process and mechanism of New Code 

Applications and Reimbursement. 

2) Comprehend the basis of the Physician and Facility 

Payment System. 


New code applications are subject to review by the CPT 

Editorial Panel, Relative Value Update Committee and Center 

for Medicare and Medicaid Services. The function and deliberations 

of each of these entities will be discussed and the responsibilities 

and current operating parameters for each will be presented. 

While this is the process and mechanism for getting a 

new code, the underlying financial, legislative, and administrative 

mandates drive how the process is implemented. The distribution 

of a fixed pool of money to fund services is a challenging 

and formidable task. Various means have been developed 

to try to provide an equitable distribution of payment 

across all of medicine and these will be explained. There are different 

categories that may be assigned to a code and the goal of

achieving Category I designation will be reviewed considering 

current requirements for success. Evidence-based medicine is 

becoming increasingly important to support Category I status 

and the means to satisfy this will be discussed. The categories 

of evidence-based medicine will be presented so that the 

requirements for code support are understood. Reimbursement 

valuation is subject to both the physician survey and a formula 

(IWPUT). The evaluation mechanism that drives physician payment 

will be presented. Strong physician participation in these 

surveys strengthens the application and validates the recommended 

RVUs. Over 90% of the (RUC) Relative Value Update 

Committee’s recommendations are accepted by (CMS) Center 

for Medicare and Medicaid Services. Therefore, all physicians 

having a stake in this need to understand how the system works 

so that they can participate effectively. 

Importance and Methodology of the RUC Survey 

Process 

Andrew Ku, MD 



able to: 

1) Understand how the RUC survey form determines potential 

physician reimbursement. 

2) Illustrate the importance of each individual being surveyed 

filling out the RUC survey form. 

3) Understand preservice time, intraservice time, and postservice 

time. 


The RUC survey document is one of the few times physicians 

are able to directly influence their own financial reimbursement 

for services and procedures. Most often the survey 

is performed when there is a new service. Accuracy in 

filling out the estimates of how much work is involved in a 

given patient treatment is vital to appropriate physician reimbursement. 

Under estimation of the amount of work and time 

for a procedure will result in your getting poor financial 

reimbursement. 

Update on Neurovascular Procedure Reimbursement 

Initiatives 

John J. Connors, III, MD 


There are two categories of Neurovascular procedural reimbursement 

initiatives: CPT codes and DRG. As of now, there 

are only three procedural codes that directly apply to interventional 

neuroradiologic procedures. The two original 

codes are embolization codes: 61624 for intracranial 

embolization and 61626 for extracranial head and neck 

embolization. Current codes for selective angiography with 

supervision and interpretation have been in existence for 

quite some time. A year ago, the addition of 61623 now gives 

neurointerventionists three codes specific to our profession. 

This code is for the distinct procedure of temporary test 

occlusion of the carotid artery for preocclusion testing in 

case sacrifice of the vessel becomes necessary. Last year, a 

103 

new procedural code for carotid stenting was passed. The 

new code has distinction for procedures “with” or “without” 

protection. Just within the last month, we have obtained 3 

new codes. These will cover intracranial angioplasty for atherosclerosis, 

intracranial angioplasty with stenting for atherosclerosis, 

and intracranial angioplasty for vasospasm. In 

addition, there is now discussion under way to raise the DRG 

for procedures involving revascularization therapy of stroke, 

as opposed to conservative management. There will be discussion 

of the implications for the practice of neuroradiology 

and interventional neuroradiology. Currently, there is no 

code for intra-arterial stroke therapy, either with just lytic 

infusion or by use of the MERCI retriever. 

Discussion 


1:00 PM - 2:45 PM 

Theatre 

(30) Head and Neck Sarcoma 

(ASHNR) 

(200) Pathology Principles 

Pathology Principles 

Bruce M. Wenig, MD 

— Bruce M. Wenig, MD 

(201) Imaging of Head and Neck Sacromas 

— Lawrence E. Ginsberg, MD 

(202) Surgical Principles: What We Need to Know 

from Imagers 

— Patrick Gullane, MB, FRCSC, FACS 

(203) Nonsurgical Therapies 

— Brian O’Sullivan, MD, FRCPC, FRCPI 

Moderators: Roy A. Holliday, MD 




able to: 

1) Understand the pathologic parameters utilized in the classification 

and diagnosis of sarcomas. 

Tuesday

Tuesday 

2) Recognize the limitations and pitfalls in the histologic 

grading of sarcomas. 

3) Become familiar with the advantages and disadvantages 

to the clinical staging of sarcomas. 


The role of the pathologist in head and neck sarcomas is to 

establish the correct diagnosis and to determine the pathologic 

parameters that assist in directing treatment and in predicting 

prognosis, including histologic grading and clinical 

staging. The classification of soft tissue tumors is predicated 

on the histogenesis (i.e., cell of origin) of a given soft tissue 

tumor. Virtually all sarcomas may occur in the head and neck 

region, but some sarcomas are more common than others. 

While sarcomas may demonstrate histologic differentiation 

and, therefore, are identifiable by light microscopic analysis, 

adjunct studies often are required in order to establish the 

correct diagnosis. These adjunct studies include histochemistry, 

immunohistochemistry, electron microscopy, cytogenetics, 

and molecular genetics. Histologic diagnosis, grading 

and staging serve as valuable guides to directed treatment 

protocols and provide useful prognostic information. This 

presentation will serve as an overview of the pathologic principles 

utilized in the diagnosis of head and neck sarcomas, 

and will include a discussion on the pitfalls and limitations 

of grading, and the advantages and disadvantages of staging. 

REFERENCES 

1. Greene FL, Page DL, Fleming ID, et al. AJCC Cancer Staging 

Manual. 6th ed. New York, Springer; 2002 

2. Fletcher CDM, Unni KK, Mertens F, eds. World Health 

Organization of Tumours. Pathology & Genetics. Soft 

Tissue Tumours. Lyon, IARC Press; 2002 

3. Weiss SW, Goldblum JR. General considerations. In: Weiss 

SW, Goldblum JR, eds. Enzinger and Weiss’s Soft Tissue 

Sarcomas. St. Louis, Mosby; 2001:1-19 

Imaging of Head and Neck Sarcomas 




able to: 

1) Review common and uncommon head and neck sarcomas. 

2) Describe those imaging findings that suggest the diagnosis 

of sarcoma. 

3) Recognize the many situations in which sarcomas may be 

radiologically indistinguishable from other cancers. 


Sarcomas of the head and neck are fairly common in a busy 

cancer hospital. As with any tumor, the role of radiology is 

to determine if indeed there is a lesion, describe the lesion in 

terms of size, extent, imaging appearance, effect on adjacent 

structures, and finally, suggest a differential diagnosis. 

Differential diagnosis is intentionally placed last, as it is 

arguably the least important goal of imaging. The lesion will 

certainly be biopsied if not resected, and the pathologist gets 

to make the ultimate diagnosis. However, we cannot resist 

the urge to suggest a likely diagnosis. For head and neck sarcomas, 

this is sometimes possible, sometimes not. The pur- 

104 

pose of this presentation is to review those situations in 

which the radiologist can reasonably predict that a tumor is 

a sarcoma, those that although the imaging may be relatively 

nonspecific still suggests the likelihood or possibility of 

sarcoma, and finally, those lesions where the imaging is 

completely nonspecific and the diagnosis of sarcoma is a 

surprise to everyone. A very wide range of common and notso-common 

head and neck sarcomas will be included, 

regardless of our ability to diagnose radiologically. 

Surgical Principles: What We Need to Know from 

Imagers 

Patrick Gullane, MB, FRCSC, FACS 

A native of Galway Ireland, Dr Patrick Gullane received his 

MB degree from Galway University, Ireland in 1970 and 

completed his residency in Otolaryngology-Head and Neck 

Surgery, University of Western Ontario in 1975. He obtained 

his fellowship from the Royal College of Surgeons Canada 

(FRCSC) in 1975 and certified by the American Board of 

Otolaryngology-Head and Neck Surgery in 1976. Dr 

Gullane then pursued advanced training in Head and Neck 

Oncology with Dr Sebastian Arena at the Mercy Hospital, 

Pittsburgh, Dr John Conley, Presbyterian Medical Center, 

New York and Dr John Marquis Converse New York 

University Medical Center Institute of Facial Plastic and 

Reconstructive Surgery, 1975-1977. He has held numerous 

academic positions. In 1983 he was recruited to the 

Department of Otolaryngology-Head and Neck Surgery at 

the University of Toronto. Dr. Gullane’s present academic 

appointments are as follows: 1989-present 

Otolaryngologist-in-Chief, University Health Network; 

1990-present Professor Department of Otolaryngology- 

Head and Neck Surgery, University of Toronto; 1993-present 

Professor, Department of Surgery, University of Toronto; 

1999-present Wharton Chair Holder in Head and Neck 

Surgery, Princess Margaret Hospital; 1999-present 

Professor, Division of Plastic Surgery, Department of 

Surgery, University of Toronto; 2002-present Professor and 

Chair, Department of Otolaryngology-Head and Neck 

Surgery, University of Toronto. Dr Gullane is recognized as 

a leader in the field of Head and Neck Surgery. Over the 

years he has received numerous awards including the 

Honour Award of the American Academy of Otolaryngology- 

Head and Neck Surgery, and the Honour Award of the 

American Academy of Facial Plastic and Reconstructive 

Surgery. In 1990 he received "The Harris P. Mosher Award” 

for his thesis submission from the Triological Society and 

was awarded the “Commemorative Medal for the 125th 

Anniversary of Canadian Confederation” in 1992. From 

1992 to present he has been selected as a member of the 

“Best Doctors in America." Other recognitions include the 

Distinguished Service Award, American Academy of 

Otolaryngology-Head and Neck Surgery in 2003 and the 

Millennium Society Award from the American Academy of 

Otolaryngology-Head and Neck Surgery in 2004. He 

presently serves as President of the American Head and 

Neck Society and President of the North American Skull 

Base Society. He was elected as Vice-President, Eastern 

Section-Triological Society. Dr Gullane is a member of both 

national and international medical societies, and has been 

invited as a visiting Professor to more than 30 countries lec-

turing on all aspects of Head and Neck Oncology. Dr 

Gullane has published 195 papers in peer-reviewed journals, 

53 chapters in textbooks, and has had 8 books published. 

He serves on the editorial board of 10 journals. 



able to: 

1) To determine the role of CT, MR imaging and PET scanning 

in the evaluation of patients with head and neck neoplasms. 

2) To identify patients with surgically respectable tumors. 

3) To evaluate neck metastases and determine whether or not 

neurovascular structures are involved. 


The role of CT/MR imaging and more recently PET scanning 

has helped significantly in the diagnosis and assessment 

of patients with head and neck neoplasms. Emphasis will be 

placed on the utility of the various imaging modalities used 

to stage tumors about the head and neck (TNM system) and 

so therefore helps determine the most appropriate form of 

therapy as per our protocols. Imaging also helps to identify 

patients who are operable candidates. With the advent of 

more surgical resections for advanced skull base tumors, MR 

scanning has helped to determine the intracranial extension 

of the neoplasm and whether or not the cavernous sinus is 

involved, especially in high-grade malignancies. An overview 

of the various imaging modalities utilized in all sites 

about the head and neck region will be reviewed and their 

contributions, limitations and pitfalls discussed. 

REFERENCES 

1. Gullane PJ, Havas T, Keller A, VanNostrand P, Broderac I, 

Shulman H. Clinical and CT correlation of pharyngolaryngeal 

cancer with whole organ serial sections in the axial 

plane - A historadiographic analysis of 60 cases. 

Transactions, American Laryngological Association 

1986;107:149-153 

2. Havas TE, Motbey JA, Gullane PJ. Prevalence of incidental 

abnormalities on computed tomographic scans of the 

paranasal sinuses. Arch Otolaryngol Head Neck Surg 

1988;114: 856-859 

3. Witterick IJ, Gullane PJ, Keller MA. Postoperastive carotid 

body tumor evaluation: Analysis using MR angiography. 

Laryngoscope 1995;105 (7):764-767 

4. Brown DH, Mulholland S, Yoo JHJ, Gullane PJ, Irish JC, 

Neligan PC, Keller A. Internal jugular vein thrombosis following 

modified neck dissection: Implications for head and 

neck flap reconstruction. Head and Neck 1998;20(2):169-174 

Nonsurgical Therapies 

Brian O’Sullivan, MD, FRCPC, FRCPI 

Dr. O’Sullivan is the Bartley-Smith/Wharton Chair and 

Professor in the Department of Radiation Oncology at 

Princess Margaret Hospital at the University of Toronto. He 

received his medical degree from the University College 

Dublin, Ireland in 1976 and completed a residency and fellowship 

training in Dublin and Toronto. A recipient of the 

Award of Honor at RSNA following his delivery of the 2004 

Annual Oration in Radiation Oncology, he combines expert- 

105 

ise in sarcomas and head and neck cancer, and is a past 

president of the Connective Tissue Oncology Society. He is 

vice chair the Sarcoma Disease Site Committee of the 

American College of Surgeons Oncology Group (ACOSOG). 

Dr. O’Sullivan also represents the International Union 

Against Cancer (UICC) in the area of staging of sarcoma 

and head and neck cancer at the AJCC. Some of his recent 

research is directed at incorporation of image-guided techniques 

into enhanced radiotherapy delivery platforms to 

achieve higher precision and for both head and neck cancers 

and sarcomas. Dr. O’Sullivan has been the author of more 

than 130 peer-reviewed papers, nearly 100 additional scientific 

articles or book chapters, and has more than 100 invited 

lectures at national and international medical meetings. 



able to: 

1) Describe clinical aspects of the varied subtypes of head 

and neck soft tissue sarcoma (STS) that impact on management 

approaches. 

2) Review the potential role of systemic treatment approaches 

in the curative management of STS of the head and neck. 

3) Elaborate on decision-making concerning the choice and 

timing of adjuvant radiotherapy combined with surgery in 

the treatment of STS. 

4) Understand target definitions and methods of precision 

image-guided radiotherapy delivery for head and neck sarcomas. 


Soft tissue sarcomas in the head and neck are rare and management 

principles are extrapolated from other more common 

anatomical sites. However, evidence that the biologic 

behavior differs by anatomical site is not apparent. What sets 

these lesions apart from tumors elsewhere is the unique functional 

and cosmetic requirements and surgery and radiotherapy 

(RT) must be applied more precisely and with greater 

complexity than elsewhere in the body because of the presence 

of critical anatomy. In general the preferred treatment 

for these lesions is surgery alone, especially in small and 

especially superficial subcutaneous tumors with origin 

above the investing fascia; this is only rarely possible due to 

the anatomical constraints in this anatomical area and there 

is the added concern that local failure in head and neck STS 

confers extremely adverse outcome that is often fatal. 

Radiotherapy can be administered preoperatively, postoperatively, 

or by brachytherapy and these all seem to be equivalent. 

Preoperative RT offers strategic advantages where the 

dose and volume to be irradiated are important (e.g., in proximity 

to critical anatomical structures such as optic 

nerves/chiasma, brain stem, and spine). Radiotherapy alone 

for resectable tumors is not advisable with the exception of 

alveolar and embryonal rhabdomyosarcoma where exquisite 

sensitivity to radiotherapy is generally the case. A wide spectrum 

of histologic subtypes is observed and the prognosis 

varies depending on the histology and grade. One tumor, dermatofibrosarcoma 

protuberans (DFSP), is of interest from a 

biologic and therapeutic standpoint because inhibition of the 

PFGF tyrosine kinase by imatinib makes it a rare tumor with 

clear response to molecular targeting. The value of adjuvant 

chemotherapy is unclear for soft tissue sarcomas in general. 

This is problematic because there is genuine risk of metastatic 

failure especially in large, high grade and deep lesions. A 

meta-analysis has shown that doxorubicin-based adjuvant 

Tuesday

Tuesday 

chemotherapy significantly improves both local control and 

distant metastases-free survival, particularly for patients 

with extremity sarcomas. Although relapse-free survival also 

was improved significantly, the improvement in overall survival 

was not statistically significant. In summary, the optimal 

treatment for adults with head and neck soft tissue sarcomas 

is complete resection. Adjuvant RT likely improves 

control in patients with close or positive margins and likely 

improves survival. Although the efficacy of adjuvant 

chemotherapy is ill-defined, it probably should be considered 

for high-grade lesions. 

REFERENCES 

1. O'Sullivan B, Gullane P, Irish J, et al. Preoperative radiotherapy 

for adult head and neck soft tissue sarcoma: assessment 

of wound complication rates and cancer outcome in a 

prospective series. World J Surg 2003;27(7):875-883 

2. Pisters P, O’Sullivan B. Soft Tissue Sarcomas. In: Holland JF, 

Frei E, eds. Cancer Medicine, 6th ed., Hamilton, BC Decker, 

Inc.; 2003, Chapter 125: 2049-2076 

3. O’Sullivan B, Audet N, Catton C, Gullane P. Soft Tissue and 

Bone Sarcomas of the Head and Neck. In: Harrison LB, 

Sessions RB, Hong WK. Head and Neck Cancer: a multidisciplinary 

approach, 2nd ed. Lippincott Williams and Wilkins; 

2003:786-823 

4. O’Sullivan B, Davis A, Turcotte R, Bell R, Catton C, Wunder J, 

Chabot P, et al. Preoperative versus postoperative radiotherapy 

in soft tissue sarcoma of the limbs: a randomized trial. 

The Lancet 2002;(29)359(9325):2235-2241 

5. Sarcoma Meta-analysis Collaboration. Adjuvant chemotherapy 

for localized resectable soft tissue sarcoma of adults: 

meta-analysis of individual data. The Lancet 1997;350:1647- 

1654 


1:00 PM - 2:30 PM 

Room 201B 

(31) ASNR Business Center - Part II 

(204) Financial Perspectives and Concepts in 

Radiology 

— Dieter R. Enzmann, MD, MBA 

(205) Private Practice vs Academic Management 

— Steve Stevens, MD 

(206) Offshore Teleradiology: Bane or Boon? 

— William G. Bradley, MD, PhD, FACR 

Moderator: Gregory L. Katzman, MD 

106 


1:00 PM - 2:30 PM 

Room 103 

(32) ELC Workshop D: Adobe 

Photoshop and Elements 


3:00 PM - 4:30 PM 

Room 105/106 

— Richard M. Berger, MD 

(33a) INTERVENTIONAL: 

Aneurysms and Spinal Vascular 

Malformations 



(33b) ADULT BRAIN: Functional Imaging and 

Advanced Techniques 

(33c) ADULT BRAIN: Functional and Advanced 

Imaging of Brain Neoplasms 

(33d) ADULT BRAIN: Degenerative, Dementias, and 

Destructive Lesions 

Moderators: Ajay K. Wakhloo, MD, PhD 

Alexander M. Norbash, MD 


Balloon-in-Stent Technique for the Constructive 

Endovascular Treatment of “Ultra Wide-Necked” 

Circumferential Aneurysms 

Fiorella, D. 1 · Albuquerque, F. C. 2 · Masaryk, T. J. 1 · 

Rasmussen, P. A. 1 · McDougall, C. G. 2 



PURPOSE 

Circumferential aneurysms, which incorporate more than 

180 degrees of the circumference of the parent vessel, present 

a unique therapeutic challenge, particularly in circumstances 

where a deconstructive treatment strategy is impos-

sible or tenuous. The current report details a novel technique 

for endovascular parent vessel reconstruction with constructive 

aneurysm embolization. 


A retrospective review of the prospectively maintained databases 

of our two institutions was performed to identify cases 

in which a balloon-in-stent technique was employed to treat 

circumferential aneurysms. During the first stage of this 

technique, a stent [Neuroform (Boston Scientific, Natick, 

MA) or Multilink Vision (Guidant, Indianapolis, IN)] is 

placed across the neck of the aneurysm to achieve parent 

vessel reconstruction. During the second stage, aneurysm 

coil embolization is performed with a compliant angioplasty 

balloon [Sentry (Boston Scientific, Natick, MA) or 

Hyperglide (Microtherapeutics, Irvine, CA)] placed within 

the stent to unambiguously demarcate and protect the parent 

vessel. Typically during the course of the embolization, coils 

project over and obscure the parent vessel in both working 

views. Prior to each coil detachment, the protection balloon 

is deflated under blank fluoroscopic roadmap visualization. 

The absence of shifting of any portion of the coil mass during 

balloon deflation indicates that the introduced coil is 

external to the stent (within the aneurysm) and can be 

detached. This process is repeated until satisfactory 

aneurysm embolization is achieved. After embolization, the 

balloon catheter may be exchanged for a stent delivery system 

to facilitate the placement of a second stent. 

RESULTS 

Seven patients underwent balloon-in-stent assisted 

embolization over a 15-month period. Three were performed 

in the setting of internal carotid aneruysms, three for basilar 

trunk or basilar apex aneurysms, and one for a 

dissecting/fusiform V4 segment vertebral artery aneurysm. 

In three cases, the presence of the inflated balloon facilitated 

the manipulation of the image intensifier into a position 

which produced a “down-the-barrel” view of the parent vessel. 

In the four additional cases, for anatomical reasons, this 

view could not be achieved and coil mass projected over the 

reconstructed parent vessel in both views. Partial aneurysm 

occlusion (75-90%), was achieved in five cases and near 

complete (> 95%) occlusion achieved in two cases. 

Complications included two retroperitoneal hematomas and 

one paramedian pontine perforator infarct. Follow up is 

available for three patients (2, 4, and 11.5 months), demonstrating 

progressive thrombosis (n = 1), no change (n = 1) 

and recanalization/coil compaction (n = 1). 

CONCLUSION 

The balloon-in-stent technique provides a practical and safe 

treatment strategy for the management of circumferential 

aneurysms which are not amenable to a deconstructive 

embolization. 

KEY WORDS: Aneurysm, Neuroform stent 

107 


Trispan ® -Assisted Coil Occlusion of 100 Wide-Necked 


Henkes, H. 1,2 · Fischer, S. 1 · Liebig, T. 1 · Weber, W. 1 · 

Reinartz, J. 2 · Miloslavski, E. 2 · Kuehne, D. 1 

1 2 Alfried Krupp Krankenhaus, Essen, GERMANY, Robert 

Janker Klinik, Bonn, GERMANY 

PURPOSE 

TriSpan ® is a modified coil, originally developed in order to 

assist the endovascular coil occlusion of wide-necked bifurcation 

aneurysms. The aim of this study was to evaluate the 

safety and efficacy of TriSpan-assisted, dual-catheter coil 

occlusion of intracranial aneurysms. 


Between October 2000 and September 2004, TriSpan 

devices were used in 107 endovascular procedures, carried 

out in 99 patients or 100 aneurysms (52 ruptured, 48 unruptured). 

The average neck and fundus sizes of the initially 

treated aneurysms were 7 ± 2.8 mm and 13 ± 5.5 mm, 

respectively. TriSpan devices were used in 82 initial and in 

25 recurrent treatment sessions. In 9 procedures a combination 

of stent and Trispan was used. 

RESULTS 

A total of 98 of these procedures resulted in an occlusion rate 

of 90-100%. Thromboembolic events were encountered during 

or after 19 procedures. Follow-up angiograms were 

available for 65 out of 107 treatment sessions and were carried 

out after a mean interval of 11.8 months. Complete or 

subtotal occlusion (90-100%) was found in 38 aneurysms 

(58%). In 27 aneurysms (42%) any type and degree of recurrence 

was encountered. Aneurysm hemorrhage after the 

treatment occurred in 2 aneurysms after 10 and 18 months, 

respectively. 

CONCLUSION 

Dual catheter-TriSpan procedures are safe and effective in 

the endovascular treatment of wide-necked bifurcation 

aneurysms. The device does, however, not prevent recurrence, 

which is frequent in these aneurysms. 

KEY WORDS: Wide-neck aneurysm, endovascular, Trispan ® 


Combined Therapy with Clopidogrel and Aspirin in 

Patients Undergoing Elective Coil Embolization of 


Gandhi, D. · Gemmete, J. · Nicol, E. · Razack, N. 

University of Michigan Hospitals 

Ann Arbor, MI 

PURPOSE 

Coil embolization is a widely accepted therapy for intracranial 

aneurysms. Thrombus formation on the coils, inside the 

aneurysmal lumen and at the neck of parent vessel-coil interface 

may be the source of most thromboembolic complications 

during GDC therapy. Pretreatment with combined therapy 

with clopidogrel and aspirin may decrease the risk of 

such complications. 

Tuesday

Tuesday 


During a 14-month period (09/03-11/04), 30 intracranial 

aneurysms were treated electively at our institution with coil 

embolization. All elective cases were pretreated with 81 mg 

aspirin and 75 mg clopidogrel orally for a period of 5 days 

before the embolization procedure. All patients also received 

intraprocedural heparin with an aim to maintain the ACT 

between 250-300 seconds. The aneurysms were treated with 

GDC (Boston Scientific) or hydrocoil embolic system 

(Microvention). Two patients additionally received intracranial 

stents for the treatment of wide-necked aneurysms. We 

reviewed the medical records, procedure notes, and 

angiograms of all patients. 

RESULTS 

None of our 30 elective patients demonstrated angiographic 

evidence of thrombus at the coil-vessel interface or distal 

embolic occlusion of intracranial arteries. There were no 

instances of intracranial hemorrhage. Two patients experienced 

easy bruising on the antiplatelet therapy. 

CONCLUSION 

In the available neuroradiology literature, formation of 

thrombus and distal embolic complications has been reported 

to occur from 2-6.7% during coil embolization of 

intracranial aneurysms. Combined treatment with aspirin 

and clopidogrel is a safe and effective means for the prevention 

of thrombus formation at the coil-parent vessel interface 

and distal thrombo-embolic complications. This pretreatment 

regimen should be considered for the electively 

planned embolization procedures. 

KEY WORDS: Aneurysm, thrombus, clopidogrel 


Monorail Snare Technique for the Recovery of Stretched 

Platinum Coils 




PURPOSE 

Coil stretching represents a potentially hazardous technical 

complication not infrequently encountered during the 

embolization of cerebral aneurysms. Often the stretched coil 

cannot be advanced into the aneurysm or withdrawn intact. 

The operator then is forced to attempt to retract the damaged 

coil which may result in coil breakage leaving behind a significant 

length of potentially thrombogenic stretched coil 

material within the parent vessel. To overcome this problem, 

we devised a technique to snare the distal, unstretched, intact 

portion of the platinum coil using the indwelling microcatheter 

and stretched portion of the coil as a guidewire. 


A 2 mm Amplatz “goose neck” microsnare (Microvena 

Corporation, White Bear Lake, MN) was placed through a 

Prowler-14 microcatheter (Cordis Corporation, Miami, FL). 

The hub of the indwelling SL-10 microcatheter (Boston 

Scientific, Natick, MA) then was cut away with a scalpel, 

leaving the coil pusher wire intact, and removed. The open 2 

mm snare then was advanced over the outside of the coil pusher 

wire and microcatheter. The snare and Prowler-14 micro- 

108 

catheter then were advanced into the guiding catheter (6 or 7 

French) as a unit over the indwelling SL-10 microcatheter. 

Using the SL-10 microcatheter and coil as a “monorail” guide, 

the snare was advanced over and beyond the microcatheter 

and stretched portion of the coil until the snare was in position 

to engage the distal unstretched coil. At this point, the snare 

was closed around the intact portion of the coil and the microcatheters, 

snare, and coil were removed as a unit. 

RESULTS 

We have used this technique successfully in four patients to 

snare coils stretched during cerebral aneurysm embolization. 

Three of these patients were undergoing Neuroform (Boston 

Scientific, Natick, MA) stent-supported coil embolization of 

unruptured aneurysms. In all cases, using the microcatheter 

as an in-line monorail guide, the snare was advanced easily 

to the targeted site for coil engagement. Once the intact distal 

segment of the coil was ensnared, coil removal was 

uneventful with no disturbance of the remainder of the 

indwelling coil pack (or Neuroform stent). 

CONCLUSION 

The monorail snare technique represents a fast, safe, and 

easy method by which a stretched coil can be removed. 

KEY WORDS: Aneurysm, embolization 


Neuroform Stent-Supported Aneurysm Embolization: 

Initial (3-6 Month) Follow-Up Results 




PURPOSE 

The current paper details our experience with the Neuroform 

stent with an emphasis on treatment durability at initial (3-6 

month) follow up. 


Patient history, indications for stent use, aneurysm dimensions, 

technical detailsof the procedures, degreeof aneurysm 

occlusion, angiographic and clinical findings at follow up 

and complications were tracked in a prospectively maintained 

database. 

RESULTS 

Sixty-four patients with 74 aneurysms were treated with 86 

Neuroform stents over a 20-month period. Indications for 

stent use included broad aneurysm neck (n = 51 stents; average 

neck 5.1 mm, aneurysm size 8.2 mm), fusiform/dissecting 

morphology (n = 17), salvage/bailout for coils prolapsed 

into the parent vessel (n = 7) and giant aneurysm (n = 11). 

Sixty-one aneurysms were stented and coiled with complete 

or near complete (> 95%) occlusion in 28 cases and partial 

occlusion (< 95%) in 33 cases. Follow-up conventional 

angiographic (n = 41) or MRA (n = 5) data (average follow 

up 4.6 months, median 4 months, range 1.5-13 months) for 

46 aneurysms (44 patients) after stent-supported coil 

embolization demonstrated progressive thrombosis in 24 

cases, recanalization in 10 cases (7 of which were retreated), 

and no change in 12 cases. Follow-up angiography in five 

additional patients with dissecting aneurysms treated with

stents alone demonstrated vascular remodeling with 

decreased aneurysm size in all cases. Delayed in-stent stenosis 

was observed in three patients. 

CONCLUSION 

The Neuroform stent facilitates the durable embolization of 

complex cerebral aneurysms. Initial follow-up data demonstrate 

very favorable rates of progressive thrombosis — particularly 

for small aneurysms with wide necks and 

fusiform/dissecting aneurysms. 

KEY WORDS: Neuroform, aneurysm 


Contrast-Enhanced MR Angiography in Spinal Vascular 

Malformations Compared with Digital Subtraction 

Angiography 

Mull, M. 1 · Nijenhuis, R. J. 2 · Wilmink, J. T. 2 · Backes, W. H. 2 

· Thron, A. 1 

1Aachen University Hospital, Aachen, GERMANY, 

2Maastricht University Hospital, Maastricht, THE 

NETHERLANDS 

PURPOSE 

To study the impact of contrast-enhanced MR angiography 

(CE MRA) in identification and classification of spinal vascular 

malformations compared with digital subtraction 

angiography (DSA). 


Eleven consecutive patients with spinal vascular abnormalities 

[9 spinal dural arteriovenous fistula (SDAVF) and 2 

spinal arteriovenous malformations (SAVM)] underwent 

both a CE MRA and DSA examination. All examinations 

were done before treatment. MR angiography (1.5 T) consisted 

of a 3D gradient spoiled-echo sequence with centrically 

ordered k-space filling of which the start was synchronized 

with the arrival of a 0.3 mmol/kg dose of gadolinium 

contrast agent in the lower aorta. Before contrast administration 

four mask scans were acquired, averaged, and subtracted 

from the contrast-enhanced scan to suppress background 

inhomogeneity, to limit noise propagation, and thus enhance 

small-vessel detectability. Voxel sizes were 0.8 × 0.8 × 1.2 

mm at acquisition. The level and side of the SDAVF and the 

feeding arteries in SAVM were determined by the creator of 

(curved) multiplanar reformatted (MPR) images, who was 

blinded for the DSA results. Three experienced neuroradiologists 

jointly evaluated the underlying vascular pathology 

and compared MRA with DSA findings. 

RESULTS 

The localization and morphology of the SDAVF by MRA was 

in agreement with the DSA results in 7 out of 9 cases. One mismatch 

of one vertebral level (not side) was noted. In the other 

case a deep sacral SDAVF (S1 right) could not be confirmed by 

MRA. The entire venous drainage was better seen in MRA. 

The level and course of the main feeding arteries in complex 

intradural SAVM were identified by MRA, other feeders were 

detected only by DSA. The type of the SAVM (fistula versus 

glomus type) could not be classified on the basis of the MRA 

findings alone. If the SDAVF and the arterialized veins were 

close to the origin of the Adamkiewicz artery, separation and 

identification of the anterior spinal artery was more difficult. 

109 

CONCLUSION 

This advanced CE MRA technique is a powerful tool to 

localize SDAVF and to identify the predominant feeders in 

SAVM. Contrast-enhanced MRA appears to be useful 

adjunct to DSA to focus the DSA examination, and may 

reduce burdensome DSA time. However, due to the lower 

spatial and temporal resolution of MRA compared with DSA 

definite pretherapeutic classification of SAVM still should 

be based on DSA. The venous drainage pattern in SDAVF 

can be demonstrated nicely on CE MRA and may open new 

insights in pathologic mechanisms. 

KEY WORDS: Spinal vascular malformations, CE MRA, 

DSA 


Negative Predictive Value of CT Angiography in Patients 

with Spontaneous Subarachnoid Hemorrhage 

Shreiber, R. 1 · Eran, A. 1 · Daitzchman, M. 1 · Abrantes, Y. 1 · 

Goldsher, D. 1,2 

1 2 Rambam Medical Center, Haifa, ISRAEL, Technion, Israel 

Institute of Technology, Haifa, ISRAEL 

PURPOSE 

To assess the negative predictive value of multislice CT 

angiography (CTA) in the evaluation of spontaneous subarachnoid 

hemorrhage (SAH). 


All patients admitted to our hospital in the last 40 months 

with spontaneous SAH underwent multislice CTA as a part 

of their routine workup, and all had digital subtraction 

angiography (DSA) within 96 hours of admission. CT 

angiography was performed using MX8000 and MX8000 

IDT (Philips) with 4- and 16-slice multidetector scanners, 

respectively. In 41 patients, no evidence of aneurysm was 

found on multislice CTA. Conventional four vessel DSA was 

performed on all 41 patients (Multistar Siemens). 

Tuesday

Tuesday 

RESULTS 

No aneurysm larger than 2.5 mm in size was found on the 

multislice CTA in our series. Of the 41 patients, 38 had no 

aneurysm demonstrated on DSA in accordance with multislice 

CTA findings. One patient had an aneurysm of the 

intracavernous segment of the right carotid siphon, masqueraded 

by enhanced blood within the cavernous sinus. On 2/41 

patients, some unusual findings drew our attention. A wide 

anterior communicating artery (AcomA) on one patient and 

abnormal appearance of vessels that seemed to be compatible 

with widened arteries in the posterior fossa were noted. 

Digital subtraction angiography disclosed an aneurysm of 

the AcomA in the first patient and a dissecting aneurysm of 

the left PICA in the other patient. A total of 38/39 patients 

demonstrated no aneurysm on both CTA and DSA, yielding 

a negative predictive value of 0.974 in cases of spontaneous 

SAH in our study. 

CONCLUSION 

We believe that with such a high negative predictive value, 

CTA can be the first and only modality to rule out intracranial 

aneurysm in cases of spontaneous SAH, provided that 

special attention in postprocessing is given to the carotid 

siphon. Because of the limited number of patients in our 

report, more studies are needed to corroborate this conclusion. 

KEY WORDS: Subarachnoid hemorrhage, CT angiography, 

aneurysm 


MR Digital Subtraction Angiography Using Parallel 

Imaging in the Diagnosis of Vasospasm after 

Subarachnoid Hemorrhage 

Tsuchiya, K. · Fujikawa, A. · Honya, K. · Nakajima, M. · 

Niitatori, T. 

Kyorin University 

Tokyo, JAPAN 

PURPOSE 

MR digital subtraction angiography (DSA) visualizes vessels 

employing a rapid T1-weighted sequence in combination 

with a bolus injection of gadolinium. We assessed 

whether MR DSA using parallel imaging is effective in the 

diagnosis of vasospasm after subarachnoid hemorrhage. 


On a 1.5 T imager, we used a three-dimensional fast fieldecho 

sequence using parallel imaging (TR/TE/excitations = 

3.1/0.9/1, flip angle = 20 deg, matrix = 128 x 256, field of 

view =26 x 28 cm, reduction factor = 2) with intravenous 

injection of 7 ml of gadolinium at a rate of 3 ml/sec. In 28 

patients examined 5-20 days after clipping of a ruptured 

cerebral aneurysm, we reviewed MR DSA images comparing 

with images of diffusion-weighted imaging, dynamic 

susceptibility contrast (DSC) perfusion imaging, and 3D 

time-of-flight (TOF) MR angiography (MRA). 

RESULTS 

Temporal resolution achieved was 0.8 sec/frame. Vasospasm 

was noted in a total of 48 arterial segments on MR DSA and/or 

3D TOF MRA. In 19 of the 48 segments (40%), MR DSA was 

able to detect vasospasms but 3D TOF MRA was not due to 

110 

artifacts from a clip. In other 26 segments (54%), MR DSA and 

3D TOF MRA showed concordant findings. In 16 of the 28 

patients (57%), diffusion-weighted images showed new 

ischemic lesions. MR DSA was able to depict vasospasm leading 

to ischemia in 10 of the 16 patients (63%). Abnormalities 

on DSC perfusion imaging noted in eight patients were consistent 

with areas with vasospasm on MR DSA. 

CONCLUSION 

MR DSA using parallel imaging can be a tool to sensitively 

visualize vasospasm after subarachnoid hemorrhage. 

KEY WORDS: Digital subtraction angiography, vasospasm, 

MR angiography 


First Line Investigation of Acute Intracerebral 

Hemorrhage Using Dynamic MR Angiography 

Evans, A. L. 1 · Coley, S. C. 1 · Wilkinson, I. D. 2 · Griffiths, P. D. 2 

1Royal Hallamshire Hospital, Sheffield, UNITED KINGDOM, 

2University of Sheffield, Sheffield, UNITED KINGDOM 

PURPOSE 

MR digital subtraction angiography (MR DSA) is a diagnostic 

tool that allows the dynamic assessment of the cerebral 

circulation in a relatively noninvasive fashion. We discuss 

our initial experience of this technique in the investigation of 

acute nontraumatic intracerebral hemorrhage. 


Twelve patients with acute intracerebral hemorrhage were 

investigated within 6 days of the ictus using a dynamic contrast-enhanced 

2D MR angiogram that produces subtracted 

images with a temporal resolution of 1-2 frame/s. The MR 

DSA examinations were assessed for evidence of an intracranial 

vascular abnormality and were compared with 1) the 

routine MR sequences, 2) nondynamic time-of-flight MR 

angiography, and 3) catheter angiogram performed during 

the same admission. 

RESULTS 

All 12 MR DSA examinations were considered to be technically 

satisfactory. MR DSA detected an intracranial vascular 

abnormality in 7 patients (3 arteriovenous malformations, 2 

aneurysms, 1 dural arteriovenous fistula, and 1 venous 

thrombosis). All abnormalities were confirmed by catheter 

angiography with the exception of one patient with venous 

sinus thrombosis found on MR imaging that did not undergo 

catheter angiography. All 4 arteriovenous shunts were 

detected by MR DSA by virtue of early venous filling.

CONCLUSION 

MR digital subtraction angiography can be performed satisfactorily 

in the setting of acute intracerebral hemorrhage and 

provides an alternative method to catheter angiography for 

identifying shunting vascular abnormalities such as arteriovenous 

malformations and fistulae; as well as large 

aneurysms and venous occlusions. MR digital subtraction 

angiography is a contrast medium-based technique that does 

not suffer from the T1 shortening effects of acute hemorrhage 

that can obscure abnormalities on conventional flowbased 

nondynamic techniques. 

KEY WORDS: Intracerebral hemorrhage, MR DSA 


Optimization of Velocity Encoding for Cerebral Blood 

Flow Quantification Using Phase-Contrast MR 

Angiography 

Guo, G. 1,2 · Wu, R. 1,2 · Lin, R. 1 · terBrugge, K. 2 · Mikulis, D. J. 2 

1The Second Hospital, Shantou University Medical College, 

2 Shantou, CHINA, The Toronto Western Hospital, 

University of Toronto, Toronto, ON, CANADA 

PURPOSE 

Accurate velocity encoding is crucial for quantification of 

artery inflow and venous outflow in intracranial diseases. 

The purpose of this study was to optimize the velocity 

encoding of phase-contrast (PC) MRA and quantify cerebral 

blood flow in volunteers. 


Nine volunteers were examined on a GE 1.5 T MR system 

with two-dimensional phase-contrast MRA sequence. The 

parameters of the sequence were as follow: TR 40 ms, TE 

6.6 ms, flip angle 20°, slice thickness 4 mm, matrix 256 x 

256, field of view 140 mm. In each cardiac circle, 40 images 

were obtained. Velocity encoding was set from 30 to 90 

cm/sec with an interval 10 cm/sec and 7 scans were performed 

for each volunteer. The scan level was chosen 

between C2 and C3 and perpendicular to the vessels of interest. 

Data were analyzed using a CV Flow software on a GE 

Advantage Windows Workstation (4.0). Artery inflow, 

venous outflow, peak velocity, and mean velocity were 

obtained for bilateral internal carotid artery (ICA), vertebral 

artery (VA), and jugular vein (JV). A paired sample t-test 

was used to determine the statistical significance of the differences. 

A P value below 0.05 was considered significant. 

RESULTS 

All the subjects had aliasing in the phase images and 6 of 

them had opposite direction of blood flow to the direction of 

velocity encoding in bilateral ICAs when Venc = 30 cm/sec. 

Significant differences were observed in artery inflow of 

bilateral ICA, peak velocity and mean velocity of right ICA 

when different velocity encodings were compared Venc = 30 

cm/sec with other velocity encoding (P < 0.05). There was 

no significant difference between paired data as velocity 

encoding ≥ 40 cm/sec (p > 0.05). There were no significant 

differences in venous outflow volumes, and blood flow volumes 

of bilateral VA when Venc was from 30 cm/sec to 90 

cm/sec (p > 0.05). Additionally, there were no significant 

differences for peak flow velocities and mean flow velocities 

between right and left ICA, VA, and JV with same velocity 

111 

encoding, respectively (p > 0.05). The maximum peak flow 

velocity of right ICA was 28.8 cm/sec, right VA was 20.5 

cm/sec, and right JV was 19.1 cm/sec, respectively. 

Therefore, for accurate cerebral blood flow measurement 

using phase-contrast MRA technique conveniently, the reasonable 

velocity encoding was 60 cm/sec for ICA, VA, and 

JV in the neck. The mean artery inflow of ICA was 655 ± 18 

ml/min and mean venous outflow of JV (C2-C3 level) was 

506 ± 186 ml/min. These results are consistent with previously 

reported flow values for these vessels. 

CONCLUSION 

In general, accuracy can be improved by increasing the spatial 

resolution and the encoded velocity and by decreasing 

the contrast between flowing blood and stationary tissue. To 

minimize error in the assessment of blood flow volume, in 

particular in the clinical environment, standardized techniques 

should be applied, and the imaging parameters should 

be selected carefully to minimize the impact of the potential 

sources of error. Combining the optimal parameters previously, 

phase-contrast MRA can be used to assess the relationship 

between major cerebral vessel inflow to outflow 

using an appropriate selection of velocity encoding. 

KEY WORDS: Phase-contrast MR angiography, quantitative 

blood flow, cerebral 


Spinal MR Angiography: Diagnosis, Pitfalls, and Follow- 

Up Assessment in Patients Treated for Spinal Dural 

Vascular Malformations 

Shah, T. C. · Woldenberg, R. F. · Setton, A. 


Manhasset, NY 

PURPOSE 

Demonstrate the utility and limitations of contrast-enhanced 

dynamic spinal MR angiography (MRA) in the diagnosis 

and long-term assessment of patients successfully treated for 

spinal dural vascular malformations and secondary pial 

venous hypertension. 


Using the protocol designated by Farb, et al. (1), MRA was 

performed on 10 patients in whom a spinal dural vascular 

malformation was suspected on conventional MR evaluation. 

An attempt to identify the level of vascular abnormality 

was undertaken; all patients subsequently underwent 

spinal catheter angiography. Short-term follow-up MRA 

(within 1 week posttreatment) was obtained in 8 patients. 

Long-term follow-up MRA (4-6 months posttreatment) was 

correlated with spinal catheter angiography. 

RESULTS 

All 10 patients demonstrated dilated venous structures on the 

surface of the spinal cord, consistent with venous hypertension 

associated with a spinal dural vascular malformation. 

Correlation with spinal catheter angiography revealed that, 

of the 10 patients with a suspected dural fistula, spinal MRA 

correctly identified the anatomical location of the fistula in 8 

patients. Pitfalls encountered in the interpretation of MRA 

findings included: 1) lack of fistula recognition secondary to 

incomplete coverage of the spinal dural surface, specifically 

Tuesday

Tuesday 

the lower lumbar area, during initial MRA examination; 2) 

misidentification of a dominant anterior spinal artery as a 

draining medullary vein; 3) the erroneous conclusion that an 

outlet vein actually represented the level of the fistula. 

Angiography revealed two patients with an epidural arteriovenous 

fistula rather than a simple dural fistula, one of 

which was identified correctly on the initial diagnostic 

MRA. The immediate posttreatment MRA and the long-term 

follow-up MRA demonstrated significantly reduced vascularity 

and myelopathy that correlated with clinical improvement 

and findings at follow-up catheter angiography. 

CONCLUSION 

Dynamic contrast-enhanced spinal MR angiography is a 

valuable and accurate modality in the diagnosis and pretreatment 

evaluation of patients with spinal dural fistulas. The 

information retrieved from preprocedure MRA has allowed 

for targeted spinal catheter angiography, thus significantly 

decreasing procedural complexity, duration, and contrast 

load. Based on our findings after correlating MRA with 

spinal catheter angiography, we are optimistic that spinal 

MRA will be an effective, reliable, and safe alternative to 

conventional angiography to monitor long-term success of 

treatment in patients who undergo obliteration of spinal 

dural fistulas. 

REFERENCES 

1. Farb RI, et al. Spinal dural arteriovenous fistula localization 

with a technique of first-pass gadolinium MR angiography: 

Initial experience. Radiology 2002;222:843-850 

2. Bowen BC, et al. Spinal dural arteriovenous fistulas: 

Evaluation with MR angiography. AJNR Am J Neuroradiol 

1995;(16)10:2029-2043 

KEY WORDS: Spinal dural malformation, spinal MRA 


3:00 PM - 4:30 PM 

Theatre 

(33b) ADULT BRAIN: Functional 

Imaging and Advanced Techniques 



(33a) INTERVENTIONAL: Aneurysms and Spinal 






Moderators: Joseph A. Maldjian, MD 

Michael A. Kraut, MD 

112 


Beyond the Diffusion Tensor: 3 T Parallel High Angular 

Resolution Diffusion Imaging of Complex White Matter 

Architecture in the Human Brain 

Hess, C. P. 1 · Mukherjee, P. 1 · Han, E. T. 2 · Xu, D. 1 · Vigneron, 

D. B. 1 

1University of California San Francisco, San Francisco, CA, 

2General Electric Applied Sciences Lab West, Menlo Park, 

CA 

PURPOSE 

High angular resolution diffusion imaging (HARDI) refers 

to the acquisition of large numbers of diffusion directional 

measurements at diffusion-weighting factors (b values) 

greatly exceeding 1000 s/mm 2 . HARDI surpasses diffusion 

tensor imaging (DTI) by enabling the visualization of complex 

white matter architecture in voxels containing converging, 

diverging, or crossing fiber tracts. In this study, we 

demonstrate implementation of HARDI at 3 T with parallel 

imaging using an 8-channel head coil, for improved signalto-noise 

ratio and spatial resolution in delineating the complex 

white matter connectivity of the human brain. 


HARDI was performed on five normal adult volunteers 

using a 3 T EXCITE scanner (General Electric, Milwaukee, 

WI). The 8-channel EXCITE head coil was employed for 

parallel imaging using the array spatial sensitivity encoding 

technique (ASSET). The 45-minute whole-brain HARDI 

acquisition was performed with a single-shot echo-planar 

spin-echo pulse sequence with 131 diffusion-encoding directions 

at b = 3000 s/mm 2 and 2.0 mm isotropic spatial resolution 

(TR = 18 s, TE = 84 ms, NEX = 1, ASSET acceleration 

factor 2). The orientation distribution function (ODF) was 

constructed for each voxel using Q-ball (1) and spherical 

deconvolution (2) methods. The resulting ODFs were compared 

with DTI analysis of the same HARDI data. 

RESULTS 

Representative results are shown from a single 2 mm cubic 

voxel (yellow box in the coronal DTI directionally-encoded 

color fractional anisotropy image, left) containing intersecting 

fibers from the corpus callosum (CC), the superior longitudinal 

fasciculus (SLF), and the centrum semiovale (CS). 

The cingulum bundle (CB) also is labeled. In all subjects, 

both the Q-ball ODF (bottom left) and the spherical deconvolution 

(SD) ODF (bottom right) provided accurate reconstructions 

of intravoxel crossing fiber tracts that are consistent 

with known white matter anatomy. In contradistinction, 

the 3D profile from DTI analysis (top) failed to correctly 

resolve crossing fibers, yielding no useful fiber orientation 

information in regions of complex white matter architecture.

CONCLUSION 

HARDI represents a major advance over DTI for delineating 

the complex white matter connectivity of the human brain. 

The gain in signal-to-noise ratio afforded by high field (3 T) 

and parallel imaging with an 8-channel head coil brings 

HARDI closer to feasibility in the clinical setting. Potential 

applications include improved fiber tracking accuracy, 

resilience to partial volume effects in which a single voxel 

contains two or more adjacent fiber pathways, and better 

characterization of white matter disease in regions of complex 

white matter architecture. 

REFERENCES 

1. Tuch DS, et al. Neuron 2003;40:885-895 

2. Tournier JD, et al. NeuroImage 2004;23:1176-1185 

KEY WORDS: Diffusion, tensor, tractography 


Fractional Anisotropy on a 3 T System: Feasibility Study 

Rumboldt, Z. · Bohning, D. · Besenski, N. · Vincent, D. · 

Nicholas, J. 

Medical University of South Carolina 

Charleston, SC 

PURPOSE 

Fractional anisotropy (FA) maps may be calculated from diffusion 

tensor imaging (DTI) data, and have been used to 

delineate white matter tracts in the brain. It is a promising 

tool for evaluation of patients with demyelinating lesions, 

brain tumors, and for investigation of brain maturation. 

Unfortunately, results obtained in evaluation of tumors and 

adrenoleukodystrophy have been conflicting, and reproducibility 

of this technique has not been evaluated fully. Our 

goal was to assess the reproducibility and variability of FA 

measurements in the brain white matter tracts on a 3 T MR 

scanner in normal adult subjects, as well as compare data 

acquisition with b values of 700 and 1000. 


Twelve normal volunteers, 6 of each sex, 25-50 years of age 

were included in the study. The imaging protocol included 

DTI with b values of 700 and 1000, performed twice, 2 to 4 

weeks apart. Two raters, blinded to any patient information, 

independently placed single regions of interest (ROIs) in the 

anterior limb, genu, and posterior limb of the internal capsule 

bilaterally, in the midline of genu and splenium of the 

113 

corpus callosum, bilaterally in the area of corticospinal tracts 

at the level of middle cerebellar peduncles, and in the right 

thalamus on the processed DTI images. The extracted 

numerical values were analyzed for reproducibility: between 

two b values, interrater, and between the two scanning sessions. 

RESULTS 

The results are summarized in Tables 1-3. At the level of 

pons the images in some subjects were compromised by susceptibility 

artifacts. In two subjects prominent artifacts, presumably 

from mineralization within the lentiform nuclei, 

completely precluded FA measurements.Inter-rater correlation 

coefficients (Pearson correlation) 

Combined results for both readers comparing b values 

of 700 and 1000 

ROI location and p value b value Mean Standard Standard 

Deviation Error 

Right Internal Capsule Anterior Limb p=.562 1000 692 .075 .016 

700 .704 .069 .015 

Right Internal Capsule Genu p=.021* 1000 .708 .055 .011 

700 .748 .060 .013 

Right Internal Capsule Posterior Limb p=.107 1000 .744. .034 .007 

700 765 .052 .011 

Corpus Callosum Genu p=.412 1000 .783 .071 .015 

700 .765 .071 .015 

Corpus Callosum Splenium p=.886 1000 .824 .076 .016 

700 .827 .053 .011 

Left Internal Capsule Anterior Limb p=.057 1000 .678 .085 .018 

700 .724 .072 .015 

Left Internal Capsule Genu p=.216 1000 .727. .059 .012 

700 746 .041 .009 

Left Internal Capsule Posterior Limb p=.118 1000 .757 .057 .012 

700 .783 .053 .011 

Right Corticospinal Tract p=.598 1000 .644 .103 .021 

700 .663 .142 .030 

Left Corticospinal Tract p=.683 1000 .679 .079 .016 

700 .669 .096 .020 

Right Thalamus p=.599 1000 .345 .037 .008 

700 .351 .042 .009 

Right Right Right Corpus Corpus Left Left Left Right Left Right 

Internal Internal Internal Callosum Callosum Internal Internal Internal Cortico- Cortical Thalamus 

Capsule Capsule Capsule Genu Splenium Capsule Capsule Capsule spinal Tract 

Anterior Genu Posterior Anterior Genu Posterior Tract 

Limb Limb Limb Limb 

b value .405 .353 .901* .624* .214 .649* .439 .894* .667* .721* .665* 

1000 

b value .572 .249 -.038 .289 .789* .035 .143 .399 .260 .285 .455 

7000 

*significant at 0.05 level 

Test-retest comparison for both b values 

ROI location Scanning Mean p value 

Session b=1000 b=700 b=1000 b=700 

Right Internal Capsule Anterior Limb 1 .687 .692 p=.744 p=.491 

2 .672 .718 

Right Internal Capsule Genu 1 .709 .737 p=.151 p=.140 

2 .737 .760 

Right Internal Capsule Posterior Limb 1 .745 .757 p=.382 p=.716 

2 .732 .751 

Corpus Callosum Genu 1 .799 .794 p=.390 p=.546 

2 .816 .782 

Corpus Callosum Splenium 1 .842 .824 p=.728 p=.162 

2 .836 .851 

Left Internal Capsule Anterior Limb 1 .688 .720 p=.882 p=.766 

2 .682 .710 

Left Internal Capsule Genu 1 .732 .750 p=.442 p=.963 

2 .746 .752 

Left Internal Capsule Posterior Limb 1 .768 .794 p=.234 p=.224 

2 .745 .767 

Right Corticospinal Tract 1 .651 .633 p=.048* p=.963 

2 .570 .631 

Left Corticospinal Tract 1 .706 .666 p=.161 p=.863 

2 .649 .672 

Right Thalamus 1 .349 .349 p=.024* p=.568 

2 .316 .340 

Tuesday

Tuesday 

CONCLUSION 

The variability in FA measurements obtained on a 3 T system 

is in some cases over 5% and statistically significant. 

Furthermore, in some individuals susceptibility artifacts may 

preclude analysis. Image acquisition with b value of 700 

appears to minimize test-retest variability, but also may 

increase interrater variability compared to the data obtained 

with b value of 1000. These findings should be taken into 

account in further studies. 

KEY WORDS: Diffusion tensor imaging, fractional 

anisotropy, reproducibility 


Abnormal Auditory Processing in Dyslexia: A Functional 

MR Imaging Study 

Milner, L. D. · Burdette, J. H. · Laurienti, P. J. · Maldjian, J. 

A. · Kraft, R. A. · Flowers, D. L. · Wood, F. B. 



PURPOSE 

Developmental dyslexia is a common language disorder that 

manifests as reading difficulty in individuals of otherwise 

normal intelligence and motivation. Several researchers have 

suggested that dyslexia results from an underlying visual 

and/or auditory sensory processing abnormality. We believe 

that dyslexia may result from abnormal interactions between 

the auditory and visual cortices in specific cross-modal 

regions, and designed an functional MR imaging (fMRI) 

experiment to test this hypothesis. 


N = 41 dyslexic patients, diagnosed on the basis of standardized 

neuropsychiatric testing, and N = 49 normal controls 

underwent an event-related fMRI examination on a 1.5 

T MR scanner while performing an auditory forced-choice 

task. Specifically, subjects had to determine whether an auditory 

syllable heard through the MR-compatible headphones 

matched an auditory target syllable given at the beginning of 

the study (e.g., a test “da” matches the target “da,” while 

“ga” represents a nonmatch). Standard echoplanar fMRI 

imaging using blood oxygen level dependent (BOLD) contrast 

was performed. Data was processed using SPM99, and 

contrast images for each subject were generated to identify 

the main effects of (1) when the presented syllable matched 

the target (match) and (2) when it did not match (nonmatch). 

Using a random effects model, brain activity was compared 

between the dyslexic and normal readers, and the resulting 

activation maps were corrected for multiple comparisons at 

p < 0.05. 

RESULTS 

When the presented auditory syllable did not match the target 

syllable (nonmatch condition), dyslexic readers showed 

increased activity in left Brodmann area (BA) 37 (Figure), 

an area at the junction of the occipital and temporal lobes. 

Conversely, during the “match” condition, there were no differences 

in activation between the two groups. Interestingly, 

dyslexics performed less accurately on the “nonmatch” condition 

compared with normal controls at p < 0.05, but performed 

similarly in the “match” condition. 

114 

CONCLUSION 

Using fMRI, we showed that dyslexic and normal readers 

used similar brain regions to perform an auditory forcedchoice 

task when the presented syllables matched the target 

stimulus. However, during the nonmatch condition, dyslexic 

readers utilized the left BA 37, a vital area within a known 

language circuit. We postulate that dyslexic readers have 

greater difficulty processing the incongruence of the stimuli 

during the nonmatch condition, requiring greater activity in 

the left BA 37. Taken further, this area, which is a known 

multisensory integration region may be utilized inappropriately 

during this nonmatch condition and may be a critical 

locus of abnormal processing causing the deficits seen in 

dyslexia. 

KEY WORDS: Functional MR imaging, dyslexia, auditory 

processing 


Structural Brain Changes Predict Functional Activation 

in Dyslexia 

Maldjian, J. A. · Laurienti, P. J. · Milner, L. D. · Kraft, R. A. 

· Flowers, D. L. · Wood, F. B. · Burdette, J. H. 

Wake Forest University 


PURPOSE 

Developmental dyslexia is a common language disorder that 

manifests as reading difficulty in individuals of otherwise 

normal intelligence and motivation. Despite the standard 

clinical diagnostic criteria for dyslexia, there is tremendous 

underlying variability in the expression of reading difficulty. 

In this paper we describe results from a multimodal integrative 

image analysis called biologic parametric mapping 

(BPM) that exploits this inherent variability. The current 

implementation of BPM utilized a voxel-by-voxel correlation 

analysis to probe functional imaging data with voxelbased 

morphometry (VBM) in a study of sensory processing 

in dyslexia.


Forty dyslexic patients, diagnosed on the basis of standardized 

neuropsychological testing, and 49 normal controls underwent 

functional MR imaging (fMRI) on a 1.5 T magnet using an 

auditory forced-choice task. Anatomical T1-weighted volumetric 

images also were acquired. The fMRI images were 

motion corrected, normalized to stereotaxic space, smoothed 

and analyzed using the General Linear Model within SPM99 

(Welcome Department of Cognitive Neurology, London, UK). 

Contrast images from a fixed effects analysis for each subject 

were generated to identify the main effect of auditory stimulation. 

Preprocessing steps for the VBM analysis of the structural 

high-resolution T1 images were performed as described 

by Ashburner (1). The contrast images from the fMRI analysis 

were used in a voxel-wise BPM correlation analysis with 

the processed structural images in a random effects model. 

The resulting correlation map was converted to a T-statistic 

and corrected for multiple comparisons at p < 0.05 using the 

false discovery rate. 

RESULTS 

The BPM analysis revealed that increases in gray matter 

concentration in the left parieto-temporal cortex and the right 

superior temporal sulcus were predictive of auditoryinduced 

activation (Fig. 1). An additional subanalysis 

restricted to the dyslexic population also revealed the region 

in the right superior temporal sulcus. However, the region in 

the left parieto-temporal area was not present. Interestingly, 

the opposite pattern was obtained with a subanalysis restricted 

to the normal readers. 

CONCLUSION 

Our BPM style analysis represents a powerful means of integrating 

multimodal functional imaging data. Important differences 

were demonstrated between dyslexic and normal 

readers in language processing areas that were not identified 

in standard fMRI or VBM analyses. These findings suggest 

that the abnormal auditory circuitry identified using BPM 

contributes to the language processing deficits in dyslexia. 

REFERENCES 

1. Ashburner J, Friston KJ. Voxel-based morphology-the methods. 

NeuroImage 2000;11:805-821 

KEY WORDS: Dyslexia, functional MR imaging, VBM 

115 


Neuroanatomical Organization of Sound Memory in 

Humans 

Kraut, M. A. 1 · Pitcock, J. A. 2 · Calhoun, V. 3 · Li, J. 2 · Hart, J. 2 

1The Johns Hopkins Medical Institutions, Baltimore, MD, 

2University of Arkansas Medical School, Little Rock, AR, 

3Yale University, Hartford, CT 

PURPOSE 

The neural interface between sensory perception and memory 

is a central issue in neuroscience, particularly initial memory 

organization following perceptual analyses. We used 

functional MR imaging to identify anatomical regions 

extracting initial auditory semantic memory information 

from environmental nonverbal sounds. 


We studied 18 normal subjects (13 men, 5 women, all right 

handed) using a 1.5 T MR instrument, and an EPI GRE pulse 

sequence. Subjects listened to binaurally-presented sounds, 

pushing one button if they identified the sounds as real, and 

another button if the sounds did not represent real items. The 

stimuli were 32 environmental sounds chosen from a set of 

sounds normed for familiarity, pleasantness, confidence in 

naming, naming accuracy, duration and reaction time. The 

real sounds were segregated into 2 categories; animals and 

nonliving objects. The categories were further equally divided 

into groups of threatening and nonthreatening based upon 

the pleasantness rating scale. We modified the original 

sounds by creating a spectrum-deforming template, and 

applied the same template to each real sound, changing the 

overall time-frequency envelope until the sounds were not 

identifiable. Two-tailed t-tests revealed no significant differences 

between the RMS power of any of the real sounds and 

their scrambled derivative nonreal sound. Image analysis 

was performed using SPM99. Imaging data for each individual 

were adjusted for time, corrected for motion, normalized 

into a standardized Talairach template and spatially 

smoothed (10 X 10 X 10 Gaussian kernel). Event-related 

analyses were conducted using a random-effects model 

excluding all false-positive and false-negative responses 

from the model for comparisons between animals and 

objects, and threatening and nonthreatening things. 

RESULTS 

Threatening stimuli, contrasted to nonthreatening sounds ( p 

< 0.001 uncorrected), elicited signal changes in the right 

superior temporal gyrus (STG) [Brodmann Area (BA) 41] 

posterior to the typical boundaries of the right primary auditory 

cortex, as well as in right inferior frontal gyrus (BA 47) 

and in the right superior parietal lobule (BA 7). Responses to 

stimuli representing animals contrasted to those representing 

nonliving objects also revealed foci of signal change in the 

right STG (BA 22), but rostral to the focus evident in the 

threatening/nonthreatening contrast and also outside of primary 

auditory cortex; in the left middle temporal gyrus (BA 

21 and 44), as well as in the left superior frontal gyrus (BA 

8) and in the medial left frontal lobe (BA 32). Small volume 

correction analysis showed that the right STG focus of signal 

change revealed with the threatening/nonthreatening 

contrast was driven mostly by the distinction between animal 

and object aspects of the stimuli (p < .001, corrected). 

Tuesday

Tuesday 

Similar analysis of the right STG focus elicited by the living/nonliving 

contrast reveals responses mostly to threatening 

stimuli. 

CONCLUSION 

Both category (living) and feature (threatening)-specific 

responses to environmental nonverbal sounds are evident in 

early stages of extraction of semantic memory information 

from these sounds, in regions adjacent to primary auditory 

cortex. Collocation of category and feature-related responses 

suggest distributed neural representations of the items 

these stimuli represent, perhaps related to an 

evolutionary/survival-related requirement for efficient processing 

and recognition of sounds related to dangerous animals. 

KEY WORDS: Functional MR imaging, memory, auditory 


Preoperative Functional MR Imaging Localization of 

Language and Motor Areas: Impact Upon Therapeutic 

Decision Making in Patients with Potentially Resectable 

Brain Lesions 

Shah, L. M. · Voyvodic, J. · Harris, K. · Friedman, A. · 

George, T. · Pekala, J. · Petrella, J. 


Durham, NC 

PURPOSE 

Preoperative functional MR imaging (fMRI) localization of 

language and motor areas: Impact upon therapeutic decision 

making in patients with potentially resectable brain lesions. 

Functional MR imaging has demonstrated potential to be an 

effective, noninvasive preoperative planning tool for showing 

the spatial relationship of functionally eloquent brain regions 

to intracranial pathology. Excellent correlation has been 

established between the results of fMRI and those of more 

invasive techniques, such as WADA testing and intraoperative 

cortical stimulation; however, definitive evidence that 

fMRI significantly impacts therapeutic decision making and 

patient outcome has not been established. Consequently, 

fMRI has not been fully integrated yet into the broader clinical 

practice of neuroradiology. The purpose of this study was 

to evaluate the impact of preoperative fMRI localization of 

language and motor areas on therapeutic decision making in 

patients with potentially surgically resectable brain lesions. 


Twenty-two consecutive patients, referred for preoperative 

fMRI by two neurosurgeons at our institution, were evaluated 

prospectively. Functional MR imaging was carried out on 

a 1.5 T scanner using GE EPI (TR/TE 2000/40) during sentence 

completion and bilateral hand squeeze tasks. 

Threshold statistical T-maps were created by correlation of 

the signal time course with an ideal hemodynamic response 

function and were overlaid in color on a set of conventional 

T2 SE images. Images were reviewed by two neuroradiologists 

in conjunction, who rendered an interpretation regarding 

the location of language and motor areas, as well as their 

proximity to the patient’s lesion. The referring neurosurgeon 

was given questionnaires regarding the proposed treatment 

plan before and after the fMRI interpretation. Questionnaires 

included 5-point confidence ratings in the fMRI results. 

116 

RESULTS 

Impact of fMRI on treatment approach is outlined in Table1. 

Functional MR imaging altered the neurosurgeon’s therapeutic 

planning in 12 of the 22 cases (54.5%), enabling a 

more aggressive approach in these12 cases. Of the 7 cases in 

which no operative intervention was planned initially, 5 

(71.4%) had awake craniotomies and 2 (28.6%) had biopsies. 

Of the 15 cases planned to have operative intervention, 

4 (26.7%) went from biopsy to awake craniotomies and 1 

(6.7%) went from awake to asleep craniotomy. Functional 

MR imaging did not alter the treatment plan in the remaining 

10 cases; however, in 8 of these 10 cases, the preoperative 

treatment plan was confirmed by the fMRI results. 

Table 1: Functional MR Imaging Affect on Treatment Planning 

Vertical (Pre fMRI Plan); Horizontal (Post fMRI Intervention) 

Total # of Patients = 22 No Surgery Biopsy Awake Craniotomy Asleep Craniotomy 

No Surgery 0 2 5 0 

Biopsy 0 0 4 0 

Awake Craniotomy 0 0 8 1 

Asleep Craniotomy 0 0 0 2 

CONCLUSION 

Preoperative fMRI localization of language and motor areas 

siginificantly impacts therapeutic decision making in 

patients with potentially surgically resectable brain lesions. 

In a significant number of cases, the results from fMRI 

enabled more aggressive therapeutic options than might 

have been considered. 

KEY WORDS: Functional MR imaging, preoperative planning 


Brainstem Serotonin 1A Receptor, Which Modulates 

Antidepressant Treatment, Is Altered in Late-Life 

Depression 

Meltzer, C. C. · Price, J. C. · Mathis, C. A. · Butters, M. A. · 

Ziolko, S. K. · Mazumdar, S. · Houck, P. R. · Reynolds, C. F. 



PURPOSE 

Late-life depression, with a prevalence of 6-10% of persons 

60 years of age or older, carries substantial and growing public 

health implications for the aging United States population. 

Depression in late life carries an increased risk of 

dementia and brittle response to treatment. Extensive evidence 

supports a key role of serotonergic dysfunction in the 

development of major depression and there has been 

increased attention on the serotonin 1A (5-HT 1A ) receptor 

as a regulator of treatment response, particularly the 5-HT 1A 

autoreceptor in the dorsal raphe nucleus (DRN). We used 

[ 11 C]WAY 100635 and positron emission tomography (PET) 

to test our hypothesis that 5-HT 1A receptor binding in the 

DRN is altered in elderly depressives and that these measures 

relate to treatment responsivity. 


We studied 20 elderly subjects with untreated (nonpsychotic, 

nonbipolar) major depression (5 men, 15 women; mean 

age: 71.6 ± 5.8 years) and 17 healthy control subjects (8 

men, 9 women; mean age: 70.0 ± 6.7 years). Patients subsequently 

were treated with paroxetine as part of a clinical 

trial. Treatment response was reflected by a time to remis-

sion measure, which was calculated as the number of weeks 

from the onset of therapy until achieving a Hamilton 

Depression Rating Score (HDRS) of 7 by weekly testing. 

Following injection of 10-15 mCi of [ 11 C]WAY 100635, 

dynamic PET imaging with arterial blood sampling over 60 

min was acquired. Volumetric SPGR MR imaging data was 

used to guide region-of-interest sampling of limbic, cortical, 

and brainstem areas and for partial volume correction. 

Regional binding potential (BP) values were derived using 

compartmental modeling. 

RESULTS 

We observed significantly diminished [ 11 C]WAY 100635 

binding in the DRN in depressed (BP = 2.25 ± 0.84) relative 

to control (BP = 3.69 ± 1.56) subjects (p = 0.004, corrected 

for multiple comparisons using the false discovery rate correction, 

p = 0.0004 prior to FDR correction). Further, the 

DRN BP was correlated with pretreatment Hamilton 

Depression Rating Scores (r = 0.60, p = 0.48 corrected for 

multiple comparisons using false discovery rate, p = 0.007 

prior to FDR correction) in the depressed cohort. 

CONCLUSION 

Our finding of decreased [ 11 C]WAY 100635 binding in the 

brainstem region of the DRN in elderly depressed patients 

supports experimental evidence of altered 5-HT 1A autoreceptor 

function in depression. Further, this work indicates 

117 

that dysfunction in autoreceptor activity may play a central 

role in the mechanisms underlying treatment response to 

selective serotonin reuptake inhibitors in late-life depression. 

KEY WORDS: PET, serotonin, depression 

Acknowledgment: Supported in part by the following grants: 

MH01210, MH59945, MH64625, MH43832, MH59769, and 

MH52247. 


Functional MR Imaging Evaluation of Deep Brain 

Stimulation for Treatment of Depression 

Phillips, M. D. 1 · Baker, K. B. 1 · Lowe, M. J. 1 · Kopell, B. 2 · 

Malone, D. 1 · Greenberg, B. D. 3 · Rezai, A. R. 1 

1 2 Cleveland Clinic Foundation, Cleveland, OH, Medical 

College of Wisconsin, Milwaukee, WI, 3Brown Medical 

School, Providence, RI 

PURPOSE 

To study the pattern of functional MR imaging (fMRI) activation 

produced by deep brain stimulation (DBS) for treatment 

of depression. 


A patient with percutaneously extended bilateral DBS electrodes 

in the anterior limb of the internal capsule with the 

distal contact in the nucleus accumbens for the treatment of 

intractable depression was studied using a 3 T Siemens 

Allegra MRI (Erlangen, Germany) on the first postoperative 

day. The externalized lead system was extended through the 

waveguide to an external pulse generator in the MR control 

room. Scanning consisted of three-dimensional anatomical 

data set with leads disconnected from the pulse generator 

and BOLD fMRI with a single lead connected to the pulse 

generator. Bold images were acquired for each DBS lead 

stimulating with lead contact 0 and 2 separately using 

parameters determined during surgery. Assessment of symptomatic 

efficacy for stimulation was performed at the time of 

surgery. Functional MR imaging examinations were performed 

with a block style paradigm consisting of 5 stimulator 

off and 4 stimulator on 32 second epochs with 10 seconds 

placed between the stimulator off and stimulator on conditions 

in order to gradually ramp the stimulator to the optimal 

stimulation level. Images acquired during the ramping 

process were discarded prior to image analysis. The MR 

imaging time series at each pixel was fit using least squares 

to a boxcar reference function plus a slope and intercept. 

RESULTS 

Stimulation with lead contact 0 produced improvement of 

the patients depression symptoms and activation in the ipsilateral 

putamen, ipsilateral dorsal medial thalamus, and contralateral 

cerebellum. Stimulation with right lead contact 2 

produced exacerbation of depressive symptoms. The left 

lead contact 2 produce equivocal symptomatic changes. 

Both left and right lead contact 2 produced a different pattern 

of fMRI activation than lead contact 0. Although activation 

was seen in ipsilateral dorsal medial thalamus and contralateral 

cerebellum, activation also was identified within the 

ipsilateral frontal lobe primarily within the superior frontal 

Tuesday

Tuesday 

gyrus Brodmann area is 8 and 9. No putaminal activation 

was seen with lead contact 2. Figure 1 shows activation patterns 

for right-sided contact 0 (1A) and contact 2 (1B-C). 

CONCLUSION 

The present study demonstrates the first-ever functional MR 

imaging performed in a patient receiving DBS therapy for 

depression. Stimulation with lead contact 0 produced symptomatic 

relief from the patients depression and a markedly 

different pattern of fMRI activation from stimulation with 

lead contact 2 which produced exacerbation of the patients 

symptoms. 

KEY WORDS: Functional MR imaging, deep brain stimulation, 

depression 


Anterior Cingulate Cortex Proton and Phosphorus MR 

Spectroscopy in Early Onset Bipolar Disorder 

Port, J. D. · Unal, S. S. · Mrazek, D. A. 

Mayo Clinic 

Rochester, MN 

PURPOSE 

To utilize both proton and phosphorus MR spectroscopy to 

examine the regional brain metabolites in children with bipolar 

disorder before and during treatment with lithium. 


Ten children between the ages 10 to 17 years diagnosed with 

bipolar disorder were recruited for the study; these subjects 

were not taking psychotropic at the time of the initial scan 

and had not been treated previously with mood stabilizers. 

Ten healthy age-, sex-, handedness-, and Tanner stagematched 

normal control subjects also were recruited. Proton 

and phosphorus MR spectroscopic imaging (MRSI) was performed 

in the bipolar subjects at four time points; 1) At baseline 

2) One week after initiation of lithium treatment 3) Six 

weeks after initiation of lithium treatment, and 4) Six months 

after the entry to the study. Control subjects received a single 

baseline scan. Single proton and phosphorus MRS voxels 

(3 x 3 x 3 cm) that included both anterior cingulate cortices, 

and whole slice phosphorus spectra were selected for 

further analysis. Proton spectra were quantified with 

LCModel; phosphorus spectra were quantified with TDFD- 

FIT. A paired statistical analysis was performed on the resulting 

metabolite ratios. 

118 

RESULTS 

Baseline anterior cingulate proton spectroscopy mI/Cr levels 

were significantly higher (p < 0.04) in bipolar subjects (0.76 

+/- 0.07) than in normal controls (0.65 +/- .11), and Glx/Cr 

levels were significantly higher (p < 0.03) in bipolar subjects 

(1.70 +/- 0.42) than in normal controls (1.34 +/- 0.30). There 

were no other significant changes in metabolite concentrations 

(regional proton or whole slice phosphorus) in bipolar 

subjects over time. However, ATP/Cr, PE/Cr, PDE/Cr, and 

Glx/Cr tended to increase in concentration over the 6-month 

time period of the treatment. Single voxel phosphorus data 

analysis at the time of abstract submission is pending. 

CONCLUSION 

This study is novel in that we acquired proton and phosphorus 

spectra in the same patient in the same scan session. 

Significant differences in regional metabolite levels were 

detected despite the very small sample size. Larger longitudinal 

studies are needed to replicate these observations. 

Recognizing early metabolic markers of bipolar disorder 

hopefully will lead to more effective developmentally sensitive 

treatment strategies. 

KEY WORDS: MR spectroscopy, bipolar disease, phosphorus 

Dr. Unal will present the paper. 


Functional MR Imaging of Truth and Deception Using a 

Modified Positive-Control Technique and Polygraph 

Correlation 

Faro, S. H. 1 · Mohamed, F. B. 1 · Platek, S. 2 · Gordon, N. 3 · 

Williams, M. 2 · Ahmed, H. 2 

1 2 Temple University, Philadelphia, PA, Drexel University, 

Philadelphia, PA, 3Academy of Scientific and Investigative 

Training, Philadelphia, PA 

PURPOSE 

The purpose of this study was to investigate the regions of 

brain activation during deception or truth by functional MR 

imaging using blood oxygenation level dependent (BOLD) 

contrast using a modified positive-control question technique 

(PQT) and compare the results with a standard polygraph 

examination. 


The experiments were performed on 11 healthy volunteers 

using a 1.5 T scanner. Informed consent was obtained and 

the local IRB approved the study. The physiologic responses 

from the normal subjects were measured by using a fourchannel 

polygraph machine. The functional MR imaging 

experiment used a box-car type block design. The auditory 

stimulus was controlled from outside the scanner using neuropsychological 

software. The subjects’ responses also was 

measured. Initially, contiguous axial high-resolution T1weighted 

images were acquired parallel to the AC-PC line. 

Functional images were acquired with echo-planar sequence 

in the same plane as the structurals. Additional parameters 

were: matrix = 128 * 128; FoV = 22 cm; st = 5 mm; TR = 4 

s; and TE = 54 ms. A relevant situation was created prior to 

the fMRI and polygraph testing. Of the eleven subjects, five 

were asked to tell the truth, that they were not involved in the 

relevant situation (shooting a toy gun), and six were asked to

deliberately lie and deny their involvement in the relevant 

situation. The subjects were presented with 5 separate blocks 

of control (irrelevant) and relevant questions alternating with 

rest period blocks. During each block (24 sec long), 6 volumes 

of EPI images were acquired, yielding a total of 120 

EPI volumes. It was expected that the subjects denying their 

involvement in the relevant situation would produce a 

greater autonomic response to the relevant question than to 

controls. Two separate fMRI experiments were carried out. 

The first trial named “Lie Condition” was carried out to 

compare the brain activity during “known lie” to control 

questions and subjective lie to relevant questions. This was 

followed by another trial named “Truth Condition,” where 

the brain activity during “known truth” to control questions 

and subjective truth to relevant questions were compared. 

The questions were randomized. Subject-level statistical 

analyses were performed using the general linear model in 

SPM2 and group-level random effects analyses for main 

effects were accomplished by entering whole brain contrasts 

into one-sample t-tests. 

RESULTS 

The results show areas of frontal lobe (medial inferior and 

precentral) [BA 9, 10, 6], temporal lobe (hippocampus, middle 

temporal [BA 19], and limbic lobe (anterior and posterior 

cingulate) to be significantly active (p < 0.005, cluster 

threshold > 10) during the deception process. During truth 

telling, activation regions were seen in the inferior and middle 

frontal [BA 46, 10], inferior temporal [BA 20], and cingulate 

gyrus. The polygraph results correlated well with both 

the lying and truth-telling subjects except in one case where 

the polygraph results were inconclusive. 

CONCLUSION 

Overall there seem to be more areas of the brain activated 

during the lying process compared to the truth-telling scenario. 

These results suggest that there may be unique area(s) 

in the brain involved in the truth-telling or deception process 

that can be measured using fMRI. 

KEY WORDS: Functional MR imaging, polygraph, deception 


Pitfalls in Lactate Measurements with MR Spectroscopy 

at High Field Strength 

Dydak, U. 1 · Lange, T. 1 · Alexander, A. L. 2 · Roberts, T. P. L. 3 

· Rowley, H. 2 · Boesiger, P. 1 

1University and ETH Zurich, Zurich, SWITZERLAND, 

2 3 University of Wisconsin, Madison, WI, University of 

Toronto, Toronto, ON, CANADA 

PURPOSE 

Lactate is a metabolically important marker of tissue 

ischemia and hypoxia. Therefore its detection and correct 

quantification by means of MR spectroscopy can be crucial 

for correct clinical diagnosis. This work demonstrates that 

using an echo time of ~144 ms, which is often applied at 1.5 

T, leads to a substantial loss of lactate signal on higher field 

strength MR scanners. 

119 


A standard brain metabolite phantom containing 5 mM of 

lactate was measured on three 3 T MR scanners from three 

different vendors (GE Healthcare, Philips Medical Systems, 

and Siemens Medical Solutions). Single voxel MRS (VOI 

size = 2 x 2 x 2 cm 3 , PRESS localization) was acquired from 

the same volume once with echo time (TE) = 144 ms and 

once with TE = 288 ms. A patient with mitochondrial 

encephalopathy also was studied using these same echo 

times. 

RESULTS 

Based on 1.5 T experience, the lactate peak seen at 1.3 ppm 

at TE = 288 ms should be smaller than the inverted doublet 

at TE = 144 ms due to T 2 relaxation. However, at 3 T spectra 

acquired with the scanners from all three vendors show 

significantly larger lactate signal at TE = 288 ms than at TE 

= 144 ms. This demonstrates that a significant amount of the 

lactate signal is lost (not visible) at TE = 144 ms (Fig1 a, b, 

c - spectra from different scanners are in arbitrary order). 

Further, the example from spectroscopic imaging in the 

patient with mitochondrial disease shows the complete loss 

of the lactate signal at TE = 144 ms, compared to clearly visible 

peaks in the same voxel at TE = 288 ms (Fig 1d). 

CONCLUSION 

Lactate peaks may be smaller than expected or even nondetectable 

on 3 T scanners when using an echo time of 144 ms. 

This potential pitfall is due to the smaller bandwidth of the 

refocusing pulses used at 3 T (1). In vivo this can lead to a 

false-negative result regarding the presence of lactate. 

Although future improvements may address this shortcoming 

on clinical 3 T scanners, at present it is advisable to 

include MR spectroscopy acquisitions with an echo time of 

288 ms when using a high field scanner and lactate is a 

metabolite of concern. 

REFERENCES 

1. Kelley et al. JMRI 1999;(9):732-737 

KEY WORDS: MR spectroscopy, lactate, high field strength 

Acknowledgment: The authors would like to thank Dr. D. 

J.Wang from the Children’s Hospital of Philadelphia, PA, for 

supplying spectra from a Siemens system. 

Tuesday

Tuesday 


3:00 PM - 4:30 PM 

Room 107 

(33c) ADULT BRAIN: Functional and 

Advanced Imaging of Brain 

Neoplasms 









Moderators: Aaron S. Field, MD, PhD 



Differentiation of Low- and High-Grade 

Oligodendroglial Tumors Using Perfusion and Proton 

Spectroscopy MR Imaging 

Spampinato, M. V. · Kwock, L. · Smith, J. K. · Camacho, D. 

L. · Grimme, J. D. · Castillo, M. 

University of North Carolina 

Chapel Hill, NC 

PURPOSE 

To determine if perfusion-weighted imaging and proton MR 

spectroscopy (MRS) may be useful in differentiating 

between high- and low-grade oligodendroglioma and 

oligoastrocytomas. 


Perfusion-weighted imaging and MRS studies were 

reviewed retrospectively in 22 patients with histologically 

proven oligodendroglioma and oligoastrocytomas (seven 

previously treated, fifteen newly diagnosed). Thirteen 

tumors had been characterized pathologically as low-grade 

(eleven oligodendroglioma and two oligoastrocytomas) and 

nine as anaplastic neoplasms (five oligodendroglioma and 

four oligoastrocytomas). Perfusion-weighted imaging was 

performed using a dynamic contrast-enhanced susceptibility-weighted 

echo-planar technique in fourteen subjects. 

Relative cerebral blood volume (rCBV) and relative cerebral 

blood flow (rCBF) within tumors were calculated using standard 

algorithms and normalized to the same values in controlateral 

normal-appearing white matter (rCBV index and 

rCBF index, region-of-interest: 2-4 mm in diameter). Multivoxel 

MRS was obtained using PRESS sequence at TE = 

135 msec in twenty patients and, additionally, with TE = 30 

120 

sec in fourteen patients, with sampling of 0.5-1 cc volume 

elements within tumors and normal-appearing brain. 

Spectroscopy data obtained at long TE were expressed as 

intratumoral metabolites ratios [choline-to-creatine (Cho/Cr) 

and choline-to-n-acetyl aspartate (Cho/NAA)] and normalized 

levels of metabolites within tumor vs normal-appearing 

white matter. In fourteen patients (six low-grade and eight 

high-grade), levels of myoinositol, recorded with MRS at 

short-echo time, were quantified and expressed as intratumoral 

myoinositol-to-creatine (Myo/Cr) ratios, and as ratios 

of levels of myoinositol within the tumor vs normal white 

matter (tumor Myo/normal tissue Myo). Mann-Whitney test 

was used for statistical analysis and results were considered 

significant when p < 0.05. 

RESULTS 

Relative cerebral blood volume index was significantly different 

(p = 0.009) between low-grade (1.52 +/- 0.62) and 

high-grade tumors (5.25 +/- 2.32). Relative cerebral blood 

flow index was also significantly different (p = 0.009) 

between low-grade (1.94 +/- 1.61) and high-grade tumors 

(4.50 +/- 1.97). Analysis of MRS data demonstrated significantly 

higher Cho/Cr ratios (p = 0.002) in high-grade than in 

low-grade tumors. Normalized myoinositol ratios (tumor 

Myo/normal tissue Myo) were significantly higher in low 

grade (4.1 +/- 3.76) than anaplastic (1.21 +/- 0.62) tumors (p 

= 0.02). 

CONCLUSION 

Our data indicate that perfusion-weighted imaging may be 

helpful in differentiating among different histologic grades 

of oligodendrogliomas. In addition, differences in metabolic 

profile demonstrated by MRS can be promising in predicting 

the grade of oligodendroglial tumors. 

KEY WORDS: MR spectroscopy, brain neoplasms, perfusionweighted 

imaging 


Proton MR Imaging and Perfusion MR Imaging of 

Anaplastic Gliomas: Correlation with Histologic 

Subtypes 

Mayo, M. · Chang, J. · Antonietti, L. · Li, X. · Butowski, N. 

· Dillon, W. · Nelson, S. · Vandenberg, S. · Cha, S. 



PURPOSE 

Anaplastic gliomas, WHO grade III, are a heterogeneous 

group of brain tumors with a diverse array of histologic, neuroradiologic, 

and biologic patterns. Recent advances in 

molecular biology foster a new and exciting platform to better 

characterize the histologic subtypes of anaplastic gliomas 

— astrocytic, oligodendroglial, and oligoastrocytic — based 

on molecular and cytogenetic alterations. Perfusion MR 

imaging and proton MR spectroscopic imaging (1H MRSI) 

allow a noninvasive, quantitative characterization of hemodynamic 

measurements and metabolite levels, respectively, 

of brain tumors. The purpose of our study was to separately 

quantify lipid and lactate levels as well as Choline to NAA 

ratio index (CNI) in treatment-naïve anaplastic gliomas

using lactate edited 1H MRSI, and to measure hemodynamic 

variables derived from perfusion MR imaging in order to 

correlate them with the three histologic subtypes. 


Twenty-two patients with newly diagnosed anaplastic 

gliomas were enrolled in the study prior to surgery. The 

patients were grouped based on biopsy-confirmed histologic 

subtype: anaplastic astrocytoma (AA, n = 10), anaplastic 

oligodendroglioma (AO, n = 6), and anaplastic oligoastrocytoma 

(AOA, n = 6). All patients underwent MR imaging on 

at 1.5 T Signa Echospeed scanner (GE Medical Systems, 

Milwaukee, WI). Axial fluid-attenuated inversion recovery 

and postcontrast T1-weighted SPGR images were acquired 

and used to define regions of contrast-enhancing lesion 

(CEL) and nonenhancing T2 lesion (T2L). Dynamic susceptibility 

contrast-enhanced perfusion MR imaging was performed 

using standard dose of Gd-DTPA. Peak height and 

abnormal recovery were measured from DSC T2* relaxivity 

signal-time curve. These variables were measured within 

two distinct anatomical image-defined regions: T2L and 

CEL. Volumes of both regions also were measured. 

Spectroscopic data also was acquired and lactate, lipid, and 

CNI were calculated within these two regions. 

RESULTS 

Anaplastic oligodendroglioma represented the most enhancing 

group of tumor and also showed the highest CNI, followed 

by AOA and AA. Lactate and lipid levels followed 

this pattern as well, with the highest values demonstrated in 

the AO group, followed by AOA and AA. Perfusion characteristics 

showed lowest recovery for AO, followed by AOA 

and AA. Peak height values did not demonstrate differences 

among these three groups. Also of interest was the correlative 

values found between the spectroscopic and perfusion 

characteristics of these tumor types. Within both T2L and 

CEL regions, peak height and abnormal recovery correlated 

strongly with the CNI for AOA. Similarly for AO, abnormal 

recovery within T2L and peak height within CEL correlated 

strongly with CNI. These correlative properties were not 

demonstrated in the AA group. 

CONCLUSION 

Our data suggest that a combination of perfusion and spectroscopic 

imaging may allow for distinction among histologic 

subtypes of anaplastic gliomas. Further study with an 

increased sample size and direct correlation with imageguided 

biopsy will strengthen the validity of our study. 

KEY WORDS: Anaplastic glioma, perfusion MR imaging, 

spectroscopy 


Diffusion Tensor Imaging and Fiber Tractography for 

Characterizing Tumoral Infiltration in the Brainstem 

Lui, Y. W. · Law, M. · Johnson, G. 


New York, NY 

PURPOSE 

Diffusion tensor imaging (DTI) and tractography has been 

demonstrated to provide a method for characterizing the 

anatomical tracts within the brain and brainstem. The pur- 

121 

pose of this study is to determine if DTI and tractography is 

useful in characterizing tumoral infiltration in the corticospinal 

tracts, transverse pontine fibers, medial lemniscus/central 

tegmental tracts within the brainstem as well as 

the brachium pontis and cerebral peduncle. 


Nine patients with brainstem tumors and 9 age- and sexmatched 

normal controls had DTI performed of the posterior 

fossa using a pulsed-gradient echo-planar imaging 

sequence (TR = 4000 ms, TE = 99 ms, 128 x 128 matrix; 240 

x 240 mm FOV, 20 5 mm contiguous slices, b = 1000 s/mm 2 

at 1.5 T). Measurements of mean diffusivity (MD) and fractional 

anisotropy (FA) were obtained within the tumor as 

well as within brainstem white matter tracts which were 

visualized using FA color maps in both patients and controls. 

Fiber tractography was performed also. Comparisons of diffusion 

tensor metrics were made between white matter tracts 

in normal controls and in patients using the Student’s t-test. 

RESULTS 

We studied 9 patients with brainstem tumors: juvenile pilocytic 

astrocytoma (n = 2), low-grade brainstem glioma (n = 

6), anaplastic oligodendroglioma (n = 1). Three patients had 

radiologic diagnosis of low-grade brainstem glioma from 

serial follow-up imaging. The remaining patients had biopsy 

proven pathology. Within enhancing portions of these brainstem 

tumors, MD was significantly higher and FA was significantly 

lower when compared to any of the visualized 

brainstem tracts in the control group (p < 0.05). Within areas 

of abnormal T2 prolongation, MD was higher in patients 

compared with controls in the pyramidal tracts, the transverse 

pontine fibers, and the medial lemniscus/central 

tegmental tracts (p < 0.05). Corresponding FA measures in 

areas of T2 signal abnormality were significantly lower compared 

with controls in the pyramidal tracts, the corticospinal 

tracts, the transverse pontine fibers, and the medial lemniscus/central 

tegmental tracts (p < 0.05). Of the anatomical 

tracts which were evaluated, the tracts found to have the 

highest FA in controls were the brachium pontis and the 

cerebral peduncle, with FA of 0.73 and 0.72, respectively. In 

patients, tensor metrics taken in regions of abnormal T2 signal 

within the brachium pontis and cerebral peduncle were 

not significantly different compared to controls. 

Measurements taken in patients within the brachium ponti 

which were normal appearing by conventional MR techniques, 

demonstrated increased FA compared to controls (p 

< 0.05). No other differences were seen between diffusion 

tensor metrics taken in norma- appearing brainstem tracts in 

patients versus controls. Fiber tractography was able to show 

that circumscribed lesions such as JPA tended to displace 

brainstem fiber tracts, whereas brainstem gliomas tended to 

infiltrate and/or disrupt these fiber tracts. 

CONCLUSION 

Diffusion tensor imaging and fiber tractography is a useful 

technique for characterizing tumoral infiltration within 

anatomical tracts of the brainstem. This may have potential 

in preoperatively and postoperatively predicting patients’ 

response to surgical and nonsurgical therapy. 

KEY WORDS: Brainstem neoplasm, diffusion tensor imaging, 

tractography 

Tuesday

Tuesday 


Proton 1H MRS Imaging of Glioblastoma Multiforme: 

Correlation with Clinical Outcome 

Chang, J. S. · Antonietti, L. · Butowski, N. · Li, X. · Dillon, 

W. · Nelson, S. · Cha, S. 



PURPOSE 

Proton MR spectroscopic (1H MRS) imaging is a noninvasive 

means of analyzing metabolic integrity within both normal 

and diseased areas of the brain. The purpose of this 

study was to use 1H MRS imaging to separately quantify 

levels of lactate (Lac), lipid (Lip), and Choline to N-acetylaspartate 

ratio index (CNI) in two groups of glioblastoma 

multiforme (GBM) patients differentiated by time to progression 

(TTP) to assess for relationships between clinical 

outcome and metabolic markers. 


Twelve newly diagnosed GBM patients were recruited for this 

study. These patients were divided into two groups based on 

clinical status: rapid progression (or RP, progression before 6 

months, n = 8) and moderate progression (or MP, progression 

after 6 months; n = 4). In addition to preoperative anatomical 

MR imaging, all patients underwent 3D J-difference lactateedited 

MRS imaging using a PRESS volume selection technique. 

The lactate-edited method required two acquisitions in 

each phase-encoding step and provided reliable separated 

quantifications of Lip and Lac. Spectroscopic data were analyzed 

for levels of Lip, Lac, and CNI. The metabolite levels 

were quantified using a software program developed in-house. 

The quantification was based on a z-score, defined as: the 

number of standard deviations (SD) a voxel presented away 

from the control voxels, given a normal distribution with a 

mean of 0 and a SD of 1. A z-score of 4+ for Lac/Lip and 2+ 

for CNI was considered abnormal. The following measurement 

variables were derived from each patient’s 1H MRS 

imaging exam: 1) Number of voxels with significant z-scores 

(sigZ); 2) Average z-score (avgZ); and 3) Max z-score 

(maxZ). The average volume of the T2L hyperintensity captured 

by the PRESS box was similar among both groups. 

MRS imaging voxels contaminated with lipid artifact were 

excluded from analysis. 

Results 

Group score derived from an average # + voxels avg Z max Z 

of each patient’s preoperative exam lac lip cni lac lip cni lac lip cni 

Rapid Progression 26.3 21.8 28.9 6.8 10.6 3.8 11.8 20.1 6.3 

Moderate Progression 19.0 8.0 21.0 6.4 7.8 4.4 10.3 13.1 8.4 

RESULTS 

Overall, the presence of Lac and CNI was comparable 

between the RP and the MP groups. Lip, however, was found 

in higher quantities in the RP group than in MP. This trend 

was observed under all measurement variables, and was particularly 

notable under sigZ. Our preliminary data suggests 

that lipid may predict clinical outcome in GBM patients and 

that lipid is also a better predictor of outcome than Lac or 

CNI. 

CONCLUSION 

The results of this study suggest that metabolic information 

derived from 1H MRS imaging may be a predictive marker 

of clinical outcome in patients with GBM. Further study 

122 

with larger sample size and longer follow up will be conducted 

to further evaluate 1H MRS variables in predicting 

clinical outcome. 

KEY WORDS: Spectroscopy, glioblastoma multiforme, lactate-edited 


Optimized Scan and Injection Protocols for Perfusion 

CT of Brain Tumors: Approaching a Full 3D Analysis 

Schramm, P. 1 · Klotz, E. 2 · Schrey, M. 1 · Erb, G. 3 · Hartmann, 

M. 1 

1University of Heidelberg Medical School, Heidelberg, 

GERMANY, 2Siemens Medical Solutions, Forchheim, 

GERMANY, 3Bracco Altana Pharma GmbH, Konstanz, 

GERMANY 

PURPOSE 

Perfusion CT has been shown to allow differentiation of several 

types of brain tumors by the assessment of cerebral 

blood volume (CBV) and blood-brain barrier (BBB) disturbances. 

So far it has been restricted to a few thick sections 

(typically 10 mm) through the center of the tumor. We studied 

different multiphasic injection protocols using high-density 

contrast material (400 mg iodine/ml) with the aim to 

reduce the slice width of functional parameter maps. In addition 

we investigated the applicability of a new multispiral 

scan protocol allowing full tumor coverage. 


We examined 4 groups of 5 patients each undergoing a 

stereotactic biopsy because of intraaxial brain tumors with 

different scan and contrast injection protocols. The patient’s 

head was fixed in a stereotactic ring. Groups 1, 2, and 3 were 

scanned with a 50 second continuous dynamic scan and a 

slice collimation of 4 * 5 mm. We used 60 ml of high-density 

contrast material (Imeron 400®, Altana, Germany) delivered 

by a dual piston power injector. The injection protocol 

for group 1 was monophasic with 5 ml/s; for group 2: 30 ml 

with 5 ml/s and 30 ml with 2 ml/s; for group 3: 30 ml with 5 

ml/s and 30 ml with 1.5 ml/s. All protocols were followed by 

20 ml saline chaser. Group 4 was examined with a new multispiral 

protocol scanning the whole tumor every 10 seconds 

for 80 seconds. For this group 30 ml of contrast were injected 

with 5 ml/s followed by 50 ml with 1 ml/s. Data were analyzed 

using the Patlak method (Perfusion-CT, Siemens, 

Erlangen, Germany) by creating maps of CBV (indicating 

hypervascularization) and permeability (indicating BBB disruptions). 

RESULTS 

All patients were examined without any side effects. All 20 

studies could be analyzed. The biphasic injection protocols 

consistently showed a relatively constant vascular enhancement 

of more than 200 HU for nearly the full scan duration. 

The image quality of the CBV and permeability maps created 

from dynamic scanning with 5 mm section thickness was 

equivalent to 10 mm sections used before with standard 

monophasic injection of standard density contrast. Despite 

sampling only every 10 seconds, Patlak analysis allowed 

creating 3D parameter maps covering the whole tumor from 

the new multispiral acquisition scheme.

CONCLUSION 

We evaluated the benefit of high-density contrast material 

delivered with various multiphasic injection protocols for perfusion 

CT of intraaxial brain tumors. Using optimized protocols 

allows increasing the spatial resolution of functional 

parameter maps from standard dynamic scanning by a factor 

of 2. Such protocols are the basis for the application of new 

dedicated multispiral protocols that for the first time allow 

covering the whole tumor in one study with perfusion CT. 

KEY WORDS: Perfusion CT, brain tumors, 3D analysis 


Short TE Chemical Shift Imaging of Brain Tumors at 3 T 

Hattingen, E. 

Johann Wolfgang-Goethe University of Frankfurt am Main 

Frankfurt am Main, GERMANY 

PURPOSE 

To evaluate the potential of short TE chemical shift imaging 

(CSI) as a reliable tool for characterization of metabolic 

changes in brain tumors. 


Twenty-one preoperative patients with primary brain tumors 

were examined on a 3 T MR scanner (Allegra®, Siemens) 

with 2D CSI (TR 1500; 28 x 28 matrix (32 x 32 after spatial 

Fourier transformation), 1 cm slice, 2 NSA with k-space 

weighting, 16 min scan time) using echo times of 144 as well 

as 30 ms. Metabolic images of cholin, N-acetyl aspartate 

(NAA) and myoinositol (short TE) were generated online 

with the postprocessing tool of the scanner and offline using 

time domain analysis software (jMRUI, http:// 

www.mrui.uab.es). Linewidth, signal (peak area), and fitting 

accuracy (SD value given by the fitting routine) in normal 

white matter were compared for spectra of short and long 

echo time. The noise was evaluated from the peak-to-peak 

distance in spectral regions without metabolite signals. 

RESULTS 

Short and long TE CSI data showed identical peak-to-peak 

noise. Reducing the echo time from 144 to 30 ms yielded in 

a signal increase of almost two but also showed increased 

linewidth. The fitting accuracy for short TE spectra was 

reduced compared to long TE spectra. For NAA in areas with 

normal brain tissue, the reduction was 30%, resulting in an 

accuracy of approximately 10%. Short TE spectra enabled 

quantification of myoinositol, which was elevated in lowgrade 

astrocytomas. 

CONCLUSION 

Short TE CSI spectra show increased linewidth of singulett 

signals, as well as an elevated background, which originates 

from macromolecules and nonsingulett signals. Thus, the 

accuracy for quantification of choline, NAA and creatin, 

which are important metabolites for noninvasive tumor grading, 

is slightly reduced but still sufficient for diagnostic purposes 

based only on these signals. While offering the same 

diagnostic power as long TE CSI, short TE CSI provides 

additional information on myoinositol other metabolites, 

which may be important for characterization of the tumor. 

KEY WORDS: Short TE chemical shift imaging, 3 T, brain tumor 

123 


Apparent Diffusion Coefficient Values as Predictors of 

Response to Stereotactic Radiosurgery for Brain 

Metastases: Preliminary Observations 

Solle, M. · Camacho, D. L. A. · Morris, D. E. · Castillo, M. 



PURPOSE 

Prior studies show that brain neoplasms with favorable clinical 

response demonstrate increased diffusion in the weeks 

after radiation treatment (1). Experimental brain neoplasms 

show increased diffusion with favorable treatment response 

and a decrease in diffusion associated with tumor recurrence 

(2). To determine if diffusion changes correlated with outcomes 

in the treatment of brain metastases, we measured the 

apparent diffusion coefficient (ADC) values in lesions 

before and after stereotactic radiosurgery (SRS). We hypothesized 

that tumor control after SRS correlates with increased 

diffusion. 


Ten metastases treated with SRS that possessed abnormal 

signal on diffusion-weighted images were identified. 

Metastases that were purely cystic, melanomas, or adjacent 

to surgical sites were excluded. Apparent diffusion coefficient 

values were measured in the enhancing portion of 

lesions and in distant white matter on pre-SRS and follow-up 

studies. For each lesion, a pre-SRS ADC ratio (lesion-to-distant 

white matter) was calculated and used as the baseline for 

subsequent measurements. Tumor control was defined as no 

evidence of local recurrence of neoplasm within 1 year as 

determined by lesion growth, MR spectroscopy, and biopsy, 

when available. Data were pooled into time frames of 0-3, 3- 

6, 6-9, and 9-12 months. The percent change of ADC ratios 

of lesions with tumor control were compared to lesions with 

evidence of recurrence using a one-tailed t test, with P ≤ 0.05 

considered significant. 

RESULTS 

All lesions initially decreased in size after SRS. Four lesions 

demonstrated tumor control; six demonstrated radiographic 

evidence of recurrence within 1 year. Lesions exhibiting 

tumor control demonstrated greater diffusion than recurrent 

lesions at 3-6 months (39.8 ± 8.5% versus -13.2 ± 7.8%, *P 

= 0.001), 6-9 months (49.3 ± 27% versus -10.1 ± 13.2%, **P 

= .05), and 9-12 months post-SRS (68 ± 37.5% versus -15.9 

± 7.7%, ***P = .03). 

Tuesday

Tuesday 

CONCLUSION 

This preliminary study demonstrates that lesions with favorable 

response in the year following SRS show increased diffusion 

when compared to those with evidence of recurrence 

during the same period of time. Apparent diffusion coefficient 

values may play a role in monitoring long-term tumor 

control following SRS. 

REFERENCES 

1. Mardor Y, Pfeffer R, Spiegelmann R, Roth Y, Maier SE, Nissim 

O, Berger R, et al. Early detection of response to radiation 

therapy in patients with brain malignancies using conventional 

and high b-value diffusion-weighted magnetic resonance 

imaging. J Clin Oncol 2003;15;21(6):1094-1100 

2. Chenevert TL, McKeever PE, Ross BD. Monitoring early 

response of experimental brain tumors to therapy using diffusion 

magnetic resonance imaging. Clin Cancer Res 

1997;3(9):1457-1466 

KEY WORDS: Brain metastasis, diffusion-weighted imaging, 

apparent diffusion coefficient 


Effect of Prior Surgery on BOLD Functional MR 

Imaging in the Preoperative Assessment of Brain Tumors 

Kim, M. J. J. · Holodny, A. I. · Hou, B. L. · Peck, K. K. · 

Moskowitz, C. · Bogomolny, D. · Gutin, P. 


New York, NY 

PURPOSE 

BOLD fMRI is a clinically useful technique for the preoperative 

mapping of eloquent cortices in patients with brain 

tumors. The purpose of this study was to determine the effect 

of susceptibility artifacts caused by prior surgery on BOLD 

fMRI accuracy by comparing volumes of activation in the 

primary motor cortex (PMC) of patients with and without 

prior brain surgery. 


The volumes of fMRI activation of the PMC were measured 

for the tumor and nontumor sides in patients with (n = 13) 

and without (n = 30) prior neurosurgery. Statistical comparison 

was performed using paired t-tests and linear regression 

analysis. The location and degree of susceptibility artifact 

was assessed subjectively. 

RESULTS 

No significant difference was found between the mean tumor 

and nontumor side volumes of fMRI activations in patients 

without prior surgery. In patients with prior surgery, the volume 

of activation was significantly smaller on the side of the 

prior operation when compared to the contralateral side (p = 

0.001) (Table). The volume of activation on the side of the 

tumor was also significantly smaller in the patients with 

prior surgery compared to those without prior surgery (p < 

0.001) (Table). Notwithstanding, the PMC was identified in 

all cases, and its location was confirmed intraoperatively. In 

each case, visual inspection of the T2* images demonstrated 

that the decrease in fMRI signal was due to the presence of 

prominent signal dropout from susceptibility artifact caused 

by prior surgery. Also, the difference in the volume of activation 

between the two sides was greater if the artifact was 

124 

closer to the PMC. Therefore, it is imperative to examine the 

T2* images and not just the analyzed fMRI data coregistered 

onto the T1-weighted images, since the susceptibility artifact 

present on the T2*-weighted images is no longer visualized 

on the coregistered images (Fig.). 

TABLE 1: Comparison of volume of activation measurements 

(mm3) between tumor side vs. non-tumor side 

Prior Surgery (n=13) No Prior Surgery (n=30) 

Tumor Side Tumor Side Non-tumor Side Tumor Side Non-tumor Side 

Mean 1045 3442 2655 2858 

Median 823 2974 2278 2539 

Standard Deviation 733 1976 1734 1753 

Range 63 - 2136 807 - 6913 380 - 5585 285 - 6819 

CONCLUSION 

Prior surgery is associated with a decrease in the volume of 

fMRI activation in patients with prior surgery. However, by 

examining the T2* images, the astute radiologist can recognize 

this phenomenon, draw the appropriate conclusions, 

and correctly identify the PMC. 

KEY WORDS: Brain tumor, functional MR imaging 


Discriminating Brain Tumor Recurrence from 

Radiation-Induced Injury Using Diffusion Tensor 

Imaging 

Sundgren, P. C. · Fan, X. Y. · Dong, Q. · Weybright, P. · 

Welsh, R. C. · Chenevert, T. L. 


Ann Arbor, MI 

PURPOSE 

To explore the ability of diffusion tensor imaging (DTI) in 

differentiating radiation-induced brain injury from brain 

tumor recurrence by measuring FA and ADC values in the 

edema surrounding the lesion and in the adjacent normalappearing 

white matter areas outside the region of edema. 


Sixteen patients with a new contrast-enhancing lesion and surrounded 

edema at the site of a previous radiation-treated brain 

tumor were reviewed retrospectively. Standardized regions of 

interest (ROIs) were manually drawn at the following areas: 

edema including tract and nontract white matter, normalappearing 

white matter tract adjacent to the edema, and corresponding 

mirror regions in the contralateral hemisphere as 

internal control. Mean ADC, and FA values, ADC ratios (ADC

values in the pathologic areas to ADC values in “mirror” contralateral 

areas) and FA ratios (FA values of lesions to FA values 

in “mirror” contralateral areas) were compared between 

the two groups (radiation injury versus tumor recurrence). 

RESULTS 

The ADC ratios in the edema in patients with radiation injury 

(mean ± SD, 1.77 ± 0.34) were significantly higher (p < 

0.0001) than those found in patients with recurrent tumor 

(mean ± SD, 1.48 ± 0.27). The ADC values of edema did not 

show significant difference between the two groups. 

Although the FA values of edema in two groups were both 

reduced compared with the internal control, the FA values or 

FA ratios did not show significant difference between the 

two groups. For the normal-appearing white matter tract 

adjacent to the edema, the FA values were significantly 

lower (p = 0.0001) in the patients with recurrent tumor compared 

to those with radiation injury, 0.23 ± 0.04 and 0.34 ± 

0.03, respectively. Also the FA ratios were significantly 

lower (p = 0.017) in the recurrence group compared to the 

radiation injury group, 0.72 ± 0.14, and 0.92 ± 0.13, respectively. 

The ADC values in the normal-appearing white matter 

tract adjacent to the edema were normal compared with 

internal controls in both groups. 

CONCLUSION 

The ADC ratios in the edema in patients with radiation injury 

were higher than in those with tumor recurrence, but their FA 

values or ratios demonstrated no statistically significant difference. 

The higher ADC ratios in radiation-induced injury 

may be due to more increased water content in the edema 

areas. The FA changes within the edema surrounding recurrent 

tumor and white matter tract adjacent to the edema 

could be attributed not only to increased water content, but 

also the tumor infiltration. Abnormal FA values in the white 

matter tract with normal appearance on T2-weighted imaging 

in patients with recurrent tumors may indicate that the 

DTI could provide a helpful method in detecting occult 

white matter tract invasion and the FA measurement could be 

more sensitive than ADC. 

REFERENCES 

1. Zhou XH, et al. Proc Intl Soc Magn Reson Med 2001(9):726 

2. Hein PA, et al. AJNR Am J Neuroradiol 2004(25):201-209 

3. Piece SJ, et al. Clin Radiol 2003(58):445-462 

KEY WORDS: Diffusion tensor imaging, radiation injury, 

recurrent brain tumor 

Acknowledgment: Supported in part by NIH grant CA 

85878. 


Changes in Arterial Spin Labeling Brain Tumor 

Perfusion as a Predictor of Progression or Effective 

Treatment 

Appignani, B. · Wong, E. T. · Wu, J. · Hackney, D. B. · Alsop, D. 

Beth Israel Deaconess Medical Center 

Boston, MA 

PURPOSE 

To utilize arterial spin labeling (ASL) perfusion imaging to 

assess changes in brain tumors as they are treated. 

125 


During a 14-month period, we performed 72 arterial spin 

labeling perfusion examinations on 37 patients with a GE 3 T 

scanner. These examinations were part of an MR evaluation of 

primary and metastatic brain tumors. Nineteen of these 

patients, all of whom had surgical resection of their tumors 

followed by radiation and/or chemotherapy, had two or more 

sequential ASL examinations. There were 15 glioma patients 

and 4 patients with metastatic masses. A subset of 4 highgrade 

glioma patients received experimental chemotherapy 

that included the antiangiogenic drug thalidomide. Arterial 

spin labeling was performed using background suppressed 

continuous arterial spin labeling with a stack of variable density 

spiral readout. Images of the entire brain were acquired at 

a 48 x 64 x 64 on an 18 x 18 x 24 cm FOV. Three ASL averages 

required 5 minutes. Three-dimensional volumes were 

reconstructed offline and resliced to match the gadoliniumenhanced 

T1-weighted scans or T2 images, for simultaneous 

viewing. Interpretation was based on the appearance of normal 

gray matter perfusion signal intensity. Signal greater than 

gray is indicative of hyperperfusion, equal to gray, isoperfusion, 

and less than gray, hypoperfusion. Changes in the standard 

MR imaging T2- and T1-weighted gadolinium-enhanced 

appearance of the masses were compared to changes in perfusion 

signal intensity. 

RESULTS 

During surveillance, there were four patients (3 high-grade 

gliomas, 1metastasis) who developed new or increasing 

extent of hyperperfusion. In these cases, areas of hyperperfusion 

did not correspond directly to areas of gadolinium 

enhancement, nor was there convincing standard radiologic 

evidence of tumor progression. In all, there was surgically 

confirmed tumor. There were intervals (during 4-12 months 

of follow up) where decreased perfusion was observed in 

tumors of three of the four patients receiving thalidomide, 

with no tumor growth on standard MR imaging. The fourth 

of these patients died of unknown causes without change 

after 2 months. In one of the thalidomide-treated patients 

with glioblastoma, there was decreased perfusion observed 

on the first 1-month follow-up ASL exam, and then a focus 

of hyperperfusion developed which subsequently grew to be 

an enhancing recurrent mass, which was resected surgically. 

Hypoperfusion developed in three initially hyperperfused 

metastases, in two patients, which were treated with stereotactic 

radiosurgery, and in one grade III/IV oligoastrocytoma. 

These patients are clinically stable. Additionally, 

seven patients who had hypoperfused tumors (6 gliomas and 

one patient with 2 metastases) or treatment sites are clinically 

well. One patient with a treated grade II/IV astrocytoma 

had a hypoperfused but enlarging and newly enhancing 

mass. No pathologic tissue has been obtained. One patient 

had an isoperfused ganglioglioma which remained stable for 

over 1 year. 

CONCLUSION 

A change of perfusion in a tumor treatment site may be 

indicative of a change in the viability or development of the 

neoplastic process. New or increasing hyperperfusion was 

indicative of progressive neoplasm in a small group of 

patients we have followed. Hypoperfusion of the treatment 

site and decreasing perfusion correlate with response to therapy 

and stability. 

KEY WORDS: ASL, tumor, perfusion 

Tuesday

Tuesday 


Effect of MR Contrast Dose on Pituitary Gland 

Enhancement, Microlesion Enhancement, and Gland-to- 

Lesion Contrast Conspicuity 

Bartynski, W. S. · Boardman, J. F. · Grahovac, S. Z. 

Presbyterian University Hospital 


PURPOSE 

To compare pituitary gland enhancement, microlesion 

enhancement and gland-to-lesion contrast difference when 

using half dose (HD), standard dose (SD) and double dose 

(DD) contrast administration in pituitary MR imaging. 


Pituitary microlesions were identified by MR imaging in 19 

patients who received HD (0.05 mmol/kg ), 10 patients who 

received SD (0.1 mmol/kg ) and 15 patients who received 

DD (0.2 mmol/kg ) contrast infusion for focused pituitary 

evaluation. Gland and microlesion enhancement over baseline 

was determined employing region of interest measurements 

and compared after normalization to temporal lobe 

white matter. Gland-to-lesion contrast ratios and differences 

were calculated and compared. 

RESULTS 

The pituitary gland and microlesions demonstrated greater 

enhancement with larger administered contrast dose (gland: 

HD - 51%, SD - 96%, DD - 132%; microlesion: HD - 16%, 

SD - 48%, DD - 79%). Relative gland-to-lesion contrast ratio 

(corrected for outlying cases: 3 slowly enhancing adenomas 

and 1 cyst) was proportional at the three contrast doses 

reflecting an expected fractional contrast distribution 

between gland and microlesion. Signal difference between 

gland and microlesion was a fixed percentage of gland 

enhancement (26%) or lesion enhancement (37.3%) with 

therefore greater contrast difference at larger contrast dose. 

CONCLUSION 

Larger contrast dose results in greater pituitary gland and 

microlesion enhancement in pituitary MR imaging. Glandto-lesion 

contrast ratio remains proportional at the three dose 

levels studied suggesting absolute gland-to-lesion enhancement 

differences will be greater at larger contrast dose levels. 

Choice of contrast dose is likely and additional important 

factor to consider in pituitary microadenoma detection 

and characterization. 

KEY WORDS: Pituitary gland, MR imaging, contrast dose 

126 


3:00 PM - 4:33 PM 

Room 205 

(33d) ADULT BRAIN: Degenerative, 

Dementias, and Destructive Lesions 









Moderators: John R. Hesselink, MD, FACR 

Meng Law, MD 


MR Imaging-Based Volumetric Analysis of Basal 

Ganglia Nuclei and Substantia Nigra in Patients with 

Parkinson’s Disease 

Li, Y. 1 · Geng, D. 2 · Zee, C. S. 2 

1Hua-shan Hospital, Fudan University, Shanghai, CHINA, 

2Keck School of Medicine, University of Southern 

California Los Angeles, Los Angeles, CA 

PURPOSE 

Parkinson’s disease (PD) is associated with well documented 

morphologic changes in substantia nigra and basal ganglia 

nuclei. This study evaluates the ability of MR imaging 

to detect these changes, and investigates the relationship 

between severity of disease and degree of morphologic 

change. This coorelation may be valuable in early diagnosis 

of Parkinson’s disease. 


Sixteen patients with early stage PD, eight patients with 

advanced PD, and eight normal controls were studied by 3 T 

MR imaging. The whole brain volume and the volumes of 

the caudate, putamen, globus pallidus, and substantia nigra 

were calculated on 3D reconstructed imagings. 

RESULTS 

Putamen volume was diminished significantly in patients 

with early PD and advanced PD compared with that in controls 

(p < 0.05), and the percentage of atrophy was 12.5% 

and 26.5% respectively. The putamen volume was negatively 

correlated with Hoehn & Yahr staging (r = -0.720, p < 

0.001). Pallidal volume reduced only in advanced PD (p = 

0.023). There were no significant differences in total brain, 

caudate, or substantia nigra among the three groups

CONCLUSION 

MR imaging-based volumetry is a sensitive method in the 

assessment of morphologic changes of PD in vivo. The putamen 

atrophy was correlated with the severity of the disease. 

The volumetric measurement of putamen could potentially 

be a useful method for the diagnosis of PD in early stage. 

KEY WORDS: Parkinson’s disease, volumetric analysis, MR 

imaging 


Change of Apparent Diffusion Coefficient in Patients 

with Parkinson’s Disease 

Park, J. K. · Jeong, A. · Choi, S. · Kang, B. · Hwang, J. · Lee, J. 

Ulsan University Hospital 

Ulsan, REPUBLIC OF KOREA 

PURPOSE 

Patients with Parkinson’s disease (PD) commonly show 

histopathologic changes in their substantia nigra (SN) and 

basal ganglia. This study was designed to evaluate the effect 

of PD to the apparent diffusion coefficient (ADC) of these 

structures. 


We examined 30 patients with PD and 64 healthy agematched 

controls with diffusion-weighted image on a 1.5 T 

MR system (Horizon Lx, GE, Milwaukee, WI). We measured 

ADC in the SN, globus pallidus (Gp), and putamen. We 

assessed any significant difference of ADC between the 

patients and controls, between the dominant side and the 

nondominant side in patients. We evaluated the correlation 

of disease duration and ADC of SN and Gp. 

RESULTS 

Comparison between patients and controls yielded significantly 

higher ADC (10 -6 mm 2 /s) in the both sides of SN (854 

vs 781 in right, 859 vs 785 in left, P < 0.001), both sides of 

globus pallidus (782 vs 739 in right, p = 0.002, 801 vs 752 

in left, p = 0.001) and putamen (787 vs 718 in right, p = 

0.003, 777 vs 726 in left, p = 0.009). In 22 patients, the ADC 

of dominant side was higher than the ADC of nondominant 

side, but the difference was not statistically significant (864 

vs 846, p = 0.1). There was no significant correlation 

between the disease duration and the ADC of SN and Gp 

(Pearson correlation, r = -0.02 in SN, r = 0.32 in Gp). 

CONCLUSION 

Significantly higher ADC in SN, Gp, and putamen was present 

in the patients with PD. These results reflect the loss of 

tissue integrity in the SN and BG in PD. Parkinson’s disease 

may affect both sides of the SN with relatively even involvement. 

KEY WORDS: Parkinson’s disease, apparent diffusion coefficient, 

substantia nigra 

127 


MR Spectroscopy Changes in the Frontal Lobes Occur in 

Presymptomatic Huntington’s Disease and Correlate to 

Oculomotor Dysfunction 

Gomez-Anson, B. M. 1,2 · Alegret, M. 3 · Muñoz, E. 1 · Sainz, 

A. 3 · Tolosa, E. 1,2 

1 2 Hospital Clinic, Barcelona, SPAIN, Institut 

d’Investigacions Biomediques August Pi i Sunyer, 

Barcelona, SPAIN, 3Fondacio/Hospital Clinic, Barcelona, 

SPAIN 

PURPOSE 

Metabolic changes have been demonstrated in the brain of 

patients with Huntington’s disease (HD) using MR spectroscopy 

(MRS). However, time of onset of these changes 

and functional significance are currently still unknown. To 

investigate the presence of brain metabolic abnormalities 

and neuropsychologic deficits in presymptomatic HD subjects 

(pre-HD). 


Seventeen genetically proven pre-HD and 17 age-matched 

(mean age 32 ± 8 years) controls (mean age 35 ± 6 years) 

with a similar educational level were included. Single-voxel 

proton MR spectra (PRESS, TR =6000, TE = 30) were 

obtained in the basal ganglia and frontal cortex (Fig. 1A) in 

13 pre-HD subjects and in all controls using a GE 1.5 T MR 

unit and the head coil. Quantitation of absolute metabolite 

concentrations and metabolite ratios was performed using an 

automated, external reference method (LC Model v 5.2-3X). 

The neuropsychologic battery comprised memory (Rey’s 

auditory-verbal learning, Rey’s complex figure), visuomotor 

speed (WAIS-III symbol search), oculomotor (the 15 

objects), premotor (Luria’s motor alternances), frontal 

(WAIS-III digits, Stroop, Trail Making, phonetic and semantic 

verbal fluency, and visuospatial (judgment of line orientation) 

tests. Statistical assessment using SPSS included a ttest 

with independent groups and ANOVA with repeatedmeasures 

analysis of variance. 

RESULTS 

Pre-HD subjects showed a significant decrease in frontal 

choline-containing compounds levels (Cho) when compared 

to controls (1.26 +/- 0.22 vs 1.54 +/- 0.30, p = 0.008). Also, 

there was a significant slowing in the oculomotor test scores 

(the 15 objects test, p = 0.004) in pre-HD when compared to 

controls. Performance in the 15 object test (time needed to 

achieve the task) for all subjects correlated significantly and 

negatively to frontal Cho (p = 0.001; R = -0.583). 

Tuesday

Tuesday 

Fig. 1. 1 H-MRS in the frontal cortex. 1a) SVS (VOI = 2 x 2 

x 2) located in the frontal cortex; 1b) Postprocessed (LC- 

Model) spectrum obtained in a control subject (Cho = 1.97 

mM; SD = 9%); 1c) Postprocessed (LC-Model) spectrum 

obtained in a pre-HD subject shows decreased choline-containing 

compounds levels (Cho = 0.958 mM; SD = 14%). 

CONCLUSION 

Our results suggest that metabolic impairment reflecting 

functionally altered connections between basal ganglia and 

frontal cortex is present already in pre-HD subjects. These 

disturbances could explain the asymptomatic ococulomotor 

dysfunction detected in these subjects. 

KEY WORDS: MR spectroscopy, Huntington´s disease, 

choline 


Functional MR Imaging in Presymptomatic 

Huntington’s Disease Reveals Global Hypoactivation 

with Motor and Verbal Fluency Tasks 

Gomez-Anson, B. M. 1,2 · Sainz, A. 3 · Alegret, M. 1 · Muñoz, 

E. 1 · Tolosa, E. 1,2 

1 2 Hospital Clinic, Barcelona, SPAIN, Institut 

d’Investigacions Biomediques August Pi i Sunyer, 

Barcelona, SPAIN, 3Fondacio/Hospital Clinic, Barcelona, 

SPAIN 

PURPOSE 

Although metabolic, structural and pathologic changes have 

been demonstrated in the brain of Huntington’s disease (HD) 

patients, it is still unknown when these changes develop. 

Functional MR imaging (fMRI) may be a valuable tool in 

evaluating early alterations in brain activation patterns pre- 

HD subjects. To search for early changes in brain activation 

patterns in pre-HD subjects using fMRI and motor and verbal 

tasks. 


We studied 13 genetically proven pre-HD (mean age 32 ± 8 

years) and 17 age-matched controls (C) (mean age 35 ± 6 

years, 3 left-handers) with a similar educational level. Realtime 

fMRI studies (TR = 3000, 17 slices, 1700 images) were 

obtained using a 1.5 T MR unit, and the head coil in all subjects. 

Both motor (Luria’s test) and verbal fluency paradigms 

were used, and studies were analyzed using SPM2. 

Statistical assessment included a one sample t-test for group 

analysis and a fixed effects model (two sample t-test) for 

comparison between groups. Significance was set at an 

uncorrected p level < 0.001, and an extent threshold > 40 

voxels. 

RESULTS 

Pre-HD subjects showed a significantly decreased, but similar 

activation pattern as controls in both motor and verbal 

fluency paradigms. Using the motor paradigm, pre-HD subjects 

showed significantly decreased activations in the motor 

cortex, SMA, right caudate, and right posterior temporal 

regions than controls (Fig. 1A). Using the verbal fluency 

paradigm, pre-HD subjects showed a significantly diminished 

cluster of activation in the left inferior frontal/pre- 

128 

frontal region when compared to controls (Fig. 1B). Similar 

results were obtained when only right-handers were considered 

for the analysis. 

Fig. 1. Results obtained after postprocessing fMRI studies 

with SPM2, and using a t-test for comparison between activation 

patterns of control and pre-HD subjects. Statistical 

significance was defined at the uncorrected p < 0.001 level, 

and a extent threshold of 40 voxels: 1a) Using the motor task 

(Luria’s test), and looking at responses when moving the 

right hand, there is less significant activation in the motor 

cortex, SMA, and right caudate in pre-HD subjects compared 

to controls; 1b) Using a verbal fluency paradigm, there 

is decreased activation in the left prefrontal/inferior frontal 

cortex in pre-HD subjects compared to controls. 

CONCLUSION 

Our results suggest that there is a general hypoactivation of 

normally occurring networks in pre-HD subjects. These 

changes may be a consequence of impaired basal gangliacortex 

connections that occur even years before the clinical 

symptoms of the disease appear. 

KEY WORDS: Functional MR imaging, Huntington´s disease, 

verbal 


Brain and Abdominal MR Imaging in Patients with 

Neurologic Presentation of Wilson’s Disease 

Kozic, D. B. 1 · Petrovic, I. 2 · Svetel, M. 2 · Semnic, R. 1 · 

Kostic, V. 2 

1Institute of Oncology, Sremska Kamenica, SERBIA AND 

2 MONTENEGRO, Institute of Neurology, Belgrade, 

SERBIA AND MONTENEGRO 

PURPOSE 

The aim of the study was to correlate the abdominal MR 

findings in patients with neurologic presentation of Wilson’s 

disease (WD) with the degree of brain lesions resolution, 

during the chronic course of chelating therapy. 


Brain MR reexamination was performed at 1.5 T imager in 

15 patients with neurologic form of WD, 3.5-7 years following 

the first MR examination. Additional abdominal MR 

imaging was done in 13 patients at the time of control brain 

MR examination.

RESULTS 

No significant abnormalities were seen on abdominal MR 

imaging in all 5 patients in whom the period between the 

appearance of clinical symptoms and correct diagnosis was 

up to 18 months, while 87% of patients with 2-5 years delay 

presented with portal hypertension (splenomegaly and less 

frequently macronodular cirrhosis or peritoneal effusion). 

Complete reversal of putaminal T2-weighted signal alterations 

was noted in 50% of patients with normal abdominal 

MR imaging while no improvement was evident in 87% of 

patients with portal hypertension. Complete regression of 

pontine base and pontine tegmentum lesions was seen in 

75% and 100% of patients with no abdominal MR abnormalities 

and short delay to correct diagnosis while majority 

of patients with longer delay and portal hypertension displayed 

significant but incomplete resolution of pontine 

lesions. Marked but incomplete regression of T2-weighted 

signal abnormalities was seen in midbrain, with no significant 

difference between the patients with normal abdominal 

MR imaging and the patients with radiologic signs portal 

hypertension. 

CONCLUSION 

Patients with neurologic presentation of WD and short (up to 

18 months) delay to correct diagnosis presented with no portal 

hypertension and more likely reversible putaminal and 

pontine T2-weighted signal alterations. Longer delay to 

diagnosis correlated with portal hypertension and less likely 

reversible putaminal changes (gliosis). Brain stem lesions 

appear to be reversible during the chronic course of chelating 

therapy even in patients with longer delay to correct 

diagnosis, and are most consistent with treatable inflammatory 

changes, edema and demyelination, probably due to 

impaired liver function or copper intoxication. 

KEY WORDS: Wilson’s disease, brain, portal hypertension 


Diffusion Tensor Imaging of the Basis Pontis Enables 

Early Diagnosis of Pontocerebellar Degeneration that Is 

not Possible on Conventional MR Imaging 

Yim, C. · Salamon, N. · Lazo, C. · Sinha, U. · Perlman, S. 

University of California Los Angeles 


PURPOSE 

Diffusion tensor imaging (DTI), a technique capable of 

examining water diffusion in the brain, can be used to evaluate 

and quantify early atrophy in the basis pontis of patients 

with clinically diagnosed pontocerebellar degeneration that 

is not yet visible on conventional MR imaging. Diffusion 

tensor imaging was used to analyze in vivo the neuropathology 

of fiber tracts of the basis pontis, the common pathway 

for cerebral-cerebellar communication. Comparisons of fractional 

anisotropy (FA) measurements were made between 

patients with pontocerebellar degeneration due to sporadic 

cerebellar ataxia (SCA) and patients with cerebellar degeneration 

without basis pontis atrophy due to drug toxicity as 

seen with epilepsy. 

129 


Ten patients with sporadic cerebellar ataxia (SCA) (n = 10) 

and ten patients with intractable epilepsy (n = 10) were 

enrolled in this study. The epilepsy patients are all undergoing 

chronic antiepileptic drug therapy half (n = 5) of which 

have visible signs of cerebellar atrophy on MR imaging presumably 

due to intrinsic drug toxicity and damage to the 

Purkinje cells and pyramindal cells, and half (n = 5) of which 

have normal appearing cerebellar hemispheres. All of the 

SCA patients have cerebellar atrophy on MR imaging. Using 

a 1.5 T Siemens Sonata scanner, diffusion tensor imaging 

was performed which consists of an axial 2D EPI diffusionweighted 

sequence is acquired with diffusion gradients along 

six noncollinear directions at three b values of 0 sec/mm 2 , 

600 sec/mm 2 and 1200 sec/mm 2 , Matrix: 256 x 256, slice 

thickness: 5 mm. The diffusion-weighted images are analyzed 

off-line to generate maps of the diffusion tensor using 

software written in IDL (IDL, Research Systems, Boulder, 

CO). Color maps reflecting the relative magnitude of the 

eigenvalues were generated and fractional anisotropy (FA) 

and mean diffusivity (ADC) were obtained on a voxel-byvoxel 

basis. Regions of interest (ROI) were drawn independently 

on the color maps outlining white matter in the 

basis pontis, cerebellar hemispheres, superior cerebellar 

peduncle, and middle cerebellar peduncle by a single operator 

who evaluated adjacent sections to ensure accurate slice 

selection. Statistical analysis was performed with the independent 

t-test to compare the differences in FA and ADC values. 

A p-value of < 0.05 will be considered statistically significant. 

RESULTS 

Fractional anisotropy values of the white matter tracts in the 

basis pontis were reduced significantly in patients with pontocerebellar 

atrophy associated with SCA when compared to 

patients with epilepsy. Apparent diffusion coefficient values 

were elevated in all patients with frankly visible imagingdetectable 

atrophy, regardless of cause. Moreover, ADC values 

were normal in epilepsy patients who had no visible atrophy 

as well as in patients who have clinically diagnosed pontocerebellar 

degeneration but have not yet manifested visible 

morphologic changes of cerebellar atrophy. 

CONCLUSION 

Diffusion tensor imaging can be a valuable tool for the in 

vivo analysis of white matter tracts in the basis pontis that 

form part of the cerebral-cerebellar communication pathway. 

Measuing FA values in the basis pontis can allow distinction 

of cerebellar atrophy due to extrinsic factors such as drug 

toxicity versus intrinsic causes such as sporadic cerebellar 

ataxia. Early pontocerebellar degeneration can be diagnoses 

by DTI where conventional MR imaging fails. 

KEY WORDS: Diffusion tensor imaging, cerebellar atrophy, 

transverse pontine fibers 

Tuesday

Tuesday 


Imaging Alzheimer’s Disease Pathology in Humans 

Using Spin Lock Technique (T 1 Rho Imaging) 

Moonis, G. · Borthakur, A. · Clark, C. · Melhem, E. · Reddy, R. 



PURPOSE 

Conventional MR imaging techniques have proven inadequate 

in observing the actual senile plaques (SP) and neurofibrillary 

tangles (NFT) in vivo underlying the pathology 

of Alzheimer’s disease (AD). An alternate contrast mechanism 

to conventional T1- and T2 weighted images is T1r, or 

“T-1-rho,” the spin-lattice relaxation time constant in the 

rotating frame, which determines the decay of the transverse 

magnetization in the presence of a “spin-lock” radio-frequency 

field. In previous studies, differences in T 2 relaxation 

times could not discern differences between patient and control 

populations. However, T1r relaxation is less prone to 

effects from diffusion and susceptibility and therefore presents 

a greater dynamic range of values compared to T 2 in 

biologic tissues. Consequently, T1r MR imaging has shown 

promise in delineating tumors, gliomas, and other cancerous 

tissue. The purpose of our study was to determine T1r relaxation 

times in vivo in patients with AD and compare these 

values against age-matched controls. 


MR imaging was performed on six patients (mean age: 77 ± 

7 years) with AD and six controls (mean age: 73 ± 7 years). 

An oblique coronal T1r-weighted image of a slice perpendicular 

to the AC/PC plane was obtained. This slice was chosen 

to include the head of the hippocampus. MR Imaging 

Parameters: T1r preencoded turbo spin-echo (TSE) pulse 

sequence with TE/TR = 12/2000 ms, TSL (duration of spinlock 

pulse) = 20, 40, 60, and 80 ms. Slice thickness - 2 mm, 

FOV = 22 cm, NEX = 1, Matrix = 256 x 128, ETL = 4. Total 

imaging time of 6 minutes. Each image pixel’s signal intensity 

was fitted as a function of TSL by a linear least-squares 

algorithm to generate T1r maps. A single user manually 

selected a region of interest in each map in the medial temporal 

lobe region and recorded average T1r values. 

Statistical analysis was performed with the JMP software 

package. A Student’s t-test was performed to determine any 

significant difference between the values obtained in patients 

and controls. 

RESULTS 

The average T1r for patients was 95.3 ± 1.4 ms (mean ± std. 

error) and for controls was 87.5 ± 1.7 ms and the difference 

was statistically significant (p < 0.005). 

CONCLUSION 

This is the first demonstration of imaging AD pathology in 

humans using T1r imaging. Our results indicate that the 

macromolecular changes underlying AD (including SP, NFT, 

edema/gliosis) results in the prolongation of the T1r relaxation 

time of brain tissue. Future applications of this technique 

would include attempts to categorize the patient population 

into incipient AD (MCI) or advanced AD on the basis 

of T1r times. 

130 

KEY WORDS: Alzheimer’s disease, spin lock technique, T1 

rho imaging 


Individual Statistical Volumetric Map (3D SSP GM) as a 

Diagnostic Aid for Mild Cognitive Impairment and Mild 

Alzheimer’s Disease 

Brinkman, W. J. · Cohen, W. · Anzai, Y. · Shibata, D. · Cross, 

D. · Minoshima, S. 


Seattle, WA 

PURPOSE 

Early diagnosis of Alzheimer’s disease (AD) permits drug 

interventions in an early stage of the disease before substantial 

neuronal loss occurs. There is accumulating evidence 

that early drug intervention appears to result in better prognosis. 

However, clinical diagnosis of early AD is still challenging. 

MR imaging has been used to exclude certain 

pathologies in dementia patients and to provide the assessment 

for regional atrophy such as the hippocampus for a 

potential diagnostic aid. Recent investigations of a group of 

AD patients also have demonstrated more widespread limbic 

and neocortical abnormalities in very early AD. The current 

investigation is to examine if such neocortical and limbic 

abnormalities can be detected in individual patients of early 

AD using voxel-based statistical assessment of regional gray 

matter (GM) density and can be used as a potential diagnostic 

marker. 


Twenty-two patients (mean age 69 years) with mild cognitive 

impairment (MCI, Clinical Dementia Rating 0.5) and 

mild AD (CDR 1), both confirmed to have probable AD by 

NINCDS/ADRDA criteria, and 6 age-matched controls 

(mean age 67 years) underwent 3D SPGR MR imaging. 

Following gray matter segmentation, each image set was 

converted to GM density by 3D spatial filter convolution, 

transformed to the bicommissural stereotactic coordinate 

system, and extracted to 3D stereotactic surface projections 

(3D SSP) format using automated software (NEUROSTAT, 

University of Washington). The 3D SSP GM maps from 

individual patients then were compared to a normal database 

created from 9 age-similar controls on a voxel-by-voxel 

basis, resulting in individual Z-score maps depicting regions 

of gray matter loss. Three observers with expertise in neuroimaging 

interpreted Z-score maps for the pattern of GM 

loss with confidence grading. Quantitative assessment of 

hippocampal volume loss was performed for comparison. 

Decision performance was evaluated using receiver operator 

characteristic (ROC) analysis. 

RESULTS 

Three-dimensional SSP GM demonstrated disproportionate 

GM loss (Z > 4) in the medial temporal lobe, the posterior 

cingulate gyrus, and the temporoparietal association cortices 

in MCI and mild AD patients. On average, hippocampal volume 

was reduced by approximately 7% in patients with MCI 

and 21% in patients with mild AD compared to age-matched 

controls. Collectively the observers were able to discriminate 

between all patients versus normal age-matched controls 

with 93% sensitivity and 100% specificity using 3D 

SSP GM Z-score maps. The sensitivity for the detection of

MCI and mild AD was 85% and 100%, respectively. The 

typical pattern suggestive of AD (preferential loss in the 

medial temporal lobe, posterior cingulate cortex, and/or temporoparietal 

association cortices) was more often recognized 

in mild AD (86%) in comparison to MCI (74%). 

CONCLUSION 

Three-dimensional SSP GM Z-score maps depict statistically 

regional cortical GM loss in individual MCI and mild AD 

patients that can be interpreted accurately by observers. Such 

diagnostic gain can be attained with conventional 3D SPGR 

MR imaging and minimal postacquisition processing time 

(typically < 5 min). Although MCI patients can be clearly 

discriminated from age-similar normals by 3D SSP GM, the 

pattern typical for AD is seen less frequently in MCI, indicating 

possible heterogeneity in clinically diagnosed MCI 

patients. 

KEY WORDS: Alzheimer’s disease, statistical volumetric 

map, mild cognitive impairment 


Fast and Robust Quantification of Temporal Horn 

Volume in Patients with Mild Cognitive Impairment and 

Alzheimer’s Disease Using MR Imaging 

Giesel, F. L. 1,2 · Hahn, H. 3 · Thomann, P. 4 · Wignall, E. 2 · von 

Tengg-Kobligk, H. 1 · Schröder, J. 4 · Wilkinson, I. D. 2 · 

Griffiths, P. D. 2 · Essig, M. 1 

1German Cancer Research Center (Deutsches 

Krebsforschungszentrum), Heidelberg, GERMANY, 

2University of Sheffield, Sheffield, UNITED KINGDOM, 

3 4 MeVis, Bremen, GERMANY, University of Heidelberg, 


PURPOSE 

Psychopathologic symptoms and cognitive deficits are the 

primary outcome measure for therapeutic trials in geriatric 

psychiatry. Objective quantification of brain structure can 

additionally aid diagnosis and therapeutic monitoring. In a 

previous study, atrophy of the medial temporal lobe was 

established as an early biomarker for Alzheimer’s disease 

(AD) (Pantel, et al. 2003). In the present study, we aimed to 

quantify temporal horn volume (THV) as an indirect regional 

measurement for hippocampal formation change in 

patients with mild cognitive impairment (MCI) and AD. 


Ten volunteers (ages 66 +/-1), ten patients with MCI (ages 

66+/-1), and eight with AD (ages 64+/-10) were enrolled. All 

subjects completed a battery of cognitive tests and were 

imaged at 1.5 T (Vision, Siemens, Germany; T1-weighted 

coronal TR = 4 ms, Flip = 13°, FOV = 250 mm, Matrix = 256 

x 256, 128 contiguous slices, 1.8 mm). Temporal horn volume 

was calculated using a watershed algorithm-based software 

package (MeVisLab, MeVis, Germany). 

Postprocessing time, per data set, was less than 3 minutes 

and the ventricular volumes were correlated with clinical 

assessments. 

RESULTS 

Successful segmentation was performed on all subjects, 

examples of each group are shown below (a = age matched 

control, b = MCI, c = AD). A significant difference in THV 

131 

was found between controls and AD patients (THV 3.48 +/- 

2.03 ml, MMSE 17.3 +/- 4), (p < 0.001). Group analysis 

showed a slight, but not significant increase in THV for MCI 

patients (THV 0.86 +/- 0.59 ml) compared to controls (p < 

0.07). In contrast THV, of the MCI group was significantly 

less than that of the AD patients (p < 0.001). 

CONCLUSION 

The fast indirect quantification of hippocampal formation 

volume using THV is feasible. Differences were observed 

between the three groups. Further evaluation of this methodology 

is warranted to establish its role in diagnosis and monitoring 

of disease status in general psychiatry. 

KEY WORDS: Alzheimer´s disease, postprocessing, temporal 

horn 


Neuroimaging of Genetically Defined Incipient 

Creutzfeldt-Jakob Disease 

Hoffmann, C. 1 · Fulbright, R. 2 · Guo, X. 3 · Schmidler, J. 3 · 

Nizan, Z. 1 · Chapman, J. 1 · Kahana, E. 1 · Kingsley, P. 4 · 

Prohovnik, I. 2 

1 2 Sheba Medical Center, Ramat-Gan, ISRAEL, Yale 

University, New Haven, CT, 3Mount Sinai School of 

Medicine, New York, NY, 4North Shore University Hospital, 

Manhasset, NY 

PURPOSE 

Creutzfeldt-Jakob disease (CJD) is the most notable of the 

human prion diseases, a group of severe, fatal neurodegenerative 

pathologies, which are transmissible (and, uniquely, 

also inherited). The Israeli cluster of hereditary CJD among 

Libyan Jews is the largest in the world, and it is associated 

with a mutation of the PRNP gene on chromosome 20. 

Neuroimaging has been attempted in CJD to elucidate the 

cerebral pattern of injury, but the literature largely consists of 

case reports and variable findings at different stages of the 

disease and in heterogeneous patient samples. 


A new study just has been funded by the NIH to study CJD 

in Israel. We will use neuroimaging to elucidate early, and 

even premorbid, cerebral abnormalities of structure and 

function in a singular cluster of high incidence occurring 

among Libyan Jews living in Israel, caused by familial transmission 

of a mutated prion protein (PrP) gene. All subjects 

will have extensive neurologic and neuropsychological 

examinations, as well as MR imaging, using both traditional 

structural imaging and newer neuroimaging methods: diffusion-weighted 

imaging (DWI) and MR spectroscopy (MRS). 

Here we present the results from the first seven subjects (4 

patients and 3 relatives). 

Tuesday

Tuesday 

RESULTS 

On clinical reading of the MR images, all three controls were 

judged normal, and all four patients were considered abnormal. 

The most common findings were FLAIR and DWI 

hyperintensity in caudate (4/4), putamen (3/4), and cortical 

white matter (3/4). Quantitative analyses showed significant 

loss of gray matter and higher ADC values in cortex. Focal 

decreases of ADC were found in the lentiform nucleus of 

patients. Spectroscopic imaging demonstrated reduced NAA 

in the patients, with the NAA/Cho ratio lowest in cingulate 

gyrus, where it was correlated also with a neurologic severity 

score. Duration of disease was correlated with this ratio in 

the caudate nucleus and putamen. 

CONCLUSION 

Both structural MR imaging and newer neuroimaging methods 

(MRS and DWI) were significantly abnormal, confirming 

and extending previous work. These preliminary data 

suggest that MR imaging abnormalities can be found early in 

the course of the disease, and that the radiologic appearance 

of this genetic cluster is identical to the classic sCJD. 

KEY WORDS: Creutzfeldt-Jakob disease, MR spectroscopy, 

diffusion-weighted imaging 


Frequency and Distribution of MR Imaging Changes in 

Sporadic Creutzfeldt-Jakob Disease 

Tschampa, H. J. 1 · Kallenberg, K. 2 · Kretzschmar, H. A. 3 · 

Knauth, M. 2 · Zerr, I. 2 · Urbach, H. 1 

1 2 University of Bonn, Bonn, GERMANY, University of 

Göttingen, Göttingen, GERMANY, 3Ludwigs-Maximilians University München, München, GERMANY 

PURPOSE 

Hyperintense signal changes in cortical and subcortical gray 

matter in T2-, FLAIR-, and diffusion-weighted MR images 

are known in sporadic Creutzfeldt-Jakob disease (sCJD). 

Recently the distribution of these changes was classified as 

striatal, cerebral cortical, and a combination of both using 

diffusion-weighted MR imaging. The aim of our study was 

to test whether this classification covers all signal abnormalities 

detectable in a large group of sCJD patients using various 

sequences. 


We included hardcopies of 125 T2-weighted, 88 FLAIR-, 

and 49 diffusion-weighted MR sequences of 133 patients 

from the German reference center for CJD (60 neuropathologically 

confirmed and 73 clinically probable sCJD cases, 

examined 2001-2003). Two observers analyzed the scans 

first individually and in a second step in consensus for high 

signal in the striatum, the thalamus, and the cortical ribbon. 

Symmetry of lesions was possible but not necessary. 

RESULTS 

High signal in the basal ganglia without cortical involvement 

was observed in 46 cases (35%), including 23 cases (17.5% 

of the whole group) with striatal lesions only (“striatal only”) 

and 23 (17.5%) with striatal and thalamic lesions (“striatothalamic”). 

In 51 cases (38%) the cortex showed high signal, 

of these 27 cases (20%) had hyperintense signal in the striatum, 

the cortex and the thalamus (“whole cerebral gray mat- 

132 

ter”), 21 (16%) involvement of the striatum and the cortex 

without thalamic signal changes (“striato-cortical”) and 3 

patients (2%) showed a “cortex only” form. One patient had 

a “thalamus only” involvement. A combination of the thalamus 

and the cortex without the striatum was not seen. No 

gray matter signal changes were observed in 35 cases (26%). 

Patterns of MRI Changes in sCJD 

Lesion Striatal Striato- Whole Striato- Cortex Thalamus Negative 

Type/Diagnosis Only thalamic Cerebralcortical only only MR 

Gray Imaging 

Matter 

Definite sCJD 14 11 13 6 2 0 14 

(n = 60) 

Probable sCJD 9 12 14 15 1 1 21 

(n = 73) 

Total (n = 133) 23 23 27 21 3 1 35 

CONCLUSION 

In addition to a totally negative MR imaging, six patterns of 

high signal lesions in sCJD can be defined: striatal only, striato-thalamic, 

whole cerebral gray matter, striato-cortical, 

cortex only, thalamus only. A correlation with clinical and 

pathologic data will further define these MR imaging subtypes. 

KEY WORDS: Creutzfeldt-Jakob disease, MR imaging, diffusion 


MR Imaging of Nonalcoholic Wernicke’s 

Encephalopathy: A Follow-Up Study 

Zhong, C. · Fei, G. · Jin, L. 

Zhongshan Hospital & Shanghai Medical College, Fudan 

University 

Shanghai, CHINA 

PURPOSE 

To investigate the correlation of MR features with pathologic 

evolution and prognosis of nonalcoholic Wernicke’s 

encephalopathy (WE). 


A retrospective review was conducted of six patients: three 

women and three men with ages ranging from 16 to 56 years, 

of nonalcoholic WE from 2000 to 2003. All patients, without 

a history of alcoholic abuse, underwent the course of iatrogenic 

fasting (from 16 to 34 days) and parenteral nutrition 

without thiamine supplement because of acute pancreatitis. 

When MR imaging was conducted, there were two of six 

patients with deep coma, two of six patients with mild coma, 

and two of six patients with drowsiness. The patients suspected 

as WE were administered with thiamine immediately 

and followed up for 1-2 years with the exception of one 

patient who died. 

RESULTS 

One of six patients died, one of six patients entered a persistent 

vegetative state, and the other four of six patients 

recovered fully from WE within 2 weeks to 1 year after 

administration of thiamine. Typical MR imaging exhibited 

areas of increased T2-weighted and FLAIR signals symmetrically 

surrounding the aqueduct and the third ventricle, at 

the floor of the fourth ventricle, in the medial thalamus and 

the caput of the caudate nucleus (Figure 1a). Two patients 

presenting without coma showed increased T2-weighted and

FLAIR signal of the periaqueductal area only. All four 

patients presenting with coma showed increased T2-weighted 

and FLAIR signals symmetrical in the medial thalamus 

and the caput of the caudate nucleus. Among four coma 

patients, two patients with deep coma showed increased T2weighted 

and FLAIR signals in the medial thalamus and caudate 

nucleus as well as in the frontal and parietal cortices 

(Figure 1b). According to the follow-up results, increased 

T2-weighted and FLAIR signals in four of six patients without 

cortical damage decreased in intensity, consistent with 

clinical recovery within 2 weeks to 1 year. The patient in a 

persistent vegetative state exhibited progressive atrophy of 

the whole brain during the 2 years of the follow-up study. 

CONCLUSION 

MR imaging is helpful not only to diagnose acute nonalcoholic 

WE but also to evaluate pathologic evolution and prognosis 

of the disorder. The involvement of the caudate nucleus 

and cortex is not an uncommon presentation but rather the 

sign of pathologic evolution. The disorder may be remediable 

if only periaqueductal area, thalamus, and caudate nucleus 

are involved. However, cortical damage may be indicative 

of irreversible damage and poor prognosis. 

KEY WORDS: Wernicke’s encephalopathy, nonalcoholic, thiamine 


3:00 PM - 4:30 PM 

Room 103 

(34) ELC Workshop C: Website 

Creation for the Novice 


4:40 PM - 5:40 PM 

Room 205 


(35) ELC Lecture E: Advanced 

Imaging Processing in MR Imaging 

— Todd B. Parrish, PhD 

133 


4:40 PM - 6:10 PM 

Room 105/106 

(36) Interventions for Head and Neck 

Neoplasms (ASITN) 

(255) Intraarterial Chemotherapy 

— Wade H.M. Wong, DO 

(256) Radiofrequency Ablation 

— Allan L. Brook, MD 

(257) Intervention for Carotid Artery "Blow-Out" 

— Gary M. Nesbit, MD 

Moderator: Gary M. Nesbit, MD 

Intraarterial Chemotherapy 

Wade H.M. Wong, DO 



able to: 

1) To describe the procedural methodology. 

2) To review the results. 

3) To discuss pitfalls and suggest practical procedural pearls. 


Intraarterial Chemotherapy for Squamous Cell Carcinoma of 

the Head and Neck. Head and neck cancers are historically 

difficult lesions to treat with long-term survival rates ranging 

between 15-40% when utilizing current conventional therapy 

of surgery, radiotherapy, and medical oncology. Often 

these cures are accompanied by diminished quality of life as 

a result of surgical impairment of speech and swallowing, 

not to mention cosmetic deformities. In the past 10 years, 

there has been the development of the intraarterial selective 

administration of high-dose cisplatin directly into the tumor 

vascular bed with concurrent circulating thiosulfate to protect 

critically sensitive organs (i.e., the kidneys and bone 

marrow) against the toxic effects cisplatin. In the most recent 

trial (Radplat) in which 4 cycles (1 cycle per week) of highdose 

cisplatin was delivered intraarterially (150 mg/m2) 

while concurrent daily radiation (100-200 gy/da) was delivered 

for 35 days, we found a complete response rate of 93%. 

This is two to four times better than conventional therapies. 

Tuesday

Tuesday 

REFERENCES 

1. Kerber CW, Wong WH, et al. An organ preserving selective 

arterial chemotherapy strategy for head and neck cancer. 

AJNR Am J Neuroradiol 1998;19:935-941 

2. Kerber CW, Wong WH. A Treatment for Head and Neck 

Cancer by Direct Arterial Infusion. In: Connors JJ, Wojak JC, 

Interventional Neuroradiology, Philadelphia:WB Saunders 

Co.1999;358-368 

3. Robbins KT, Storniolo AM, Kerber CW, et al. Phase I study of 

highly selective supradose cisplatin infusions for advanced 

head and neck cancer. J Oncol 1994;12:2113-2120 

Radiofrequency Ablation 

Allan L. Brook, MD 



able to: 

1) Describe the various techniques that have been developed 

for the application of radiofrequency ablation in the palliative 

treatment of recurrent and advanced head and neck cancer. 


To develop the methods and describe how CT guidance 

allows for the safe insertion of the radiofrequency needle and 

allows for avoidance of the critical neurovascular structures. 

Intervention for Carotid Artery "Blow-Out" 



Carotid hemorrhage associated with head and neck cancer, 

the so-called “carotid blowout syndrome”, is often a lifethreatening 

emergency caused by radiation necrosis, cutaneous 

fistula, abcess, or recurrent tumor. Patients will present 

with hemoptysis, hemorrhage from cutaneous fistula or 

tracheostomy, or hemorrhage into the GI tract or pulmonary 

tree. The site of hemorrhage may be small external carotid 

branches within a tumor bed, false aneurysms of larger 

branches of the external carotid, common, or internal carotid 

artery, or due to venous extravasation from the internal jugular 

vein. Cross-sectional imaging is critical in the therapeutic 

plan and should be obtained if the patient is clinically stable. 

False aneurysms, extravasation, proximity of abcess, 

tumor, or air from necrosis or cutaneous fistula to vascular 

structures are all features that can assist in determining the 

appropriate site when no active angiographic extravasation 

can be seen. Careful angiographic evaluation is also critical 

as the findings can often be quite subtle and incidental irregularities 

due to radiation, post-operative, neoplastic or atherosclerotic 

changes. These complexities often require correlation 

with cross sectional imaging studies and clinical evaluation 

from direct or endoscopic visualization to determine 

the actual site of hemorrhage. Surgical therapy is very limited 

due to the necrotic or sclerotic bed in the region in cancer 

patients and usually requires and extra-anatomic bypass, if 

no endovascular solution can be devised. Endovascular therapeutic 

options include particulate or cyanoacrylate 

134 

embolization of small tumor branches, vascular sacrifice of 

larger vessels using coils or balloons, or stent and stent-graft 

therapy. Although these treatments are usually definitive, 

they can also be adjunctive, and direct discussion of the therapeutic 

plan with the head and neck surgeon is necessary to 

optimize the therapy and ultimate outcome. Each of these 

endovascular techniques and the clinical situation where 

they are indicated will be discussed in a case presentation 

format. 


4:40 PM - 6:10 PM 

Room 107 

(37) General Neuroradiology Session 

(ARS)* 

Panelists 


— James G. Smirniotopoulos, MD 

— Robert D. Zimmerman, MD 

— David M. Yousem, MD, MBA


4:45 PM - 6:15 PM 

Theatre 


Perfusion 

(258) MR and CT Perfusion Techniques in Stroke 

— Pamela W. Schaefer, MD 

(259) MR Permeability Imaging 

— Timothy P.L. Roberts, PhD 

(260) Arterial Spin Labeling and CMRO2 

Measurements 

— Xavier Golay, PhD 


Moderators: Timothy P.L. Roberts, PhD 


MR and CT Perfusion Techniques in Stroke 

Pamela W. Schaefer, MD 



able to: 

1) Summarize the basic theory of methods for obtaining CT 

and MR perfusion images and creating blood volume, blood 

flow, and tissue transit time maps. 

2) Recognize the changes that occur on CT and MR perfusion 

images with acute stroke. 

3) Illustrate how CT and MR perfusion image findings may 

predict tissue outcome and affect acute stroke patient management. 

4) Recognize the pitfalls of CT and MR perfusion techniques 

and the differences between the two. 


Perfusion imaging has become essential in the imaging of 

acute stroke patients. This lecture will review the basic theory 

behind and practical technical aspects of performing IV 

first pass CT and MR dynamic imaging techniques on acute 

stroke patients. Pearls and pitfalls will be discussed and the 

processing of raw data into three major parameter maps 

(cerebral blood volume, cerebral blood flow, and tissue transit 

time) will be reviewed. MR and CT techniques will be 

compared and the advantages and disadvantages of both 

modalities will be discussed. Incorporation of perfusion 

techniques into acute stroke protocols will be discussed. This 

135 

lecture also will cover the physiologic changes that occur in 

perfusion during acute stroke and illustrate typical changes 

on parameter maps. Concepts such as operational penumbra, 

perfusion diffusion mismatch, early reperfusion, luxury perfusion, 

and misery perfusion will be discussed. This lecture 

also will illustrate how CT and MR perfusion image findings 

may suggest tissue outcome and how the findings on CT and 

MR perfusion images guide patient management. 

REFERENCES 

1. Sunshine JL. CT, MR imaging, and MR angiography in the 

evaluation of patient with acute stroke. J Vasc Interv Radiol 

2004;15(1Pt2):S47-S55 

2. Latchaw RE. Cerebral perfusion imaging in acute stroke. J Vasc 

Interv Radiol 2004;15(1Pt):S29-S46 

3. Grandin CB. Assessment of brain perfusion with MRI 

methodology and application to acute stroke. 

Neuroradiology 2004 Jan 20;(Epb ahead of print) 

4. Yamada K, Wu O, Gonzalez RG, Bakker D, Ostergaard L, 

Copen WA, Weisskoff RM, et al. Magnetic resonance perfusion-weighted 

imaging of acute cerebral infarction:effect of 

the calculation methods and underlying vasculopathy. 

Stroke 2002;33(1):87-94 

5. Hoeffner EG, Case I, Jain R, Gujar SK, Shah GV, Deveikis JP, 

Carolos RC, et al. Cerebral perfusion CT: technique and 

clinical applications. Radiology 2004;231(3)632-644. Epub 

2004 Apr 29 

MR Permeability Imaging 




able to: 

1) To appreciate the concept of microvascular permeability, 

as distinct from perfusion. 

2) To recognize the analysis of progressive enhancement on 

dynamic CE MR imaging and its modeling in terms of permeability. 

3) To appreciate the value of permeability in grading 

gliomas. 

4) To appreciate the value of permeability in the assessment 

of antiangiogenic therapies. 


If dynamic magnetic resonance imaging is performed in concert 

with bolus injection of gadolinium-based contrast 

media, it is possible to “track” the bolus as it passes through 

the tissues of the brain. In the context of tumor imaging, it is 

not only the first pass (perfusion-related) transit of the contrast 

agent which is of interest, but also progressive enhancement 

revealed over the course of the next 3-5 minutes. 

Conceptually, initial enhancement will reflect the vascular 

fractional volume of the tissue (a tissue with 5% vascular 

volume will initially enhance 5% as much as a reference 

blood vessel containing 100% blood); any progressive 

enhancement, however, will relate to the extravasation of 

contrast medium from the intravascular to the interstitial 

space. Given the lack of such extravasation in the healthy 

brain (where contrast media cannot cross the intact bloodbrain 

barrier), such hyperpermeability can be viewed as an 

indicator of pathologic compromise of blood-brain barrier 

Tuesday

Tuesday 

integrity. Kinetic modeling of the dynamic signal intensity 

vs time curves from tissue and a reference blood vessel (e.g., 

the sagittal sinus) allows interpretation of the dynamic data 

in terms of fBV and microvascular permeability, directly (1- 

3). Performing this analysis on a pixel-by-pixel basis, it is 

possible to generate parameter maps or pseudo-images, 

where image intensity relates to the physiological parameter 

of interest (fBV or Kps). Studies have demonstrated superior 

distinction between tumors of various grade by the adoption 

of such measures of microvascular permeability, over 

consideration of relative blood volume alone (2). However, 

assessment of microvascular permeability is not only useful 

for grading tumors, but also should be taken in context of 

emerging evidence that microvascular permeability elevation 

may be an indicator of the process of angiogenesis and 

thus may be a sensitive marker of the process of antiangiogenesis 

(4) via new generations of pharmaceutical, such as 

Avastin (Genentech Inc., South San Francisco, CA). Such 

agents which reduce tumor growth rate rather than killing 

tumor cells directly demand an effective indicator of biological 

efficacy such as quantitative assessment of microvascular 

permeability reduction as traditional endpoints may no 

longer be appropriate. Finally, recent application of dynamic 

contrast-enhanced imaging and kinetic modeling of 

microvascular permeability shows promise for the detection 

of blood-brain barrier disruption in acute stroke, a potential 

indicator of subsequent hemorrhagic transformation. 

REFERENCES 

1. Shames DM, Kuwatsuru R, Vexler V, Muhler A, Brasch RC. 

Measurement of capillary permeability to macromolecules 

by dynamic magnetic resonance imaging: a quantitative 

noninvasive technique. Magn Reson Med 1993;29:616-622 

2. Roberts HC, Roberts TPL, Brasch RC, Dillon WP, 

Quantitative estimation of microvascular permeability in 

human brain tumors: correlation with histopathological 

grade. AJNR Am J Neuroradiol 2000;21(5): 891-899 

3. Roberts HC, Roberts TPL, Bollen AW, Ley S, Brasch RC, 

Dillon WP. Correlation of microvascular permeability 

derived from dynamic contrast-enhanced MR imaging with 

histologic grade and tumor labeling index: a study in 

human brain tumors. Acad Radiol 2001;8(5):384-391 

4. Gossmann A, Helbich TH, Kuriyama N, Ostrowitzki S, Roberts 

TP, Shames DM, van Bruggen N, et al. Dynamic contrastenhanced 

magnetic resonance imaging as a surrogate marker 

of tumor response to anti-angiogenic therapy in a 

xenograft model of glioblastoma multiforme. J Magn Reson 

Imag 2002;15(3):233-240 

Arterial Spin Labeling and CMRO2 Measurements 

Xavier Golay, PhD 

Dr. Xavier Golay is currently a member of the Research 

Faculty at the Department of Neuroradiology at the National 

Neuroscience Institute of Singapore. He also holds an 

adjunct Associate Professor in the Department of Electrical 

and Electronic Engineering of Nanyang Technological 

University, as well as an affiliate faculty position in the 

Department of Radiology of the Johns Hopkins University. 

He received his PhD from the Swiss Federal Institute of 

Technology of Zurich (ETHZ) in 1998, and moved to the 

Johns Hopkins University as a postdoctoral fellow. In 2002, 

136 

he was promoted to Assistant Professor, and left 1 year later 

to go to Singapore. Dr. Golay is the author or coauthor of 

more than 40 articles and four book chapters of which five 

recent articles and three review articles deal specifically 

with measurement of cerebral perfusion using arterial spin 

labeling. 



able to: 

1) Explain how Arterial Spin Labeling (ASL) works. 

2) Illustrate how advanced ASL techniques can provide new 

information in a variety of neurologic diseases. 

3) Evaluate how to combine ASL with BOLD measurements 

to infer on CMRO2 changes. 


Arterial spin labeling is a magnetic resonance method for the 

measurement of cerebral blood flow. In its simplest form, the 

perfusion contrast in the images gathered by this technique 

comes from the subtraction of two successively acquired 

images, one with, and one without proximal labeling of arterial 

water spins. Over the last decade, the method has moved 

from the experimental laboratory to the clinical environment. 

Furthermore, numerous improvements, ranging from 

new implementations to extensive theoretical studies have 

broadened its reach and extended its potential applications. 

In this lecture, the multiple facets of this powerful yet difficult 

technique are discussed. Different implementations will 

be compared, the theoretical background will be summarized, 

and potential applications of the latest developed techniques 

will be demonstrated. Finally, the combination of this 

technique with gradient-echo BOLD imaging will allow 

demonstrating one of the most promising applications of this 

technique in the field of functional neuroimaging research, 

the ability to noninvasively assess changes in the metabolic 

rate of oxygen CMRO2. 

REFERENCES 

1. Golay X, Hendrikse J, Lim TCC. Perfusion imaging using 

arterial spin labeling. Top Magn Reson Imag 2004;15(1):10- 

27 

2. Barbier EL, Lamalle L, Decorps M. Methodology of brain 

perfusion imaging J Magn Reson Imag 2001;13(4):496-520 

3. Hendrikse J, van der Grond J, Lu H, van Zijl PC, Golay X. 

Flow territory mapping of the cerebral arteries with regional 

perfusion MRI (RPI). Stroke 2004;35(4):882-887 

4. Golay X, Petersen ET, Hui F. PULsed STAR labeling of arterial 

regions (PULSAR): a robust regional perfusion technique 

for high field imaging. Magn Reson Med 

2005;53(1):15-21 

5. Lu H, Golay X, Pekar J, van Zijl PC. Sustained poststimulus 

elevation in cerebral oxygen utilization after vascular recovery. 

J Cereb Blood Flow Metab 2004;24(7):764-770 

Panel Discussion


5:00 PM - 6:00 PM 

Room 103 


(NLM): When PubMed ® is Not The 

Answer 

— Maryanne Blake 

137 

Tuesday

Tuesday 

Notes:

Wednesday Morning 

8:00 AM - 9:30 AM 

Room 105/106 

(40) Special Session: Surviving 

Change: Neuroradiology Practice 

(262) Overview of NER Foundation and Resource 

Utilization 

— A. James Barkovich, MD 

(263) Report from the Clinical Practice Committee: 

Aligning ASNR CPC Activities with Other 

Societies 

— Patrick A. Turski, MD 

(264) The Neurovascular Coalition 

— Gary R. Duckwiler, MD 

(265) Update on Maintenance of Certification (MOC) 

— Glenn S. Forbes, MD, FACR 

139 






8:00 AM - 9:30 AM 

Room 205 

(41) Perspectives on Submission of 

Grant Applications 

(268) Grantsmanship: The Art of Writing a Grant 

— Colin P. Derdeyn, MD 

(269) The National Institutes of Health Review 

Process 

— R. Nick Bryan, MD, PhD 

(270) Research Funding Opportunities for 

Neuroradiologists 

— James M. Provenzale, MD 

Moderator: James M. Provenzale, MD 

Grantsmanship: The Art of Writing a Grant 




able to: 

1) Explain and review the format and structure of grant 

applications. 

2) Define key points of emphasis in writing a grant. 

3) Discuss common critical mistakes made in preparing 

grant applications. 


Obtaining funding for research is a competitive process. The 

basic foundation for a competitive grant application was 

reviewed by previous speakers: the development of a 

testable hypothesis, the design of an experiment to adequately 

test the hypothesis, and the results of preliminary 

data indicating feasibility. No grant application lacking these 

fundamental items will be successful. Once you have assembled 

these items, your next two steps are to select a funding 

source and write a grant application. It is an unfortunate fact 

that many grant applications with great ideas, good design, 

Wednesday

Wednesday 

and strong preliminary data fail simply because their authors 

did not communicate effectively this information to the 

reviewers. The ability to write a good grant application is a 

specialized skill that investigators learn and develop over 

time. This skill in packaging has been called grantsmanship. 

This talk will focus on the key aspects of grant composition. 

First, we will discuss targeting the application for specific 

funding mechanisms - NIH RO1, R21, for example. Then we 

will use the current National Institutes of Health PHS 398 

application form as a template and go through it page by 

page, emphasizing key points and common pitfalls. Case 

examples from successful and unsuccessful grant applications 

will be used to illustrate these points and pitfalls. 

The National Institutes of Health Review Process 

R. Nick Bryan, MD, PhD 



able to: 

1) Appreciate the NIH structure. 

2) Understand the Center for Scientific Review process. 

3) Experience a R01 Grant “Case Presentation.” 


The National Institutes of Health (NIH) utilize a highly 

structured peer review process that is used for evaluating, 

funding, and monitoring its extramural grants. It is important 

to understand the system and follow its procedures closely as 

deviations from protocol can result in rejection of an application. 

While the system is stringent, it can be very helpful 

and it is fair. All NIH policies and practices are available on 

the WEB and the peer review process is well documented at: 

http://www.csr.nih.gov/review/peerrev.htm. A typical grant 

(R01/R21) is received initially by the Center for Scientific 

Review (CSR) and, using written guidelines, is assigned by 

a Referral Officer to an Integrated Review Group (IRG). 

IRGs are clusters of Study Sections that review similar science. 

The grant then is assigned to a specific Study Section 

and specific Institutes are identified that might be suitable 

for funding such a grant. Referral Officers do take into 

account written requests for specific IRG and Study Section 

assignments. A Study Section has approximately 20 members 

who are active researchers in the particular area of 

research covered by the Section. The Principle Investigator 

is notified of the assignment and further communication is 

between the PI and the Scientific Review Administrator 

(SRA) of that Study Section. The SRA reviews the grant for 

content, completion, etc. and assigns it to 2 or 3 members of 

the Study Section for written review and 1 or 2 additional 

members for discussion. Submission of Supplementary 

Material to the original grant may be arranged between the 

SRA and PI. Each assigned reviewer provides the SRA a 

detailed written evaluation of the grant and a score based on 

a 100 to 500 point system (100 being best possible). 

Approximately a week before the Study Section meets, the 

SRA compiles all scores and those grants that score in the 

lower half are “streamlined.” These grants are not scored or 

discussed at the meeting and will not be funded; however the 

PI is given the written critiques and subsequently may revise 

and resubmit the grant. A Study Section meeting lasts 2 days, 

during which the members sit around a large table and 

140 

review in detail all those grants not “streamlined.” After 

active discussion of a grant, every member of the Study 

Section privately marks his/her Priority Score on a form provided 

by the SRA. After the meeting, a computer-generated 

average Priority Score and percentile is calculated and this, 

along with the written critique, are sent to the PI. These 

scores also are sent to the individual Institutes that select, on 

the basis of these scores and institutional priorities and 

resources, those grants to fund. At this point the applicant’s 

link to the NIH changes from the SRA and CSR to an official 

of the particular Institute to which the grant has been 

assigned. For institutional funding, grants are usually well 

within the top quartile. Unfunded grants can be resubmitted. 

In fact this might be considered the norm for first time submissions, 

as many grants are rejected initially for funding. 

Hence it is very important to not be discouraged by an initial 

rejection; however, resubmissions should carefully address 

concerns raised in the written critiques from the original 

Study Section. 

Research Funding Opportunities for Neuroradiologists 




able to: 

1) Describe federal funding mechanisms for investigator-initiated 

grants. 

2) Show the variety of funding mechanisms available 

through the NIH. 

3) Detail eligibility requirements for a number of NIH funding 

mechanisms. 


Advances in diagnostic and therapeutic neuroradiology have 

led to greater stature of neuroradiologists in the medical 

community. Clinicians and researchers in other disciplines 

rely heavily on the expertise and imaging tools available to 

neuroradiologists. However, with increasing clinical 

demands, less time is available for neuroradiologists themselves 

to perform research studies. This lecture is a first step 

in an attempt to reverse this trend. Extramural funding is one 

means of insuring protected research time. Levels of available 

funding are reported to be at their highest level ever. 

The purpose of this lecture is to familiarize neuroradiologists 

with (1) a few of the tools needed to write successful 

research protocols and (2) some funding mechanisms that 

could provide funded research time needed for writing grant 

applications. A wide variety of funding mechanisms are 

available to radiologists. These mechanisms include federally 

funded resources, private foundations (e.g., NER 

Foundation and RSNA Research and Education Fund) and 

for-profit agencies such as industry. This lecture will focus 

on federally funded mechanisms that might be of interest to 

young investigators. In particular, The NIH K-series awards 

will be discussed in detail. This series of awards provide 

funding for 3-5 years for young researchers with potential to 

become independent investigators. They require mentorship 

by a more senior researcher in the young investigator's field 

or other scientific fields and generally require between 50% 

and 75% effort by the young investigator. In addition, the 

R21 funding mechanism, which is for a maximum of 2 years

and provide direct costs for the 2-year period not to exceed 

$275,000, will be discussed. This funding mechanism is one 

that allows for investigation of exploratory hypotheses with 

little preliminary data. As such, it may provide young investigators 

with a means of obtaining funding that will allow 

them to obtain preliminary data needed for submission of 

grant applications, such as the R01 grant, that provide funding 

for a longer period. 


10:00 AM - 10:15 AM 

Room 105/106 

(42) ASNR Presidential Address 

(271) ASNR Presidential Address 

— Victor M. Haughton, MD, ASNR President 

141 


10:15 AM - 11:10 AM 

Theatre 

(43) ASNR Awards Presentation 

Presentation of 2005 Gold Medal Award 

Moderators: Victor M. Haughton, MD, ASNR President 

Michael Deck, MD, FACR, Chair, 

Gold Medal Award Committee 

Presentation of 2005 ASNR Honorary Member 

Moderator: Norman E. Leeds, MD, Chair, 

Honorary Member Committee 

Announcement of 2004 Outstanding Presentation 

Awards 

Moderator: James G. Smirniotopoulos, MD, Chair, 

Education Committee 

Presentation of the Neuroradiology Education and 

Research (NER) Foundation Award for 

Outstanding Contributions in Research 

Moderator: A. James Barkovich, MD, Chair, 

Neuroradiology Education and 


Announcement of 2005 NER Foundation Scholar 

Award in Neuroradiology Research 

Announcement of 2005 NER Foundation Outcomes 

Research Grant in Neuroradiologic Imaging 

Announcement of 2005 NER Foundation/Boston 

Scientific-Target Fellowship in 

Cerebrovascular Disease Research 

Announcement of 2005 Berlex/NER Foundation 

Fellowship in Basic Science Research Awards 

Moderator: Howard A. Rowley, MD, Chair, 

Research Committee 

Announcement of Symposium Neuroradiologicum 

(SNR) XVII/World Federation of 

Neuroradiological Societies (WFNRS) 

Michael S. Huckman, MD, 

WFNRS President 

Wednesday

Wednesday 

WednesdayMorning 

11:10 AM - 11:40 AM 

Room 105/106 

(44) American Society of 

Neuroradiology (ASNR) Annual 

Business Meeting (Members Only) 

WednesdayMorning 

11:50 AM - 12:50 PM 

Room 201F 

(45) American Society of Functional 

Neuradiology (ASFNR) Annual 


Wednesday Afternoon 

1:00 PM - 2:30 PM 

Theatre 

(46) Neurodegenerative Disorders 

(ASFNR) 

(272) Neuroradiology of Neurodegenerative 

Disorders 

— Sasan Karimi, MD 

(273) Functional Imaging of Neurodegenerative 

Disorders 

— Jeffrey R. Petrella, MD 

(274) From Molecules to Neuroimaging 

— R. Gilberto Gonzalez, MD, PhD 

Moderator: R. Gilberto Gonzalez, MD, PhD 

142 

Neuroradiology of Neurodegenerative Disorders 

Sasan Karimi, MD 



able to: 

1) Identify and discuss the more common and radiologically 

relevant neurodegenerative disorders in the adult patient. 

2) Review the important and common neuroradiologic imaging 

findings of the selected neurodegenerative disorders. 


The development of magnetic resonance (MR) imaging has 

greatly enhanced the radiologic assessment of neurodegenerative 

disorders, including the dementias, movement disorders, 

hydrocephalus, and temporal lobe epilepsy. 

Distinguishing among various neurodegenerative disorders 

is often a difficult clinical task. The normal aging process of 

the brain further adds complexity to making diagnosis of 

neurodegenerative disorders. At times it is difficult to differentiate 

a pathologic process from the normal aging brain. 

MR imaging with its multiplanar ability to image structure 

and more recently function of the brain, may be sensitive as 

well as specific in diagnostic evaluation of neurodegenerative 

disorders. The importance in understanding the MR 

imaging appearance of neurodegenerative disorders will 

increase as the population ages and as new therapies for 

these disorders are developed. The demographics, clinical 

features and typical MR imaging findings of the more common 

and radiologically relevant neurodegenerative disorders 

in the adult patient will be reviewed and discussed. This 

includes diseases or conditions such as: Alzheimer's, Pick's, 

and Creutzfeldt-Jakob disease, multiple infarct dementia, 

cerebral autosomal dominant arteriopathy with infarcts and 

leukoencephalopathy (CADASIL), Parkinson's and 

Huntington's disease as well as normal pressure hydrocephalus. 

REFERENCES 

1. Holodny AI, George AE, Golomb J, de Leon MJ, Kalnin AJ. The 

perihippocampal fissures: normal anatomy and disease 

states. Radiographics 1998;18:653-665 

2. Drayer, BP. Imaging of the aging brain II. Pathological conditions. 

Radiology 1988;166:797-806 

3. Pantoni L, Garcia JH. The significance of cerebral white matter 

abnormalities 110 years after Binswanger’s report. A 

review. Stroke 1996;26:1293-1301 

4. Barboriak DP, Provenzale JM, Boyko OB. MR diagnosis of 

Creutzfeldt-Jakob disease: significance of high signal intensity 

in the basal ganglia. AJR Am J Roentgenol 

1994;162(1):137-140 

5. Bradley WG, Jr. Diagnostic tools in hydrocephalus. Neurosurg 

Clin N Am 2001; 2(4):661-84, viii

Functional Imaging of Neurodegenerative Disorders 

Jeffrey R. Petrella, MD 



able to: 

1) Demonstrate the application of fMRI to assessment of 

dementia and movement disorders. 

2) Identify necessary requirements for use of fMRI as a diagnostic 

and therapeutic assessment tool in neurodegenerative 

disorders. 


Application of functional MR imaging (fMRI) to memory 

and movement disorders with an emphasis on Alzheimer’s 

disease will be discussed. Reference will be made to previous 

FDG PET studies. The pathologic basis and evolution of 

Alzheimer’s disease will be reviewed along with its clinical 

manifestations as well as therapies on the horizon. It has 

become clear that there is pathologic evidence of the disease 

decades before clinical symptoms become evident. The early 

presence of neuronal dysfunction, which precedes neuronal 

death and tissue atrophy, may offer the opportunity for early 

diagnosis with functional imaging modalities. Functional 

MR imaging studies in patients with Alzheimer’s disease, 

mild cognitive impairment, and other memory disorders, 

such as frontotemporal dementia and dementia with lewy 

bodies, will be reviewed. Recent fMRI studies in patients 

with movement disorders, such as Parkinson’s and 

Huntington’s disease, will be discussed briefly. Finally, 

requirements for fMRI as tool for early diagnosis of neurodegenerative 

disease and drug assessment will be 

addressed. 

From Molecules to Neuroimaging 




able to: 

1) Describe the major molecular changes that occur in the 

brain in neuroinflammatory and neurodegenerative diseases. 

2) Describe the probable biological basis of these molecular 

changes. 


A large number of neurological diseases demonstrate 

changes in the amount of molecules that occur in high concentration 

in the brain that are detectable by proton MR 

spectroscopy. Diseases include inflammatory processes such 

as multiple sclerosis, neurodegenerative diseases such as 

Alzheimer’s disease, posttraumatic syndromes and even psychiatric 

disorders such as schizophrenia. The spectrum from 

these disorders follows a characteristic pattern. They typically 

include a decrease in NAA and relative increases in 

Cho and MI. There is substantial evidence to indicate that the 

decrease in NAA is a result of injury to neurons. The elevations 

in Cho and MI generally are ascribed to gliosis. 

However a close reading of the literature reveals that there is 

no strong experimental evidence to support this supposition. 

To further understand the biological basis of the altered pro- 

143 

ton MR spectrum, in these diseases, investigations have been 

conducted employing post-mortem human brains (1, 2) and 

a nonhuman primate model (3, 4). These studies indicate that 

frank neuronal loss is highly correlated to a decline in NAA. 

Investigations of reversible NAA decline indicate that this 

molecule is a marker of early neuronal injury, and best correlates 

with the synaptic marker synaptophysin. The Cho and 

MI resonances are far more complex. Detailed studies comparing 

these resonances with the astrogliosis marker GFAP 

shows that during an acute neuroinflammatory process there 

is concordance in changes of these three markers early in the 

process; however they diverge subsequently. Moreover, 

studies of brain extracts reveal that the changes in Cho and 

probably the MI resonances that are observed in vivo are a 

result of contributions from small molecular weight, water 

soluble molecules and larger macromolecules. 

REFERENCES 

1. Cheng LL, Ma MJ, Becerra L, Ptak T, Lackner A, González RG. 

Quantitative neuropathology by high resolution magic angle 

spinning proton magnetic resonance spectroscopy. Proc Natl 

Acad Sci USA 1997;94:6408-6413 

2. Cheng LL, Newell K, Mallory A, Hyman BT, Gonzalez RG. 

Quantification of neurons in Alzheimer and control brains 

with ex vivo high resolution magic angle spinning proton 

magnetic resonance spectroscopy and stereology. Magn 

Reson Imag 2002;20:527-533 

3. Lentz MR, Kim JP, Westmoreland SV, Greco JB, Fuller RA, 

Ratai EM, He J, et al.Quantitative neuropathological correlates 

of changes in macaque brain NAA/Cr. Radiology 2005; 

In press 

4. Kim JP, Lentz MR, Westmoreland SV, Greco JB, Ratai EM, He 

J, Halpern EF, et al. Relationships between astrogliosis and 1H 

MRS measures of brain Cho/Cr and MI/Cr in a primate 

model. AJNR Am J Neuroradiol 2005; In press 


1:00 PM - 2:30 PM 

Room 105/106 

(47) Evaluation and Management of 

Intractable Seizures in Children 

(ASPNR) 

— O. Carter Snead, III, MD 

— James T. Rutka, MD, PhD 

Moderators: Marvin D. Nelson, MD 

David G. Ashley, MD 

Wednesday

Wednesday 

Evaluation and Management of Tractable Seizures in 

Children 

O. Carter Snead, III, MD 



able to: 

1) Understand seizures and epilepsy. 

2) Know the criteria for epilepsy surgery in children. 

3) Understand the preoperative diagnositic evaluation for 

epilepsy surgery. 

4) Appreciate the role of magnetoencephalography in epilepsy 

surgery. 


Twenty-five percent of children with epilepsy continue to 

seize despite best medical management and may be defined 

as medically refractory. Many children with medically 

refractory localization-related epilepsy (i.e., seizures which 

originate in a particular area of brain and secondarily spread 

to involve other brain regions) may benefit from a variety of 

surgical treatments including hemispherectomy, corpus callosostomy, 

focal cortical resection of the temporal lobe, focal 

cortical resection of extratemporal regions of brain, and multiple 

subpial resections. A successful outcome from epilepsy 

surgery generally is defined as a seizure-free state with no 

imposition of neurologic deficit. In order to achieve these 

twin goals two criteria must be fulfilled. First, precise localization 

of the epileptogenic zone in the brain is necessary. 

The epileptogenic zone may be defined as the region of 

epileptogenic cerebral cortex whose removal will result in a 

seizure-free state. Second, one must determine the anatomical 

localization of eloquent cortex in brain in order to spare 

these areas during any planned cortical excision of epileptogenic 

cortex. Several diagnostic measures may be used to 

achieve a successful surgical outcome. A clinical history to 

ascertain the earliest symptom in the clinical progression of 

the seizure (semiology) is imperative as is ictal and interictal 

scalp EEG, neuropsychological testing, magnetic resonance 

(MR) imaging, positron emission tomography (PET), singlephoton 

emission computed tomography (SPECT), and interictal 

magnetoencephalography (MEG). This technique, married 

to surgical navigation techniques, has proven particularly 

useful at our institution. Magnetoencephalography is a 

technique which detects the magnetic fields associated with 

the intracellular activity of thousands of neurons, unlike 

EEG which measures extracellular volume currents. 

Superconducting quantum interference devices (SQUIDS) 

are used to amplify these very small magnetic field signals 

which are output as real-time wave forms. Magnetic source 

imaging (MSI) is defined as the combination of high temporal 

resolution electrophysiologic data which is derived from 

MEG recordings and coregistered to high spatial resolution 

structural MR images. The utility of MSI lies in the combination 

of the submillisecond temporal resolution of MEG 

with the precise anatomical images provided by MR imaging. 

Magnetic source imaging allows for rapid, efficient 

sampling and superior neuronal source localization. 

Therefore, MSI is a useful tool for noninvasive localization 

of the epileptogenic zone in children who are candidates for 

epilepsy surgery. In the typical child undergoing evaluation 

for epilepsy surgery, if the clinical, neuropsychological, 

EEG, and radiologic data are all concordant and point to the 

144 

same area of epileptogenicity in brain, cortical excision of 

the suspected epileptogenic zone is undertaken. However, if 

the data are discordant, and/or the epileptogenic zone resides 

wholly or in part within eloquent cortex, invasive intracranial 

monitoring from depth and/or subdural electrodes during 

a seizure is required to map out the areas of epileptogenicity 

in brain. The assessment of potential risks and benefits 

for this type of epilepsy surgery in children involve 

complex age-related issues, including the possible impact of 

uncontrolled seizures, medication, or surgery, on learning 

and development. 

REFERENCES 

1. Minassian BA, Otsubo H, Weiss S, Elliott I, Rutka JT, Snead 

OC. Magnetoencephalography localization in pediatric 

epilepsy surgery: comparison with invasive intracranial electroencephalography. 

Ann Neurol 1999;46:627-633 

2. Snead OC. Surgical treatment of medically refractory epilepsy 

in childhood. Brain and Devel 2001;23:199-207 

3. Otsubo H, Snead OC. Magnetoencephalography and magnetic 

source imaging in children. J Child Neurol 2001;16:227-235 

4. Holowka SA, Otsubo H, Koji I, Pang E, Sharma R, Hunjan A, 

Xiang J, et al. Three dimensionally reconstructed magnetic 

source imaging and neuronavigation in pediatric epilepsy. 

Neurosurgery 2004;55:E1244-E1248 

Evaluation and Management of Intractable Seizures in 

Children 

James T. Rutka, MD, PhD 

Born in Toronto, and educated at Princeton University 

(1975-1977), and Queen's University Medical School (1977- 

1981), Dr. Rutka did an internship at McGill University 

(1981-1982) before entering the University of Toronto 

Neurosurgery Training Program in 1982. His training 

included a research fellowship at the Brain Tumor Research 

Centre, the University of California San Francisco where he 

obtained his PhD in Experimental Pathology (1984-1987). 

Upon conclusion of his neurosurgical residency in 1989, Dr. 

Rutka did a postdoctoral research fellowship in molecular 

immunology at Juntendo University, Tokyo (1990). Dr. Rutka 

assumed his appointment in the Department of Surgery, 

Division of Neurosurgery in 1990, and has been on the surgical 

staff at the Hospital for Sick Children in the Division of 

Pediatric Neurosurgery since that time. Dr. Rutka's primary 

research and clinical interests relate to the science and surgery 

of human brain tumors. His laboratory interests lie in 

the molecular biology of human brain tumors - specifically 

in the determination of the mechanisms by which brain 

tumors grow and invade. He has over 200 peer review publications. 

He is on the editorial boards of Neurosurgery and 

the Journal of Neurosurgery. He is Professor of 

Neurosurgery at the University of Toronto. He was appointed 

as the first Director of the Arthur and Sonia Labatt Brain 

Tumor Research Centre at the University of Toronto in 1998. 

In 1999, he received a Scientist Award from the Medical 

Research Council of Canada. This same year, Dr. Rutka was 

appointed Chairman of the Division of Neurosurgery at the 

University of Toronto, and sits in the Dan Family Chair of 

Neurosurgery. In 2001 Dr. Rutka received the Lister Award 

in the Department of Neurosurgery for sustained contributions 

to surgical research at the University of Toronto. In

2002, Dr. Rutka was listed by Globe and Mail as one of the 

Nation Builders in Canada, in the section of Science and 

Technology. This same year, he was appointed to the Board 

of Directors of the AANS. In 2004, Dr Rutka received the 

Grass Award from the Society of Neurological Surgeons for 

excellence in neurosurgery research, and delivered the 

Bittner Lecture at the AANS annual meeting. 



able to: 

1) Learn the neurosurgical options available to aid children 

with intractable epilepsy. 

2) Understand the benefits and potential risks of neurosurgery 

for pediatric epilepsy. 

3) Be aware of the variety of imaging modalities which can 

aid the neurosurgeon in performing epilepsy surgery. 


Neurosurgery remains a valuable treatment option for children 

with intractable epilepsy. The gamut of procedures 

ranges from lesionectomy for tumors and cortical dysplasia, 

to focal cortical resections in nonlesional cases, to hemispheric 

procedures such as hemispherectomy. Neurosurgery 

for epilepsy is highly dependent on detailed neuroimaging 

procedures including advanced MR sequences, MEG, PET, 

and SPECT. Brain mapping before or at the time of surgery 

enables the neurosurgeon to avoid major neurologic deficits. 

In this regard, the utility of neuronavigation cannot be 

emphasized strongly enough. In lesional epilepsy, the vast 

majority of children (> 90%) are improved following 

lesionectomy. In nonlesional epilepsy, we are achieving 70- 

80% significant improvement rates for children with 

intractable epilepsy. Patients undergoing hemispherectomy 

for neurosurgical disease (Sturge Weber, Rasmussen's 

encephalitis) can be made seizure-free after surgery provided 

early surgery is offered. 

145 

- 


1:00 PM - 2:30 PM 

Room 107 

(48) Spine Case-Based Review 

(ASSR) (ARS)* 

Panelists: 


1:00 PM - 1:45 PM 

Room 205 

— Gordon K. Sze, MD 

— David M. Yousem, MD, MBA 

— Erin M. Simon, MD 

— Mauricio Castillo, MD 

Moderators: Gordon K. Sze, MD 



(49) Socioeconomic Scientific Paper 

Session 


Moderators: Irwin A. Keller, MD 



Evidence-Based Screening Criteria for Blunt 

Cerebrovascular Injury 

Bub, L. D. · Hollingworth, W. · Blackmore, C. C. · Jarvik, J. G. 


Seattle, WA 

PURPOSE 

To evaluate clinical and radiographic predictors of blunt 

cerebrovascular injury (BCVI) and develop a clinical prediction 

rule which could facilitate rapid diagnosis and treatment. 

Wednesday

Wednesday 


A retrospective case-control study was performed on 104 

consecutive blunt trauma patients who sustained carotid or 

vertebral artery injuries (cases) and 220 randomly selected 

blunt trauma patients without radiographic or clinical evidence 

for BCVI (controls). Data on 24 possible predictors of 

BCVI were collected for each patient from emergency room 

records, clinical notes, and radiology reports. Each potential 

predictor was evaluated through simple and multivariate 

logistic regression. Clinically similar predictors were 

grouped into composite variables, and using recursive partitioning, 

a clinical prediction rule was generated. Receiver 

operator characteristic (ROC) curves were calculated. The 

absolute probability of BCVI will be estimated for each risk 

group using the odds ratio form of Bayes’ theorem. 

RESULTS 

Composite predictors of carotid or vertebral artery injury 

included transverse foramen fractures (OR 27.7; confidence 

interval (CI) 9.6, 80.2); cerebral or cerebellar infarct (OR not 

calculable, no control patients with infarct); skull base fracture 

through the carotid canal (OR not calculable, no control 

patients with carotid canal fracture); focal neurologic deficit 

(OR 8.3; CI 2.6, 26.5); cervical vertebral body subluxation 

or atlantooccipital dissociation (OR 15.2; CI 1.2, 184.5); and 

anterior neck injury, including thyroid, hyoid, airway injury, 

or mandible fracture (OR 6.3; CI 2.5, 15.7). Intracranial 

hemorrhage (OR 3.2; CI 1.3, 7.3) and altered mental status 

or unconsciousness (OR 1.4; CI 0.6, 3.2) were moderately 

and mildly predictive respectively. The prediction rule stratified 

patients into risk groups with diminishing absolute 

probability of BCVI. The absolute probability of injury for 

each of the above risk groups will be determined by the time 

of presentation. The area under the ROC curve was 0.91. 

After accounting for the above factors, other clinical variables 

such as vertebral body fractures, facial fractures, seat 

belt sign, aortic injury, first or second rib fractures, skull base 

fractures not involving the carotid canal, and posterior element 

fractures not involving the transverse foramen were not 

found to be significant predictors of BCVI. 

Composite Predictors of BCVI Based on Multivariate Logistic 

Regression 

Composite Predictors of BCV Odds Ratio Confidence p value 

Interval 

Transverse Foramen fracture 27.7 9.6, 80.2

CONCLUSION 

The apparatus shows excellent sensitivity and specificity for 

detecting even small ferromagnetic articles on patients prior 

to MR imaging. The design of the apparatus, resembling an 

airport metal detector allowed easy acceptance by patients. 

Proper screening of patients and personnel using a ferromagnetic 

detector may help decrease MR imaging-related 

adverse effects. 

KEY WORDS: Safety, screening, ferromagnetic 


Racial and Economic Disparities in Access to Acute 

Stroke Treatments 

Bateman, B. T. · Schumacher, H. C. · Berman, M. F. · Pile- 

Spellman, J. 


New York, NY 

PURPOSE 

Thrombolytic therapy, delivered either intravenously or 

intraarterially, is an effective treatment for acute stroke. The 

purpose of this study was to assess whether there are economic 

or racial disparities in the delivery of this therapy. 


This study used data from the Nationwide Inpatient Sample 

(NIS), the largest database of inpatient admissions available 

in the United States, for the years 1999-2002. The database 

contains the discharge records from approximately 8 million 

hospitalizations annually. It represents a stratified sample of 

approximately 20% of all inpatient hospitalizations in the 

United States. Patients with an acute stroke admitted from 

the emergency room were identified (n = 282,519). Those 

patients for whom race was coded were selected for further 

147 

analysis (n = 214,419). Race was simplified to either 

Caucasian or non-Caucasian for purposes of analysis. 

Median income quartile for the patients’ zip code, which was 

taken as a measure of patients’ economic status, was recorded. 

Those patients who received thrombolytic therapy then 

were identified (n = 2603). Patients who were coded as both 

receiving thrombolytics and having angiogram were designated 

as having IA thrombolysis (n = 371). Logistic regression 

analysis was performed to assess the effect of race and 

the median income quartile of the patient’s zip code on the 

patients’ likelihood of receiving thrombolytics, adjusting for 

age, sex, and co-morbid medical conditions. 

RESULTS 

Unadjusted analysis shows a higher percentage of 

Caucasians than non-Caucasians received thrombolytic therapy 

(1.4% vs . 1.0%, p < .025). There was also a higher rate 

of receiving IA thrombolytics in the Caucasian group 

(0.18%) than the non-Caucasian group (0.16%), although 

this difference was not statistically significant. Logistic 

regression shows a significant predictive effect of both race 

and median income of patients’ zip-code (divided into quartiles) 

on both the likelihood of receiving thrombolytics generally, 

and on the likelihood of receiving IA thrombolytics. 

Compared with Caucasians, the odds ratio (OR) of receiving 

thrombolytics for non-Caucasians was 0.719 (95% CI, 

0.650-0.796). For IA, the OR for non-Caucasians was 0.768 

(95% CI, 0.596-0.988). Compared with patients in the highest 

quartile for zip code median income, the OR of receiving 

thrombolytics for patients in the first, second, and third quartile 

was 0.559 (95% CI, 0.455-0.685), 0.729 (95% CI, 0.660- 

0.806), and 0.878 (95% CI, 0.801-0.964), respectively. For 

IA, the OR were 0.462 (95% CI, 0.274-0.779), 0.535 (95% 

CI, 0.515-0.847), and 0.953 (95% CI, 0.947-0.959). 

CONCLUSION 

Non-Caucasian race and lower socioeconomic level profoundly 

decrease a patient’s probability of being treated with 

thrombolysis for acute stroke. There are two possible explanations 

for this. First, as this therapy can only be given if the 

patient presents within 3-6 hours of the onset of stroke symptoms, 

it may be certain groups are less likely to identify the 

symptoms of stroke and activate emergency services. This 

explanation would suggest the need for aggressive education 

of these populations about the symptoms of acute stroke. 

Alternatively, it may reflect systemic issues within the 

healthcare system that disadvantage certain groups. 

Interventional neuroradiologists need to be at the forefront of 

efforts to correct these disparities. 

KEY WORDS: Thrombolysis, race, economics 


Screening for Unruptured Intracranial Aneurysms: A 

Cost-Effective Analysis 

Lin, E. · Platnick, J. · Shen, P. · Pile-Spellman, J. 


New York, NY 

PURPOSE 

Intracranial aneurysms are common and when they rupture 

may cause significant morbidity and mortality. Screening 

and treating unruptured aneurysms would significantly 

Wednesday

Wednesday 

reduce the incidence and complications from rupture. Using 

MR angiography (MRA) and an endovascular approach, 

aneurysms can be diagnosed and treated with minimal risk to 

the patient. However, the economic implications of a screening 

program is yet unknown. We assessed the hypothesis that 

it is cost effective to screen and treat unruptured intracranial 

aneurysms in the general population. 


Decision tree and Markov analysis were used to determine 

the cost effectiveness of screening a random population for 

intracranial aneurysms. The models grouped the patients into 

four health states each with an associated health-related 

quality of life score. Additionally, each health state was 

assigned a designated medical cost based on current reports. 

A comparison was made between the two hypothetical 

groups, the screened population and the nonscreened population, 

to determine the additional costs and quality-adjusted 

life years (QUALYs) gained. 

RESULTS 

The model assumed a screening age of 50 years old, an incidence 

of unruptured intracranial aneurysms of 3.52%, and a 

rupture rate of 2%. The additional cost of the screening program 

as compared to no screening, when using an endovascular 

approach for therapy, was found to be $19,500 per 

QUALY, with a net gain of 0.04 QUALY. Differences in the 

rupture rate had significant impact on the results; with a 1% 

rupture rate, the cost per QUALY significantly increased to 

$280,000. Endovascular therapy for unruptured aneurysms 

yielded a significantly lower cost per QUALY ($19,500) as 

opposed to surgery ($220,000). The analysis was most sensitive 

to changes in the rupture rate and the cost of an MRA. 

CONCLUSION 

Based on current reported rupture rates of 2%, screening and 

treating unruptured intracranial aneurysms is cost effective 

and should be recommended. Furthermore, endovascular 

repair of an aneurysm is more cost effective than surgery and 

should therefore become the standard method of therapy. 

KEY WORDS: Screening, aneurysms, cost effective 

Discussion 


1:00 PM - 2:30 PM 

Room 103 


PowerPoint 



148 


3:00 PM - 4:30 PM 

Room 105/106 

(51a) PEDIATRICS: Epilepsy and 

Nonneoplastic Miscellaneous 



(51b) ADULT BRAIN: Demyelinating and 

Miscellaneous Nonneoplastic Lesions 

(51c) SPINE: Functional and Degenerative 

(51d) ADULT BRAIN: New Imaging and 

Postprocessing Techniques 

Moderators: Jill V. Hunter, MD 

A. James Barkovich, MD 


Quantitative and Qualitative Comparison between 3 and 

1.5 T Imaging in the Evaluation of Epilepsy 

Helwig, J. A. 1 · Nesbit, G. 1 · Anderson, J. 1 · Wang, P. 1 · 

Hamilton, B. 2 

1 Oregon Health & Science University, Portland, OR, 

2 University of Utah, Salt Lake City, UT 

PURPOSE 

Patients with medically refractory epilepsy may benefit from 

image or electrophysiologic-guided surgical resection. 

Patients found to have no focal lesion by MR imaging may 

be excluded from surgery while detection of focal parenchymal 

abnormalities depends on the expertise of readers and 

quality of the MR examination. The recent introduction of 3 

T MR clinical imaging offers the potential of detecting with 

greater sensitivity and specificity parenchymal abnormalities 

such as hippocampal sclerosis or focal cortical dysplasia in 

comparison to standard 1.5 T imaging. Additionally, the 

greater signal to noise properties of 3 T scans may aid in 

improved lesion characterization and diagnosis that influence 

management given differing surgical outcomes of various 

lesions. The current study is to evaluate and compare 

epilepsy pulse sequences and lesion conspicuity and diagnosis 

on 3 T and 1.5 T MR imaging. 


The imaging studies of seventeen patients who underwent 

both 3 T and 1.5 T dedicated brain MR imaging for the evaluation 

of epilepsy were reviewed retrospectively by four 

CAQ certified neuroradiologists. 3 T and 1.5 T dedicated 

MR epilepsy protocols consisted of sag T1, ax FSE PD, ax 

FSE T2, cor FLAIR, cor FMPIR, and ax 3D SPGR pulse 

sequences. Quantitatively, pulse sequences were graded on a

four-point scale regarding distortion/artifact, lesion conspicuity, 

gray matter/white matter differentiation, and 

motion. Qualitatively, reviewers assessed site of abnormality 

and most likely diagnosis. Overall imaging assessment 

was judged against surgical pathology or, if pathology not 

available, clinical neurologic impression constituting physical 

examination, EEG, and SPECT/PET results. Sensitivity 

and specificity of lesion detection and diagnostic accuracy 

between 1.5 T and 3T then were compared. 

RESULTS 

Preliminary retrospective evaluation of seventeen patients 

indicates a trend towards higher quantitative grading of 3 T 

epilepsy-tailored pulse sequences in four broad imaging 

parameters compared to 1.5 T. Distortion/artifact, lesion conspicuity, 

gray matter/white matter differentiation, and 

motion all graded better with 3 T than 1.5 T by each neuroradiologist. 

In particular, SPGR imaging and lesion conspicuity 

were graded better with 3 T than 1.5 T. Data collection 

and statistical analysis of quantitative data is ongoing. 

Lesion conspicuity and diagnostic sensitivity/specificity 

appear improved with high field strength MR imaging for 

epilepsy imaging. 3 T improved lesion detection and diagnosis 

in 12/51 cases while 1.5 T improved lesion detection and 

diagnosis in 2/51 cases from analysis of three CAQ certified 

neuroradiologists. Analysis of the fourth tester is pending. 

CONCLUSION 

Preliminary results show 3 T demonstrating an overall trend 

towards higher grading of broad imaging parameters for 

pulse sequences commonly used in epilepsy evaluation and 

increased lesion conspicuity and diagnostic accuracy with 3 

T compared to 1.5 T imaging. 

KEY WORDS: Epilepsy, 3.0 T, seizure 


Thalamic and Hippocampal Diffusion Tensor Imaging 

Abnormalities in Children with Temporal Lobe Epilepsy 

Bernardi, B. · Kimiwada, T. · Juhasz, C. · Makki, M. · Muzik, 

O. · Chugani, D. · Chugani, H. 

Children’s Hospital of Michigan 

Detroit, MI 

PURPOSE 

Previous diffusion tensor imaging (DTI) studies on temporal 

lobe epilepsy (TLE) reported abnormal mean diffusivity 

(ADC) and fractional anisotropy (FA) measurements in the 

epileptogenic hippocampus (HP), as compared to the contralateral 

side. Most commonly, DTI of adults with relatively 

long history of medial TLE, showed in interictal state, an 

increased ADC and decreased FA in the HP presumed 

seizure focus. Since the thalamus (TH) is considered important 

for the regulation of the cortical excitability and seizure 

propagation, we analyzed FA and ADC in the thalami and 

hippocampi of children with partial TLE, with and without 

secondary generalization. We hypothesized that changes in 

FA and ADC, might be associated to secondary generalization 

of the seizures (sGSZ). 

149 


We evaluated 14 children with unilateral TLE and 6 normal 

controls. Three patients had unilateral hippocampal sclerosis 

(HS), 1 HS and temporal cortical dysplasia, 3 temporal 

tumors (1 DNET, 2 low-grade gliomas), 2 temporal 

choroidal cyst, and 5 negative MR images. No thalamic signal 

abnormalities were noted. Diffusion tensor imaging was 

acquired in patients seizure free for at least 12 hours. Singleshot 

EPI was performed on 40 axial planes with diffusion 

gradients in six directions (b = 1000 s/mm 2 ). The same imaging 

parameters were applied to acquire one T2-weighted 

imaging volume. Fractional anisotropy and ADC were 

obtained in the thalami and hippocampi, including almost 

the whole thalamus and the hippocampal head. 

RESULTS 

All patients had significantly increased ADC and decreased 

FA values in the HP ipsilateral to seizure focus. The ADC 

and FA in the TH ipsilateral to the epileptic focus showed 

significantly higher values than on the contralateral side in 

children with sGSZ. Patients without seizure generalization 

showed increased ADC but did not show FA difference 

between the two thalami (Table). 

ADC FA 

*10-9[mm 2 /s] 

Ipsilateral Contralateral p- Ipsilateral Contralateral p 

Thalamus Whole 682±17 676±18 0.01 0.307±0.043 0.295±0.039 0.001 

group(n=14) 

PE + sGSZ 684 ± 16 676 ± 18 0.015 0.314 ± 0.055 0.297 

± 0.049 0.006 

(n=7) 

PE - sGSZ 680 ± 19 676 ± 20 0.26 0.302 ± 0.031 0.292 ± 0.031 0.062 

(n=7) 

Hippocampus WG 831 ± 79 779 ± 35 0.031 0.162 ± 0.028 0.180 

± 0.020 0.019 

(n-14) 

PE + sGSZ 869 ± 96 781 ± 32 0.048 0.165 ± 0.030 0.180 

± 0.020 0.106 

(n=7) 

PE - sGSZ 800 ± 47 778 ± 40 0.39 0.160 ± 0.028 0.108 ± 0.023 0.111 

(n-7) 

[PE + sGSZ] = partial epilepsy with sGSZ [PE - sGSZ] = partial epilepsy without sGTC 

CONCLUSION 

Increased ADC and decreased FA has been interpreted as 

result of structural organization loss and enlarged extracellular 

space. The increased ADC in the TH ipsilateral to the 

seizure focus, observed in all patients, are consistent with 

extension of DTI changes to the TH. In the thalami ipsilateral 

to the seizure focus, different changes in FA are demonstrated, 

according to the presence or absence of sGSZ. In 

sGSZ, the increased FA, suggests involvement of this structure, 

perhaps due to its recruitment into epileptic network. 

Potential explanation might be increasing coherence in preserved 

connections, when there is selective fibers disruption. 

Moreover, because several patients were children with cortical 

lesions, developmental alterations of the cortico-thalamic 

connectivity, might be partially responsible for DTI abnormalities. 

Nevertheless, DTI is sensitive to detect remote subcortical 

abnormalities, even in negative conventional MR 

imaging. The potential significance of the observed changes 

in relation to the seizure semiology and the role of DTI in 

outcome prediction require further investigations in a larger 

patient population. 

KEY WORDS: Diffusion tensor imaging, temporal lobe 

epilepsy, thalamus 

Wednesday

Wednesday 


Presumable Cause of Reversible MR Signal Changes in 

the Splenium of the Corpus Callosum in Miscellaneous 

Pathology: Assessed by Diffusion-Weighted MR Images 

and MR Spectroscopy 

Kawamura, Y. 1 · Watanabe, K. 2 · Tsukahara, H. 1 · Mayumi, 

M. 1 · Itoh, H. 1 

1 2 University of Fukui, Fukui, JAPAN, Okazaki Municipal 

Hospital, Okazaki, JAPAN 

PURPOSE 

In this study, reversible signal abnormality in the splenium 

of the corpus callosum was revealed on T2-weighted images, 

FLAIR images, and EPI diffusion images in various central 

nervous disorders (Table). 


Although the exact mechanism of the lesion in the splenium 

is unclear, its reversibility indicates that ischemic damage 

and cytotoxic edema is not likely the cause. The marked high 

intensity abnormality seen on diffusion-weighted images 

was reduced when motion probing gradient (MPG) was 

applied in X direction parallel to the myelin fibers course in 

the splenium. On T2-weighted images and FLAIR images, 

high intensity change is not as remarkable as on diffusionweighted 

images. 

RESULTS 

MR spectroscopy obtained in three patients (two infectious 

encephalopathy cases and one epileptic case) did not show 

significant decrease in NAA and no significant lactate peak. 

Patients with High Intensity Leison in the Splenium of the Corpus Callosum 

Case Age/sex Symptoms Pathology 

Number 

1 8/female fever, convulsion, diarrhea, Salmonella 

Enteritidis 

2 2/female fever, convulsion, delirium Rotavirus 

3 9/female fever, convulsion, disorientation Influenza A 

4 5/male fever, convulsion, disorientation Influenza A 

5 5/male fever, disorientation Adenovirus 

6 4/male fever, convulsion, status epilepticus Influenza A 


8 8/male status epilepticus Epilepsy 


10 15/male irritation Anticonvulsants for 

previous seizure 

11 21/female no symptom Anticonvulsants for 


12 25/female lethargy Anticonvulsants for 

schizophrenia 

13 20/female headache Shering injury 

(traffic accident) 

14 18/female somnolence Osmolarity Disorder due 

to Anolexia Nervosa 

150 

CONCLUSION 

The authors presume that the myelin edema is the main 

cause of this transient signal abnormality in the splenium of 

the corpus callosum in various central nervous disorders. 

KEY WORDS: Corpus callosum, MR imaging, diffusion 


Diffusion-Weighted MR Imaging and Proton MR 

Spectroscopy of the Brain in Children with Chronic 

Liver Disease 

Abdel Razek, A. A. · Abdalla, A. · Ezz, M. · Sadek, A. · 

Kandel, A. · Rabeah, O. 



PURPOSE 

To determine the role of diffusion-weighted MR imaging 

and proton MR spectroscopy in children with chronic liver 

disease. 


Diffusion-weighted MR imaging and multivoxel proton MR 

spectroscopy of the brain were done for 27 pediatric patients 

(19 M, 8 F aged 4-14 years: mean 9 years) with chronic liver 

disease and 20 volunteers as controls. Diffusion MR weighted 

imaging was done using a single-shot echo-planar imaging 

with a diffusion-weighted factor b of 0.500 and1000 

sec/mm2. The apparent diffusion coefficient (ADC) map 

was reconstructed. Proton MR spectroscopy (multivoxel 

PRESS technique: TR/TE = 1500/135, 1.6 X 1.6 matrix, 

FOV = 100 mm, 2 averages with 2 ml spatial resolution) was 

performed at the level of most significant MR abnormality. 

Assignment of the resonance peaks of NAA, creatine, 

choline, glutamate, myoinositol, lipid and lactate were determined 

in the basal ganglion, thalamus, gray and white matter. 

RESULTS 

Restricted diffusion (high signal intensity at b = 1000 images 

and low signal intensity at ADC maps) within the corticospinal 

tract in hepatic encephalopathy (n = 17), putamen 

and caudate nucleus in Wilson’s disease (n = 4) and frontal 

and occipital white matter in Gaucher’s disease. MR spectra 

of hepatic encephalopathy showed increased glutamate peak 

with decreased myoinositol and choline peaks. Wilson’s disease 

showed decreased choline and myoinositol that associated 

with lactate. MR spectra in Gaucher’s disease revealed 

presence of lipid and lactate that associated with decreased 

choline and NAA. 

CONCLUSION 

Diffusion-weighted MR imaging and proton MR spectroscopy 

added valuable information to routine MR imaging 

of the brain in children with chronic liver disease. Further 

studies will be needed to determine if abnormalities detected 

with diffusion MR imaging and proton MR spectroscopy can 

be used as a marker to assess long-term functional outcome. 

KEY WORDS: Diffusion, MR spectroscopy, pediatric


MR Imaging and CT in Central Nervous System 

Involvement of Hemolytic Uremic Syndrome: Prognostic 

Value of Lesion Detection 

Donnerstag, F. G. F. · Hartmann, H. · Luecke, T. · Haubitz, B. 

· Das, A. M. · Ehrich, J. · Becker, H. 

Hannover Medical School 

Hannover, GERMANY 

PURPOSE 

Cerebral involvement occurs in approximately 30% of children 

with hemolytic uremic syndrome (HUS) and correlates 

with poor prognosis. We aimed to investigate the value of 

cerebral imaging, especially MR studies, for the prognosis of 

cerebral involvement. 


We reviewed the CT and MR images from 22 patients with clinically 

proven HUS who were referred to our clinic between 

January 1996 and September 2003. In 16 cases enterohemorrhagic 

E. coli were identified as causative organisms, 1 was associated 

to pneumococcal disease and in one a significant antibody 

titer to brucellosis was found. All patients required dialysis. 

Clinical signs of central nervous system (CNS) involvement 

included altered conciousness in all patients. Twelve had seizures 

or status epilepticus. Four patients died due to HUS, 3 showed a 

severe residual syndrome and 5 neurologic deficits including 1 

hemiparesis. Electroencephalographic recordings were obtained 

in 21 patients and showed abnormalies in 20 patients. Ten patients 

only had CT, 11 had CT and MR imaging, and 1 only MR imaging. 

All MR imaging but one included diffusion-weighted imaging. 

Sequential CT and MR imaging were analyzed; 3 patients 

had only 1 CT, 1 patient only 1 MR imaging. 

RESULTS 

Abnormal CT findings included hypodensities of the basal 

ganglia and thalami with or without the adjacent white matter 

in 9, in the deep white matter in 2, in the subcortical white 

matter in 2 and generalized edema in 5 patients. 

Hemorrhagic transformation was seen in 2 patients. 

Abnormal MR imaging findings included abnormal signal 

intensities in the basal ganglia and thalami in 6 patients, 2 

had reduction of the apparent diffusion coefficient in the 

acute phase. Four patients showed reduction of ADC in the 

fronto-parietal white matter. Signs of cortical infarction were 

seen in 2, infratentorial abnormalities in another 2 patients. 

From the group of patients who were examined with both CT 

and MR imaging, in 4 patients only diffusion-weighted 

imaging could show pathologic intensities with a normal 

FLAIR. In 2 patients both modalities were normal in the 

early exam; one of them had no residual neurologic sequelae, 

the other slight persistent neurologic impairments. The 

extent of lesions did not correlate to the severity of residual 

neurologic damage. 

CONCLUSION 

MR imaging, especially diffusion-weighted imaging, may 

help to demonstrate early cerebral involvement in HUS. 

However, signal changes do not appear to predict the course 

of the disorder and the severity of residual neurologic damage. 

KEY WORDS: Hemolytic, uremic, syndrome 

151 


Study of the Spectrum of Childhood Leukodystrophies: 

Correlation of Imaging, Spectroscopic, and Biochemical 

Features 

Sinha, A. · Sharma, R. · Gupta, A. K. · Gulati, S. · Kabra, M. 

· Kalra, V. 

All India Institute of Medical Sciences 

New Delhi, INDIA 

PURPOSE 

Characterization of suspected leukodystrophies with MR 

imaging and proton MR spectroscopy with biochemical correlation. 


Twenty-five patients with suspicion of primary white matter 

degenerative disease or evidence of white matter abnormalities 

on preliminary imaging were evaluated by MR imaging 

and ¹H MR spectroscopy (MRS) on a 1.5 T MR scanner. T1weighted 

axial, fast spin-echo T2-weighted axial, sagittal, 

coronal and FLAIR images were acquired. Single voxel ¹H 

spectroscopy was carried out using point resolved spectroscopy 

(PRESS) with an echo time (TE) of 135 ms and 30 

ms, a repetition time of 6000 ms, and 64 acquisitions. ¹H MR 

spectroscopy also was done in 19 control patients. NAA (Nacetyl 

aspartate): Cr (Creatine), Cho (Choline): Cr, 

NAA:Cho, Ins(Myo-Inositol):Cr, Glx (Glutamates):Cr ratios 

were compared between controls and patients. Values less 

than or more than one standard deviation from the mean in 

controls were taken as abnormal. 

RESULTS 

The patients were classified into four groups based on imaging 

features: 1. Megalencephalic leukoencephalopathy with 

subcortical cysts (van der Knaap disease) - 9 patients had 

diffuse white matter hyperintensity with subcortical cysts in 

temporal and/or frontoparietal lobes. 2. 

Adrenoleukodystrophy (ALD) - 4 patients had posterior predominant 

white matter hyperintensity with involvement of 

corpus callosum. 3. Unclassified leukodystrophy - 8 patients 

had evidence of leukodystrophy on MR imaging but could 

not be definitively classified into any known category. On 

the basis of MR imaging features, a differential diagnosis of 

the likely leukodystrophies was given. 4. Equivocal - 4 

patients had evidence of mild, subtle, or no changes in white 

matter on MR imaging and the presence or absence of 

leukodystrophy on imaging could not be established. In nine 

patients in our first subgroup (van der Knaap disease), we 

could offer a definitive diagnosis on the basis of clinical features, 

imaging, and spectroscopic analysis. Ours is the first 

study to report increase in glutamates in this condition. There 

is no biochemical marker available for confirmation. FLAIR 

sequence was of great benefit in this subgroup, not only for 

demonstration of cysts, but also for voxel localization. Two 

patients in the ALD subgroup had occipital predominant disease 

with inhomogeneous zones of enhancement, and 

showed decreased NAA:Cr and NAA:Cho and increased 

Cho:Cr ratios. These two patients were confirmed biochemically 

by raised very long chain fatty acids (VLCFA) levels. 

However, in the majority of our patients, there was significant 

overlap in the MRS findings among the subgroups. 

Wednesday

Wednesday 

CONCLUSION 

MR imaging and MRS were sensitive for detection of white 

matter abnormality but were nonspecific. In its present state 

of evolution, we found MRS to be inadequate for characterization 

of leukodystrophies. In the absence of standardized 

values, it was difficult to come to a definitive conclusion. 

Larger studies, with more patients and more easily available 

and definitive biochemical and genetic evaluation are needed 

for further characterization of this rare group of diseases. 

KEY WORDS: Leukodystrophy, MR imaging, proton MR 

spectroscopy 


Brain Asymmetry on MR Imaging Analyzed with Voxel- 

Based Morphometry in the Deaf 

Shibata, D. K. 


Seattle, WA 

PURPOSE 

The lateralization of language function is clinically important 

in preoperative planning and lesion functional localization 

although the factors which influence the development of 

brain asymmetry are unclear. Speech acquisition is felt to be 

an important factor in brain lateralization and purpose of this 

study was to compare the structural asymmetries of deaf versus 

control volunteers with volumetric brain MR imaging 

scans using an automated image-processing technique, 

voxel-based morphometry (VBM). 


Fifty-three right-handed prelingually deaf students from a 

deaf college were compared with 51 hearing right-handed 

subjects. 3D SPGR 1.5 mm contiguous axial T1-weighted 

images on a 1.5 T scanner were processed with an optimized 

protocol for VBM analysis performed using SPM2. Gray 

matter, white matter, and CSF segmentations were spatially 

normalized and then these deformation parameters were 

applied to the original images which then were segmented 

again. The images then were modulated for volume changes 

and smoothed and right-left reversed images were compared 

with the original images using a T-test statistic to determine 

asymmetries. 

RESULTS 

The deaf and hearing groups showed foci of asymmetry in 

similar regions, but the largest two foci, the left perisylvian 

cortex and the right occipital cortex were of different extent 

in the two groups. For the perisylvian cortex, the control 

group had larger gray matter clusters at higher statistical significance 

compared with the deaf (82,797 vs 69,773) and for 

the occipital cortex, the deaf group was larger ( 8,128 vs 

5,332). Similar differences were seen in both the gray matter 

and white matter maps. Differences also were seen in smaller 

cortical and subcortical foci. 

CONCLUSION 

These results support the hypothesis that early speech may 

play a role in the structural lateralization of perisylvian language 

cortex. Although VBM did detect asymmetries in the 

deaf brains, the degree was less than the hearing controls. 

152 

This supports the neurologic and functional imaging findings 

on deaf sign language which has shown a left lateralization. 

The greater lateralization of visual cortex may reflect 

the use of sign language. Voxel-based morphometry may be 

a powerful tool in accessing the role of early sensory and 

language function on brain development. 

KEY WORDS: Deaf, anatomy, asymmetry 

Acknowledgments: The National Technical Inst of the Deaf 

at the Rochester Inst of Tech, Rochester, NY provided assistance 

with subjects. Rachel Yotter, Electrical Engineering, 

Univ of Washington, assisted with data analysis. 


Triangular Crossroads: A “Wetterwinkel” of the Fetal 

Brain 

Prayer, D. 1 · Brugger, P. C. 1 · Judaš, M. 2 · Radoš, M. 2 · 

Kasprian, G. 1 · Kostovic, I. 2 

1 2 Medical University of Vienna, Vienna, AUSTRIA, School 

of Medicine, University of Zagreb, Zagreb, CROATIA 

PURPOSE 

To assign the fetal MR imaging finding of transient MR 

imaging signal intensity changes in the triangular area lateral 

to the exit of the posterior limb of the internal capsule to a 

histologic background and to discuss the significance of 

apparently “pathologic” appearance of this area. 


In vivo MR images of 430 fetal brains between the 18th and 

39th gestational week (GW) were performed on a 1.5 T 

superconducting unit, using ultrafast T2-weighted, T1weighted 

and diffusion-weighted sequences in 3 orthogonal 

section-planes. The normal group (287 cases) resulted from 

examination done for extra CNS pathologies. In the other 

group (143 cases) cerebral pathology had been suspected by 

ultrasound or was diagnosed with fetal MR imaging. In addition 

to routine assessment of the fetal brain (including the 

assessment of lamination of the brain parenchyma, gyration 

and sulcation, and premyelination), the area lateral to the exit 

of the posterior limb of the internal capsule was evaluated 

with respect to signal intensity, size, and shape. Findings in 

normal fetuses were compared with Nissl-stained, AChEstained, 

and PAS-Alcian blue stained sections of agematched 

specimens. 

RESULTS 

In fetuses between the 19th-22nd GW a triangular hyperintense 

area is present on axial slices adjacent to the germinal 

matrix in the continuity of the posterior limb of the internal 

capsule, tip pointing medially (dorso-lateral to the basal ganglia). 

These triangles were hyperintense (but not CSF 

intense) on T2-weighted images, and iso or hypointense to 

the subplate on T1- and diffusion-weighted images until the 

24th GW, to fade eventually until the 28th GW and became 

isointense with adjacent structures by the 30th GW. The size 

of the structures did not exceed 4 side length. In different 

pathologies, such as periventricular leukomalacia (2), metabolic 

diseases (1), malformations (9) the triangular regions 

were larger than normal, and/or presented with CSF intensity, 

and/or persisted until after the 28th GW. In the remaining

131 cases with cerebral pathology these regions were either 

not involved or the severity of brain pathology precluded 

their evaluation. Histologically, normal findings showed a 

high extracellular matrix content in this area, with a merging 

point of two intensely AChE-staining stripes at the tip of the 

triangle. The region was identified as an area where internal 

capsule fibers fan out towards the occipital lobe and intersect 

with long corticocortical fiber bundles. 

CONCLUSION 

The transient MR imaging features of the triangular crossroads 

may be explained by transiently increased water content 

of the extracellular matrix (ECM); these MR features 

gradually become less conspicuous with ongoing cellular 

maturation, and finally become indistinct with respect to 

their surroundings. Increased signal intensity/size or persistence 

probably reflects oxidative stress of ECM-producing 

cells and axonal guidance cues or impairment of fiber development 

in this region, due to a cellular damage, or disturbed 

mechanisms of intercellular signaling. Normal triangular 

crossroads should not be misdiagnosed as ischemic lesions. 

However, this region seems to be a “Wetterwinkel” of the 

fetal brain, as altered morphology and signal intensities in 

this area may be an early indicator of cerebral pathology. 

KEY WORDS: Fetal MR imaging, fetal brain, diffusionweighted 

imaging 


Correlation between the Neuroanatomy and the 

Endocrinology in Patients with Optic Nerve Hypoplasia 

Mehta, H. 1 · Mehta, A. 2 · Thompson, C. 1 · Dattani, M. T. 2 · 

Chong, W. K. 1 

1 Great Ormond Street Hospital, London, UNITED KING- 

DOM, 2 Great Ormond Street Hospital & Institute of Child 

Health, London, UNITED KINGDOM 

PURPOSE 

Optic nerve hypoplasia (ONH) when occurring with midline 

forebrain defects (MFD) [abnormal septum pellucidum (SP) 

or corpus callosum (CC)] and/or hypothalamo-pituitary (HP) 

abnormalities is termed septo-optic dysplasia. We aimed to 

assess if neuroimaging in patients with clinical ONH predicts 

endocrine dysfunction. 


MR imaging [size of anterior pituitary (AP) and ON; posterior 

pituitary (PP) morphology; morphology of the SP, CC 

and HP stalk] and endocrine function of 45 patients with 

ONH (22 males, mean age 2.3 years) were analyzed retrospectively. 

Anterior pituitary hypoplasia (APH) was defined 

as size < -2SD from the mean of normal children (published 

data). 

RESULTS 

Thirty-nine of 45 patients had bilateral ONH, 14 of whom 

had hypopituitarism [growth hormone deficiency 100%, thyrotropin 

deficiency 71.4%, adrenocorticotropin deficiency 

78.6%, diabetes inspidus (DI) 7.1%]. Optic nerves were normal, 

on imaging, in 4/45 patients. Hypopituitarism was significantly 

(p = 0.04) associated with MFD (11/25) as compared 

to those without (3/20). APH, diagnosed on gross 

153 

inspection, was more sensitive (sensitivity 92.9%, specificity 

45.2%) in predicting AP dysfunction than a measured AP 

size (sensitivity 78.6%, specificity 77.4%), although the latter 

was more specific. An ectopic (8) or absent (11) PP did 

not predict DI. Prolactin concentrations were significantly 

greater in patients with an abnormal (480 mU/L) compared 

to those with a normal stalk (258 mU/l; p = 0.03). Patients 

with normal AP, SP, and CC had normal endocrinology. 

CONCLUSION 

Neuroimaging in patients with ONH can help predict 

endocrine dysfunction. All patients with ONH should have 

neuroimaging. MR imaging is less invasive than dynamic 

endocrine tests and in this study a normal AP and forebrain 

was predictive of normal endocrine function. Patients with 

abnormal imaging but normal endocrinology may indicate a 

subgroup who evolve into hypopituitarism later. 

KEY WORDS: Septo-optic dysplasia, neuroanatomy, 

endocrinology 


Correlation of Brain Pathology with Chromosomal 

Abnormalities: Development of a MR Imaging 

Phenotypic Database 

Forbes, K. P. 1 · Vezina, G. 2 · Gropman, A. 3 · Rosenbaum, K. N. 2 

1 Institute of Neurological Sciences, Glasgow, UNITED 

KINGDOM, 2 Children’s National Medical Center, 

Washington, DC, 3 Georgetown University Medical Center, 


PURPOSE 

Genetic control of brain development is complex, involving 

many different genes on many chromosomes. Precise gene 

expression at various stages of brain development depends 

both on chromosomal structure and, as yet unknown, temporal 

signals. The role of specific genes and chromosomal 

regions in brain development is largely unknown. MR imaging 

offers a noninvasive tool to examine brain structure, providing 

a means of obtaining detailed phenotypic information 

on genetic abnormalities. While a number of case reports and 

series have documented brain findings in a range of chromosomal 

abnormalities, there has been no prior report of a database 

systematically correlating brain pathology with abnormal 

autosomal karyotype. The aim of this project is to document 

MR imaging brain findings across a broad range of 

chromosomal abnormalities to determine whether there are 

patterns of brain development associated with specific chromosomal 

alterations. 


Subjects with abnormal karyotyping of autosomal chromosomes, 

excluding trisomy 21, were identified from a pediatric 

genetics clinic, over a 6-year period. All subjects underwent 

1.5 T MR imaging of the brain as part of routine clinical 

work up of their chromosomal abnormality. T1-, proton 

density, and T2-weighted images were reviewed by two 

pediatric neuroradiologists and scored for structural or signal 

changes to the brain (cortex, cerebral white matter, corpus 

callosum, basal ganglia, pituitary-hypothalamic axis, optic 

pathways, brainstem, cerebellum), CSF spaces, and skull. 

Wednesday

Wednesday 

RESULTS 

Forty-four subjects were identified, comprising 23 females 

and 21 males, with a mean age of 4 years 7 months +/-14 

months. Chromosomal abnormalities were widespread, 

affecting 19 of the autosomal chromosomes. Thirty-six subjects 

(82%) showed abnormal MR imaging findings. 

Atrophy/hypoplasia of the cerebral hemispheres was most 

common, seen in 17 subjects (39%), with a wide range of 

chromosomal abnormalities. White matter abnormalities 

comprised delayed myelination in 5 subjects (4p-,dup 5p, t 

(10,12), 3 NOS (not otherwise specified), 15q-), and T2 

hyperintensities in 7 subjects (tri 8, dup 10p, t (11,15), tri 13, 

t(X,16), 18q-, 18 NOS). Corpus callosal agenesis or hypogenesis 

was present in 5 subjects (2p-, tri 8 (2 subjects), dup 

10p, t (11;15)). Polymicrogyria occurred in 3 subjects (2p-, 

tri 8, dup 8). Abnormalities of the posterior fossa comprised 

pontine hypoplasia in 7 subjects (3 NOS, 4 NOS, tri 8, tri 18 

(2 subjects), t (10,12), t (10,NOS) and vermian hypoplasia in 

4 [3 (NOS), dup 10p, tri 13 and tri 18]. Anomalies of the 

skull base or craniocervical junction occurred in 3 subjects [t 

(6;15), tri 6p, 7q-/10q-]. A number of chromosomal abnormalities 

(1q-, mosaic ring 2,tri 6, 7q-, dup 11q, tri 13,20 

NOS,22-) were associated with normal MR imaging findings. 

CONCLUSION 

Pathologic brain findings are common in chromosomal 

abnormalities. The size of the current database precludes 

firm conclusions but suggests that while some MR imaging 

findings may be nonspecific e.g., atrophy, delayed myelination, 

the presence of similar findings in multiple subjects 

with the same affected chromosome provides tentative support 

for a potential genetic link for callosal agenesis, polymicrogyria, 

and pontine hypoplasia. Continued expansion of 

this database will allow further assessment of patterns of 

brain development in chromosomal abnormalities. 

KEY WORDS: Genetic, chromosome, MR imaging 


Dynamic MR Perfusion and Proton MR Spectroscopic 

Imaging in Sturge-Weber Syndrome: Correlation with 

Neurologic Symptoms 

Lin, D. D. · Hatfield, L. A. · Comi, A. M. · Barker, P. B. 

The Johns Hopkins University 

Baltimore, MD 

PURPOSE 

Sturge-Weber syndrome (SWS) has well characterized neuroimaging 

features marked by atrophy, cortical calcification, 

and leptomeningeal angiomatosis. In this study we investigated 

some of the physiologic alterations including cerebral 

perfusion and metabolite profile in these patients using 

dynamic susceptibility contrast MR perfusion imaging and 

multislice spectroscopic imaging, and correlated the findings 

with neurologic status. 


Six consecutive patients with clinically established diagnosis 

of SWS underwent MR imaging using a 1.5 T scanner. In 

addition to routine brain MR sequences, contrast-enhanced 

gradient-echo echo-planar PWI was acquired, with recon- 

154 

struction of hemodynamic data using first-pass tracer kinetic 

methods. Spin-echo proton MR spectroscopic imaging 

was performed with three 15 mm axial slices covering from 

the midbrain to the centrum semiovale. Neurologic scores 

were established independently by a pediatric neurologist for 

all six patients based on neurologic examinations, history, 

and seizure log. The quantified image parameters then were 

correlated with neurologic scores using nonparametric correlation 

analysis. 

RESULTS 

Four patients had classic neuroimaging findings of predominantly 

unilateral cerebral atrophy with corresponding leptomeningeal 

enhancement and hypoperfusion. Perfusion 

scans showed significant deficit as demonstrated by prolonged 

MTT, with signal intensity time-series curve demonstrating 

normal bolus arrival but protracted recovery phase. 

One patient had bilateral disease with hypoperfusion in bilateral 

occipital lobes and right parietal lobe. Two of six 

patients with an established diagnosis of SWS had normal 

perfusion patterns. Among clinical measures of seizure 

activity, hemiparesis, cognitive deficit, visual field cut, and 

total disability, there was highest correlation between hemiparesis 

and hypoperfused tissue volume (Spearman’s correlation 

coefficient, ρ = 0.943, P < 0.05). There was also a 

trend of correlation, although not statistically significant 

(most likely related to small sample size) (P = 0.06), between 

hemiparesis score and NAA/Cr ratio in the mid to posterior 

centrum semiovale, lateral gray matter, and splenium of the 

corpus callosum, regions closely matching those of leptomeningeal 

angiomatosis and perfusion deficits in the most 

affected hemisphere. A similar trend was observed between 

the NAA/Cr, Cho/Cr ratios, and total neurologic score in the 

mid centrum semiovale and splenium (P = 0.06). 

CONCLUSION 

In SWS, perfusion MR imaging indicates cerebral hypoperfusion 

(prolonged MTT) due to impaired venous drainage. 

The extent and severity of perfusion abnormality, as well as 

neuronal loss/dysfunction (as indicated by reduced NAA/Cr) 

directly reflect the severity of neurologic symptoms and disability, 

with the highest correlation with degree of hemiparesis. 

MR perfusion imaging can be a viable alternative to PET 

and SPECT for the investigation of the pathophysiology of 

SWS, and together with MR spectroscopy may be useful as 

quantitative measures of disease burden. 

KEY WORDS: Sturge-Weber syndrome, MR perfusion, MR 

spectroscopy


3:00 PM - 4:32 PM 

Theatre 

(51b) ADULT BRAIN: Demyelinating 

and Miscellaneous Nonneoplastic 

Lesions 



(51a) PEDIATRICS: Epilepsy and Nonneoplastic 

Miscellaneous 




Moderators: Robert D. Zimmerman, MD 

Soonmee Cha, MD 


Corticospinal Tract Degeneration in Brainstem in 

Patients with Multiple Sclerosis: Evaluation with 

Diffusion Tensor Tractography 

Ge, Y. · Tuvia, K. · Law, M. · Johnson, G. · Herbert, J. · 

Mannon, L. · Grossman, R. I. 

New York University School of Medicine 

New York, NY 

PURPOSE 

The hallmark of multiple sclerosis (MS) is mutlifocal lesions 

characterized by inflammatory demyelinating changes that 

primarily involve the white matter tracts in the brain. There 

is increasing evidence demonstrating that distal wallerian 

degeneration in MS may occur secondary to axonal transection 

or damage due to MS lesions more proximally (1). 

Diffusion tensor imaging provides physiologic information 

with in vivo visualization and quantitative assessment of 

these white matter pathways. The purpose of this study was 

to evaluate the effects of supratentorial lesions on corticospinal 

tract function and integrity by measuring fractional 

anisotropy (FA) and fiber tract number (FT) at the brainstem 

level. 


Nine patients with clinical relapsing remitting MS disease 

were studied on a 3.0 T MR unit. After conventional MR 

imaging, which included T2- and enhanced T1-weighted 

images, axial DTI data were acquired with a pulsed gradient, 

double spin-echo, echo-planar imaging (5300/74; 128 x 128 

matrix; 220 x 220 mm FOV; forty-five 3 mm contiguous 

slices; b = 1000 s/mm2) in six directions. Fixed seeds/ROIs 

155 

(4 voxels with a total of 32 tracts) were placed within each 

side of corticospinal tract in cerebral peduncle at middle 

brainstem level without observed lesions. The measurements 

for FA and FT then were made in both patients and controls. 

Fiber tracks were constructed using software that allowed 

variation of the threshold to follow the primary eigenvector 

voxel-by-voxel based on both quantitative magnitude and 

directional information. Both lesion number and lesion volume 

in both cerebral hemispheres were calculated in 

patients. 

RESULTS 

Fiber tractography accurately delineated the corticospinal 

tract from cerebral peduncle in brainstem through internal 

capsule to cerebral cortical gyri. The FA was significantly 

lower (FA = 0.52) in patients with higher lesion load (volume 

> 1360 mm 3 ) as compared with those patients (FA = 0.63) 

with lower lesion load (p = 0.03). Correspondingly, there 

were fewer fibers generated in the corticospinal tracts in 

patients with higher lesion load. Also patients with lesions in 

the corticospinal tracts showed lower FA values and lower 

number of fiber tracts as compared with those who don’t 

have lesions on this pathway. In addition, the total number of 

fiber tracts was lower in the hemisphere with more lesions. 

CONCLUSION 

Wallerian degeneration in white matter tracts has been 

shown to occur in cerebral infarction with decreased 

anisotropy values (2). We have demonstrated that the degree 

of fiber tract loss in corticospinal tracts at the level of the 

brainstem level in MS patients is likely to relate to the lesion 

load in the supratentorial brain. This suggests fiber tractography 

can provide a method for quantifying wallerian degeneration 

and axonal transection from remote lesions in MS. 

REFERENCES 

1. Trapp BD. N Engl J Med 1998;338:278-285 

2. Zelaya F. Magn Reson Imag 1999;17:331-348 

KEY WORDS: Multiple sclerosis, diffusion tensor imaging, 

wallerian degeneration 

Acknowledgment: This work was supported by grants R37 

NS29029-11 from NIH and NCRR M01 RR00096 (GCRC). 


Optimization of Echo Time Increases Observer 

Performance in the Detection of Artificial Multiple 

Sclerosis Lesions on Simulated FLAIR Images of the 

Brain 

Pikus, L. · Woo, J. H. · Wolf, R. L. · Herskovits, E. H. · 

Moonis, G. · Jawad, A. · Krejza, J. · Melhem, E. R. 

University of Pennsylvania 


Withdrawn 

PURPOSE 

There have been no previous studies conducted to establish 

optimal FLAIR parameters for detecting multiple sclerosis 

(MS) lesions; thus our study was designed to determine an 

optimal echo time (TE) in detecting simulated MS lesions on 

FLAIR images of the brain. 

Posters 

Wednesday

Wednesday 


MR parametric maps were derived at supra and infratentorial 

levels from a mixed multiecho inversion recovery scan of 

a healthy volunteer. Similar maps of a MS lesion were 

obtained in a patient with MS. Pixel-wise solution of the 

Bloch equation enabled us to create images of FLAIR 

(TR/TE/TI: 11000/100-200/2600 ms) contrast, in which we 

placed simulated lesions of varying size, number, and location. 

For each of the 11 TE values spaced 10 ms apart from 

100 to 200 ms, 12 images were created at either axial level, 

for a total of 264 images. In each group of 12 images, nine 

contained the same 18 lesions distributed pseudo-randomly, 

and the other three had no lesions. Selected lesion sizes (2, 

3, and 4 pixels) and parenchymal locations (cortical, deep 

white matter, and periventricular) were represented equally. 

Automated software displayed the same 264 images to 7 

neuroradiologists with various levels of experience, who 

independently rated suspicious lesions on a 4-point scale 

reflecting their confidence level. Undetected “true” lesions 

were assigned a rating of 0. Observer performance was 

measured by the area under the alternative FROC curve 

(A1). Standard errors were estimated by the jackknife 

approach implemented in the JAFROC software. 

RESULTS 

Average performance of all observers across all locations 

differed significantly among the TE values (repeated measures 

ANOVA, p < 0.05), with a substantial decrease in performance 

with increasing TE. In the supratentorial region, 

the average A1 value at TE of 100 ms was higher compared 

to other TEs: 0.93 ± 0.09 versus 0.88 ± 0.08, 0.83 ± 0.12, 

0.81 ± 0.12, 0.80 ± 0.13, and 0.77 ± 0.17 at TE of 120 ms, 

140 ms, 160 ms, 180 ms, and 200 ms, respectively (Fig. 1). 

In the infratentorial region, the respective values at TE of 

100 ms were also higher compared to other TEs: 0.87 ± 0.10 

versus 0.81 ± 0.10, 0.72 ± 0.12, 0.73 ± 0.14, 0.63 ± 0.17, and 

0.58 ± 0.15 for TE of 120 ms, 140 ms, 160 ms, 180 ms, and 

200 ms, respectively (Fig. 2). 

CONCLUSION 

Observer performance in the detection of MS lesions on simulated 

FLAIR images was very high at TE of 100 ms, with 

lower performance at other TEs, both supra and infratentorially. 

This result suggests that most current implementations 

of FLAIR imaging that use a higher TE may not be optimizing 

the detectability of MS. 

KEY WORDS: Multiple sclerosis, FLAIR, echo time 

156 


Lesion Evolution Using Apparent Diffusion Coefficient 

in a Patient with Balo’s Concentric Sclerosis 

Aviv, R. I. 1 · Ball, T. 2 · Quest, R. 2 · Roncaroli, F. 2 · Malik, O. 2 

1 2 Sunnybrook Hospital, Toronto, ON, CANADA, Charing 

Cross Hospital, London, UNITED KINGDOM 

PURPOSE 

The mechanism of evolution of lesions in Balo’s concentric 

sclerosis has not been described previously. Opinions differ 

whether rings of enhancement and T1 and T2 isointensity 

represent demyelination or remyelination and whether the 

lesion grows progressively or synchronously. We present 

apparent diffusion coefficient (ADC) data that demonstrates 

a progressive expansion of the lesion and values, supported 

by histology, that suggest demyelination is the predominant 

process. 


A 27-year-old patient presented with subacute left hemiparesis 

and dysaesthesia evolving stepwise over a 4-week period. 

Examination revealed left pyramidal hemiparesis and loss of 

left proprioception. Cerebrospinal fluid and VEP’s were normal. 

The patient was treated with 1g IV methylprednisolone 

with little effect. Due to worsening left weakness and 

impaired right proprioception she was treated with dexamethasone. 

Brain biopsy showing an inflammatory infiltrate 

consistent with demyelination. Cerebrospinal fluid now 

showed unmatched oligoclonal bands. At 2 months the 

patient relapsed with transient dysphagia, poor concentration, 

and a right hemiparesis, progressing over 3 weeks. She 

was treated with 3 days of IV methylprednisolone with gradual 

improvement. Three MR studies were performed during 

disease course. 

RESULTS 

The mean ADC value for NAWM was 840*10-3 comparing 

favorably to that for MS published data. Regions of interest 

(ROIs) were placed in the lesion core and 2 peripheral rims 

and copied to each ADC study. Values in the second study 

demonstrate an increase in ADC at the outer rim of the initial 

lesion core of 1369*10-3 (60% elevation vs NAWM p < 

0.05) and 32% increase from the initial lesion. Both studies 

demonstrated perilesional restriction of 32-42% or 575*10-6 

and 486*10-6 for the inner rim. The outer rim, initially similar 

to NAWM (792*10-6) increased in size and restriction 

on the subsequent scan (557*10-6, 33% reduction NAWM). 

The rings of enhancement corresponded to the perimeter of 

each of the regions of restriction. At 2 months follow up, the 

T2-weighted hyperintense lesion was seen to extend into the 

zone of previous restriction. 

CONCLUSION 

We show a leading edge of restriction around lesion core 

present for 4 weeks not visible on conventional imaging, into 

which the T2 hyperintense or demyelinating lesion progressively 

expanded. Restriction increased progressively and 

corresponds with an inflammatory infiltrate (and demyelination) 

supported by the biopsy findings. 

KEY WORDS: Apparent diffusion coefficient, Balo’s sclerosis, 

MR imaging


Extrahippocampal Changes in Patients with Mesial 

Temporal Lobe Epilepsy with Hippocampal Sclerosis 

Salamon, N. · Lin, J. · Lee, A. · Dutton, R. · Thompson, P. · 

Toga, A. · Engel, J. 



PURPOSE 

There is emerging evidence that patients with mesial temporal 

lobe epilepsy (MTLE) and hippocampal sclerosis have 

structural abnormalities that extend beyond the hippocampus. 

We report a quantitative volumetric MR imaging study 

of patients with pathologically confirmed hippocampal sclerosis. 

Gray and white matter changes were correlated with 

clinical variables including seizure duration, history of 

febrile seizures, and medication history. 


Quantitative volumetric analyses were performed on preoperative 

high resolution MR imaging brain scans of 15 left 

and 15 right mesial temporal lobe epilepsy 

(LMTLE/RMTLE) patients who underwent anteromesial 

temporal resection and have been seizure-free for at least 2 

years, and 20 age-matched normal controls. MR images 

were registered linearly to the International Consortium for 

Brain Mapping (ICBM) space. Images were normalized and 

classified into gray matter, white matter, and cerebrospinal 

fluid. Tissue maps for each hemisphere were parcellated into 

lobes. Lobar and whole hemisphere gray and white matter 

volumes were compared to normal controls. Regression 

analyses were performed to correlate tissue volumes with 

age, seizure duration, history of febrile seizures, and number 

of antiepileptic drugs (AED). 

RESULTS 

In LMTE patients, a 35.6% average gray matter deficit was 

found in the left and a 34.6% average deficit in the right 

hemisphere (both P < .0001). In the RMTLE patients, a 

39.4% gray matter deficit was found in the left and 40.4% in 

the right hemisphere (both P < .0001). There was also significant 

white matter loss in both MTLE groups with maximal 

deficit in the frontal and temporal lobes. No significant 

asymmetry was found for tissue volumes between the right 

and left hemispheres. Regression analysis showed that the 

age of the patient and duration of seizures were correlated 

negatively with gray matter volume ipsilateral (P < .04) and 

contralateral (P < .05) to the side of seizure onset. The number 

of AEDs negatively correlated with white matter but not 

gray matter deficits (P < .05). A history of febrile seizures did 

not correlate with tissue volumes. On region-specific correlation, 

gray matter deficits in frontal (P < .01) and parietal 

lobes (P < .05) correlated with age and seizure duration. 

When data from the affected hemispheres of both MTLE 

groups were pooled, cortical gray matter deficits correlated 

significantly with seizure duration in superior frontal 

regions. 

CONCLUSION 

Quantitative volumetric analysis of mesial temporal lobe 

epilepsy with hippocampal sclerosis showed widespread 

deficits in gray and white matter. Gray matter loss in frontal 

and parietal lobes correlated with age and duration of 

157 

seizures. This suggests MTLE may be a progressive disease 

involving multiple specific brain regions outside of the temporal 

lobe. 

KEY WORDS: Mesial temporal lobe epilepsy, voxel-based 

morphometry, neocortical volume map 


Diffusion Tensor Imaging of Mesial Temporal Lobe 

Epilepsy Patients 

Loya, A. · Salamon, N. · Lazo, C. · Selva, L. · Sinha, U. 



PURPOSE 

Mesial temporal sclerosis (MTS) is a progressive disease, 

and is the most common cause of complex partial seizures. 

Surgical treatment is curative in only 61% of patients after 

anterior temporal lobectomy, and there remains a high percentage 

of patients with residual seizures. There is evidence 

that mesial temporal sclerosis may affect extrahippocampal 

structures. Diffusion tensor imaging (DTI) has shown abnormalities 

in normal MR imaging-appearing regions in different 

disease processes, such as multiple sclerosis, thus suggesting 

a possible method to analyze the extent of cellular 

damage. Diffusion tensor imaging explores the directionality 

and the magnitude of diffusional motion of water molecules 

within tissue, which can be a measure of cellular 

integrity. The purpose of this study is to evaluate the extent 

of abnormality in patients with mesial temporal sclerosis 

using diffusion tensor imaging. 


Diffusion tensor imaging was performed on seizure patients 

who had both mesial temporal sclerosis and normal MR 

imaging findings. Diffusion tensor imaging was obtained 

using 1.5 T Siemens Sonata scanner. Axial 2D EPI diffusionweighted 

sequence was acquired with diffusion gradients 

along six noncollinear directions at three b values of 0 

sec/mm2, 600 sec/mm2, and 1200 sec/mm2, with slice thickness 

of 5 mm. Diffusion-weighted images were analyzed off 

line to generate maps of the diffusion tensor using software 

written in IDL (IDL, Research Systems, Boulder, CO). 

Diffusivity (ADC) and fractional anisotropy (FA) values 

were calculated by bilateral voxel sampling from the white 

matter in the temporal, occipital, and temporal stem white 

matter and hippocampus. Ten patients with known mesial 

temporal epilepsy with hippocampal sclerosis and 10 

patients with nonfocal epilepsy with normal MR imaging 

findings were examined. Mesial temporal sclerosis patients 

had mean age of 27 years and 60% were women. Patients 

with nonfocal MR imaging findings had a mean age of 21 

years, and 60% were male. Fractional anisotropy and ADC 

values were in the hippocampus, temporal stem, anterior 

temporal white matter, and occipital white matter. The values 

were plotted and compared to the contralateral side. A p 

value of < 0.5 will be considered statistically significant. 

RESULTS 

In MTS patients, there was increased ADC and decreased FA 

in the hippocampus, temporal lobe white matter, and temporal 

stem when compared to the nonaffected side. Increased 

ADC values with equivocal values of FA were found in the 

Wednesday

Wednesday 

occipital lobe white matter of the affected side in mesial temporal 

sclerosis patients when compared to the nonaffected 

side. 

CONCLUSION 

In MTS patient, there are extrahippocampal abnormalities 

showing high ADC and low FA values in the hippocampus 

and ipsilateral white matter along the temporal stem and 

anterior temporal lobe. Diffusion tensor imaging findings are 

seen before T2-weighted high signal abnormalities appear on 

conventional MR imaging. Diffusion tensor imaging is useful 

in evaluating extrahippocampal abnormalities in MTS 

patients. 

KEY WORDS: Epilepsy, hippocampal sclerosis, diffusion tensor 

imaging 


MR Imaging Findings in Chronic Lyme Encephalopathy 

Lignelli, A. 1 · DeLaPaz, R. L. 1 · Fallon, B. 2 · Corbera, K. 2 · 

Sackeim, H. 2 

1 Columbia University, New York Presbyterian Medical 

Center, New York, NY, 2 Columbia University, New York, 

NY 

PURPOSE 

Patients with late neuroborreliosis may exhibit encephalopathy, 

encephalomyelitis, or axonal polyneuropathy. Chronic 

lyme encephalopathy may be difficult to diagnose clinically. 

MR imaging is typically normal or shows scattered, nonspecific 

white matter hyperintensities. White matter hyperintensities 

have been reported in up to 40 % of subjects with lyme 

encephalopathy. This is the first study of lyme encephalopathy 

to evaluate T2-weighted and FLAIR hyperintensities in 

patients and compare them to age, sex, and educationmatched 

healthy controls. 


Patient group: age 18 to 65 years, lyme diagnosis based on 

CDC clinical and laboratory values, current positive IgG 

Western blot, prior treatment with at least 3 weeks of antibiotics 

and willingness to be off antibiotics for 4 weeks before 

and 24 weeks during the study. T2-weighted and FLAIR MR 

sequences were reviewed in 33 patients and 19 age-matched 

controls. Semiquantitative regional hyperintensity scores 

accounting for size, number, and anatomical distribution 

were generated by two neuroradiologists (AL and RD) using 

the Scheltens system. The neuroradiologists were blinded to 

disease status and to each others scores. 

RESULTS 

Overall scoring demonstrated abnormal hyperintensities in 

22/33 patients (67%) and 15/19 controls (79%). 

Periventricular white matter lesions were present in 8/33 

(22%) patients and 3/19 (15%) controls. Deep white matter 

lesions were present in 21/33 (63%) patients, and 11/19 

(58%) controls. Basal ganglia hyperintensities were seen in 

4/33 (12%) patients and 5/19 (26%) controls. Infratentorial 

lesions were observed in 6/33 (18%) patients and 4/19 (21%) 

controls. No statistically significant difference was seen 

between patient and control groups in total or regional 

Scheltens scores. Total Scheltens scores in patients: n = 32, 

mean = 4.13, standard deviation of 6.74 and standard error of 

158 

mean 1.19. Total Scheltens scores in controls: n = 19, mean 

= 3.32, standard deviation 4.31 and standard error of mean = 

.99. Scheltens score correlated with age, Pearsons’s r = 

0.396, p = 0.04, and with duration of illness, Pearson r = 

0.382, p = 0.05 after regression analysis, but not with severity 

fatigue scale, physical function (measured by short form 

36), number of joints involved, McGill pain scale, or length 

of antibiotic use. A trend was observed between the 

Scheltens score and both cardiovascular risk factors and total 

pain scale. 

CONCLUSION 

Contrary to expectation, lyme patients did not demonstrate a 

greater number or size of abnormal hyperintensities as compared 

to controls. Neither the presence of hyperintensities in 

any of four anatomical areas nor the total number of hyperintensities 

discriminated between lyme group and control 

group. When controlling for the effects of age and CVD 

score in a regression analysis the duration of illness did show 

a statistically significant correlation with the Scheltens 

score. Among lyme patients, longer duration of illness led to 

a greater hyperintensity burden. Although in the population 

studied MR imaging did not differentiate chronic lyme 

patients from age-matched controls, hyperintensities in 

chronic lyme may reflect long term disease related structural 

pathology. 

KEY WORDS: Lyme, white matter hyperintensities, 

encephalopathy 


Apparent Diffusion Coefficient Analysis of the 

Nonlesional Brain in Central Nervous System Vasculitis 

White, M. L. · Hadley, W. L. · Zhang, Y. 


Iowa City, IA 

PURPOSE 

Conventional MR imaging in central nervous system (CNS) 

vasculitis has been found to underestimate the extent of 

abnormalities present when compared to brain perfusion 

studies. Diffusion-weighted imaging (DWI) can demonstrate 

occult changes in the brain that would not be detected otherwise 

with conventional MR imaging studies. Diffusionweighted 

imaging has not been used to study the areas of the 

brain without lesions in patients with CNS vasculitis. 

Changes in the nonlesional brain (NLB) if present may help 

to facilitate diagnosis, better demonstrate extent of injury, 

and provide prognostic information. We studied the NLB in 

CNS vasculitis patients by DWI to evaluate for evidence of 

occult damage. 


We retrospectively reviewed MR imaging studies of patients 

diagnosed with CNS vasculitis from 1998 to present. There 

were 15 patients (average age 52 years old, 3 male, 12 

female) who had DWI during their initial presentations. 

Sixteen subjects (average age 51 years old, 5 male, 11 

female) with normal brain MR images were used as controls 

for the DWI apparent diffusion coefficient (ADC) analysis. 

All diffusion studies were performed with b=1000 s/mm 2 at 

1.5 T. Slice thickness was 5 mm. ADC maps were generated 

off-line with an ADW 4.0 GE workstation using FuncTool.

Regions of interest (ROIs) were drawn bilaterally on the 

ADC maps in the caudate heads, putamina, thalami, cerebellar 

white matter (WM), frontal WM, frontal-parietal WM, 

parietal WM, and the posterior internal capsules. The average 

size of the ROIs was 40 mm 2 . The ADCs of the frontal 

WM, the frontal-parietal WM, and the parietal WM were 

measured at three levels. Statistical analysis was the student 

t-test with Bonferroni correction. A p-value of < 0.00625 was 

taken to be significant. 

RESULTS 

The CNS vasculitis patients’ average ADCs versus the controls 

were found to be 7.7 x 10 -4 mm 2 /sec versus 7.4 x 10 -4 

mm 2 /sec for the caudate heads, 7.3 x 10 -4 mm 2 /sec versus 7.1 

x 10 -4 mm 2 /sec for the putamina, 7.4 x 10 -4 mm 2 /sec versus 

7.2 x 10 -4 mm 2 /sec for the thalami, 6.7 x 10 -4 mm 2 /sec versus 

6.4 x 10 -4 for the cerebellar WM, 8.1 x 10 -4 mm 2 /sec versus 

7.3 x 10 -4 mm 2 /sec for the frontal WM, 7.3 x 10 -4 mm 2 /sec 

versus 6.7 x 10 -4 mm 2 /sec for the frontal-parietal WM, 7.7 x 

10 -4 mm 2 /sec versus 7.1 x 10 -4 mm 2 /sec for the parietal WM, 

and 7.3 x 10 -4 mm 2 /sec versus 6.9 x 10 -4 mm 2 /sec for the posterior 

internal capsules. The vasculitis patients’ ADCs of the 

frontal WM, frontal-parietal WM, parietal WM, and the posterior 

internal capsules all were found to be significantly elevated 

(p < 0.00625). The vasculitis patients’ ADCs of the 

caudate heads, putamina, thalami, and the cerebellar WM 

were not elevated significantly (p > 0.00625). 

CONCLUSION 

Significantly elevated ADCs that could indicate occult damage 

were detected in the supratentorial NLB white matter in 

patients with CNS vasculitis. This unique information may 

provide additional data that in the future could be incorporated 

into facilitating the diagnosis and the treatment plan of 

CNS vasculitis patients. 

KEY WORDS: Central nervous system vasculitis, nonlesional 

brain, ADC 


MR Imaging Findings and Radiologic-Pathologic 

Correlation in a Case of Postvaccination 

Meningoencephalitis in Alzheimer’s Disease 

Gomez-Anson, B. M. 1,2 · Sanchez-Guerra, M. 3 · Ferrer, I. 4,5,6 · 

Boada, M. 3 

1 Hospital Clinic, Barcelona, SPAIN, 2 Institut 

d’Investigacions Biomèdiques August Pi i Sunyer, 

Barcelona, SPAIN, 3 Fundació ACE, Barcelona, SPAIN, 

4 Hospital Bellvitge, Barcelona, SPAIN, 5 University of 

Barcelona, Barcelona, SPAIN, 6 Brain Bank Hospital Clinic, 

Barcelona, SPAIN 

PURPOSE 

To describe the MR findings, and their pathologic correlates, 

in a case of asymptomatic Aβ-postvaccination meningoencephalitis 

(ME) for Alzheimer’s disease (AD). 


A 76-year-old mild/moderate AD patient (GDS = 5) enrolled 

in a multicentric, international phase II trial with Aβ-42immunization 

including 3 repeated MR imagings. Six 

months after the second injection, he experienced progres- 

159 

sive worsening of speech and gait, initially related to progression 

of AD. There was no fever, focal neurologic signs, 

or acute symptoms. On routine MR imaging (9 months after 

the second injection), a diagnosis of ME was suggested. 

After treatment with corticosteroids, and improvement, reactivation 

occurred 11 months after the second injection. The 

patient then suffered from endocarditis and aspiration pneumonia, 

the neurologic status worsened, and died 4 months 

after the reactivation episode at a nursing home. The brain 

was donated to the Brain Bank. 

RESULTS 

Bihemispheric, large, mass-like lesions were seen mainly 

affecting the white matter, and reaching the cortex, in the 

frontal and temporal regions. The lesions were hyperintense 

on DP/T2-weighted FSE MR-images (Fig. 1a), hypointense 

on T1-weighted and 3D-SPGR-MR images, and there was 

mass effect and minor midline shift (Fig. 1b). No parenchymal, 

but some pachymeningeal and leptomeningeal contrast 

enhancement were present (Fig. 1c). On diffusion-weighted 

imaging, the lesions were hypointense, having increased 

apparent diffusion coefficients (ADCs) within them (Fig. 

1d). A diagnosis of postvaccination ME was suggested. After 

treatment, repeat MR images showed improvement. 

However, clinical worsening and MR imaging reactivation 

occurred 3 months after. Post-mortem brain pathology: 

Neuropathologic study showed ME, with focal reduction of 

diffuse and neuritic plaques and inflammatory changes in the 

vicinity of collapsed plaques. These changes were more pronounced 

in the frontal and temporal regions, in keeping with 

MR findings. There was no reduction of vascular amyloid, or 

regression of tau pathology regarding neurofibrillary tangles 

and neuropil threads. 

Fig.1. MR findings. 1a) T2-weighted FSE axial images show 

large, hyperintense, bilateral fronto-temporal mass-like 

lesions affecting subcortical white matter, and reaching the 

cortex. 1b) The lesions appear hypointense on T1-weighted 

3D-SPGR MR images. The right-sided lesion is larger, and 

there is associated mass-effect and some midline shift. 1c) 

There is no parenchymal, but meningeal contrast enhancement; 

1d) There are increased ADCs in the right fronto-temporal 

lesion. 

Wednesday

Wednesday 

CONCLUSION 

This is the first case of ME to be diagnosed on MR imaging 

in an otherwise asymptomatic patient after Aβ-42-immunization 

for AD. Radiologic findings resemble those of acute 

disseminated encephalomyelitis (ADEM), and correlate to 

findings on histopathology. 

KEY WORDS: Meningoencephalitis, immunization, 

Alzheimer´s disease 


Reproducibility of MR Spectroscopy in Terms of Signalto-Noise 

Ratio: Importance for Diagnosis of Alzheimer’s 

Disease 

Okada, T. 1 · Sakamoto, S. 1 · Nakamoto, Y. 1 · Kohara, N. 2 · 

Senda, M. 1 

1Institute of Biomedical Research and Innovation, Kobe, 

JAPAN, 2Kobe City General Hospital, Kobe, JAPAN 

PURPOSE 

Several studies advocate diagnostic capability of proton MR 

spectroscopy ( 1 H-MRS) for Alzheimer’s disease (AD) and 

increased amount of myoinositol (mI) relative to creatine (Cr) is 

considered to give good criteria for diagnosis, but previous 

results of test-retest measurement of mI have shown more than 

twice variability of other metabolites, which reduced in measurement 

at a higher magnetic field. The aims of this study were 

to analyze test-retest variability of 1 H-MRS measurements in 

terms of signal-to-noise ratio (SNR). 


Ten subjects clinically suspected of mild cognitive impairment 

(MCI) or mild AD were examined twice (2-14 days apart) with 

1 H-MRS at TE of 35 ms from voxels placed at anterior and posterior 

cingulated cortex. The SNR was calculated as Cr signal 

over standard deviations of residual signals. 

RESULTS 

The percent differences between two measurements (%Diff) 

of mI/Cr showed significant linear trend to decrease as average 

SNR increased, but %Diff of N-acetylaspartate 

(NAA)/Cr and choline (Cho)/Cr did not (Fig. 1). The average 

of %Diff was 10.5 in NAA/Cr, 15.0 in Cho/Cr, and 20.8 

in mI/Cr, indicating the prominent deterioration of mI/Cr 

measurement reproducibility, that decreased to 6.96, 15.4, 

and 9.87, respectively when MRS measurements with SNR 

over 25 only were included in the analysis. Significant differences 

were observed in NAA/Cr (P = 0.024) and mI/Cr (P 

= 0.0077) between measurements with SNR higher and 

lower than 25 (Fig. 2). 

160 

CONCLUSION 

Under usual clinical setups, MRS measurements with high 

SNR should be used for reliable measurement of mI/Cr and 

more accurate diagnosis of AD. 

KEY WORDS: MR spectroscopy, test-retest reproducibility, 

Alzheimer’s disease 


MR Imaging of BK Virus Encephalitis 

Friedman, D. P. · Flanders, A. E. 

Jefferson Medical College/Thomas Jefferson University 

Hospital 


PURPOSE 

To describe dramatic MR imaging findings in a case of BK 

virus encephalitis. 


During treatment of hemorrhagic cystitis secondary to BK 

virus, a 38-year-old woman developed mental status changes 

and lethargy. Neurologic examination revealed generalized 

psychomotor slowing and dysarthria, but otherwise the 

examination was nonfocal. In 1989, the patient had been 

treated for nodular sclerosing Hodgkin’s lymphoma; in 

2002, she suffered a relapse that was treated with a stem cell 

transplant, radiation, and chemotherapy. Although her lymphoma 

went into remission, she developed multiple urologic 

problems, including CMV viremia, BK virus viremia, and 

intractable hemorrhagic cystitis. Cranial MR imaging was 

performed to evaluate her neurologic decline; multiple 

abnormalities prompted a brain biopsy. The biopsy showed 

nonspecifc reactive astrocytosis and rare perivascular lymphocytes; 

no viral inclusions were identified. Tissue polymerase 

chain reaction for BK virus revealed 8 viral DNA 

copies/1000 cells. The patient continued to receive antiviral 

agents for treatment of her BK viremia and encephalitis. Her 

neurologic status improved, although she expired 5 months 

later due to multiple medical problems. BK virus infects 80- 

90% of the population; most primary infections are asymptomatic 

or minimally symptomatic (fever, upper respiratory 

symptoms). After primary infection, a latent infection is 

established in renal epithelial cells, and possibly other tissues 

(including the brain). Significant clinical manifestations 

are almost always limited to patients with impaired immune 

function, especially renal and bone marrow transplant 

patients. BK virus is associated most often with urologic disease, 

including hemorrhagic cystitis and tubulo-interstitial 

nephritis. The spectrum of clinical and imaging manifestations 

in BK virus encephalitis is not known; only a few cases 

(diagnosed by PCR assay of CSF) have been reported. In our 

patient, dramatic imaging abnormalities were present, 

despite only relatively mild clinical symptoms. 

RESULTS 

MR imaging demonstrated widespread increased signal on 

T2- and FLAIR-weighted images, along with restricted diffusion, 

in the cerebellar cortex and cerebral hemispheric 

white matter. Similar findings were present in the bilateral 

globus pallidi, and, to a lesser extent, the thalami. The cerebellar 

white matter and cerebral cortex appeared spared. The

cerebellum was diffusely swollen. There was no abnormal 

enhancement or hemorrhage. The cause for the distribution 

of these abnormalities is unknown. 

CONCLUSION 

Although rare, BK virus can cause encephalitis in immunocompromised 

patients. The simultaneous presence of urologic 

abnormalities (especially hemorrhagic cystitis or 

nephritis), and nonfocal neurologic deficits, especially in 

renal and bone marrow transplant patients, should prompt 

consideration of BK virus encephalitis. 

KEY WORDS: BK virus, encephalitis, MR imaging 

Dr. Flanders will present the paper. 


Quantitative Serial MR Imaging: Prospective Evaluation 

of Solitary Cerebral Cysticercosis 

Nalini, A. · Yeshraj, G. · Sivakumar, D. · Thennarasu, K. · 

Jayakumar, P. N. 

National Institute of Mental Health & Neuro Sciences 

Bangalore, INDIA 

PURPOSE 

This prospective randomized trial was to study 1) the evolution 

of single cysticercal cyst (NC), 2) changes noted in adjacent 

parenchyma in patients on a randomized albendazole 

(Alb) trial using MR quantitative parameters like magnetization 

transfer (MT) and T2 relaxometry (T2R) and 3) to study 

the changes in perilesional parenchyma (PLP) and its correlation 

with patient-dependent parameters of age, sex, MT 

and T2R values of normal brain and effect of Alb drug on 

161 

these. Single cysticercal cyst is thought to be modified by 

cysticidal drugs. There are conflicting reports citing the role 

of Alb on evolution of the cyst and appearance of adjacent 

perilesional gliosis. 


MR imaging was done on 1.5 T machine. There were 81 

cases with solitary NC randomized for treatment with Alb. 

All had MR imaging at 0, 3, and 6 months of follow up and 

serial MR image was done in 17 patients. All underwent 

TSE, T2-weighted, SE T-weighted with and without MT. 

Following independent parameters were analyzed: stage of 

the lesion, MT and T2 relaxation values from the cyst, PLP 

and opposite normal white matter (WM), and the mean values 

were calculated. These parameters were analysed with 

follow-up MR imaging at 3 and 6 months, and distribution of 

variables among treatment groups was done by multivariate 

analysis to look for differences among treatment groups in 

evolution of lesion and PLP changes. 

RESULTS 

Man:woman ratio was 46:35, age range was 8-50 years. The 

first MR imaging revealed 27 lesions were of colloid vesicular 

(CV) type, 52 granulomatous (Gr), and 2 nodular (No). 

T2 relaxometry values of cyst were 264.7 ± 214.6, PLP 

350.2 ± 141.7, and opposite normal WM was 123.7 ± 11.9 

which were significantly different. The T2R values differed 

significantly between CV (321.2 ± 179.6) Gr (389.3 ± 

674.7), and No (84.0 ± 19.7) and on follow-up MR imaging 

at 3 and 6 months the T2R values of the lesion(162.500 ± 8.5 

and 136.2 ± 36.0) and PLP (160.1 ± 55.4 and 159.4 ± 66.1) 

decreased progressively, but the T2R values were more than 

opposite normal white matter (123.7 ± 11.9). Magnetization 

transfer values of cyst (9.7 ± 6.0) and PLP (16.1 ± 3.9), were 

significantly less than MT values of opposite side normal 

WM (21.8 ± 3.9). The MT values differed significantly 

between CV (7.7 ± 10.7), Gr (7.9 ± 10.6), and No (16.7 ± 

4.9) stages. Further at 3 and 6 months follow-up MR imaging, 

MT values of cyst (11.9 ± 7.0 and 11.9 ± 5.2) and adjacent 

region (16.2 ± 5.8 and 19.5 ± 15.4) increased progressively, 

but the MT values were below the opposite normal 

WM(21.8 ± 3.9).The changes in MT and T2R values did not 

show any significant differences between patient groups, 

age, and sex. 

CONCLUSION 

MR morphometric parameters, T2R and MT values of NC 

and PLP show significant differences compared to normal 

parenchyma. These values were significantly different, at 

various stages of the lesion and progressively changed with 

evolution of the cyst in course of time . However, there were 

no significant differences in the evolution of the cyst and 

associated perilesional changes between patients who 

received Alb and who did not. Gender and age also did not 

have any significance in the findings. 

KEY WORDS: Neurocysticercosis, magnetization transfer 

imaging 

Wednesday

Wednesday 


Comparison of Gadolinium-Enhanced T1 and FLAIR 

Sequences in Patients with Acute Meningitis 

Ponzo, J. A. · Simonetta, A. · Calderwood, G. 

Scott and White Clinic 

Temple, TX 

PURPOSE 

Cranial MR imaging currently is utilized in the evaluation of 

patients with suspected meningitis and significant findings 

have been described on both enhanced and unenhanced MR 

imaging sequences, particularly FLAIR sequences. The purpose 

of this study was to analyze the sensitivity and specificity 

of gadolinium-enhanced sequences and FLAIR 

sequences independently in the clinical setting of meningitis. 


A retrospective blinded evaluation of 14 patients from over a 

2-year period with a documented clinical diagnosis of 

meningitis, documented by cerebrospinal fluid analysis, who 

also have been studied with cranial MR imaging was performed. 

Enhanced MR imaging and FLAIR sequences were 

reviewed separately in a blinded fashion by 2 neuroradiologists 

as to the presence or absence of abnormality such as 

hyperintensity (enhancement), mass, presence of debris or 

architectual distortion of the brain or meninges. Typical 

small vessell ischemic changes in the white matter were not 

reported as a positive finding on FLAIR. A control group of 

random patients during the same period who have both 

enhanced and nonenhanced cranial MR imaging with an 

official normal interpretation also were reviewed by the neuroradiologists. 

RESULTS 

Sensitivity/specificity for abnormal findings suggestive of 

intracranial infection was 50/100% for FLAIR sequences 

and 42.8/100% for gadolinium-enhanced MR imaging 

seqences. Differences were not statistically significant. 

CONCLUSION 

FLAIR imaging in the clinical scenario of meningitis is at 

least as sensitive and specific for abnormalities attributable 

to intracranial infection as gadolinium-enhanced MR imaging. 

Findings suggest that gadoliniuim administration is not 

warranted for initial evaluation of patients with suspected 

meningitis. 

KEY WORDS: FLAIR, meningitis, cost effectiveness 

162 


Reversible Progressive Multifocal Leukoencephalopathy: A 

Case Report 

Talmi, D. · Crowder, C. · Drachenberg, C. B. · Ramos, E. · 

Morales, R. E. 

University of Maryland Medical Center 

Baltimore, MD 

PURPOSE 

To present radiologic correlation of a case of reversible progressive 

multifocal leukoencephalopathy (PML) in the setting 

of organ transplantation. Progressive multifocal 

leukoencephalopathy is a rare, progressive, and ultimately 

fatal demyelinating neurologic disorder of the brain that is 

associated with polyomavirus infection. BK and JC viruses 

(BKV and JCV) are human polyomavirus (HPV) which are 

members of the papovavirus family. After primary infection, 

there may be reactivation in situations of immune impairment. 

Immunocompromised patients such as those with 

chronic lymphocytic leukemia, AIDS and following organ 

transplantation are at increased risk. Since the prognosis of 

patients with PML is poor, reversible cases commonly have 

not been reported. 


A forty-seven-year old man with end stage renal disease secondary 

to Alport syndrome developed PML following a second 

living related donor renal transplantation. Tacrolimus 

was one of the drugs included in the patient’s immunosuppressive 

regime following the transplant. The disease was 

manifested by focal seizure, low grade fever, myalgias and 

moderate cognitive changes 5 months after transplantation. 

MR imaging showed cerebral foci of demyelination and 

gliosis. Abnormal increased signal within the subcortical 

white matter of the right parietal lobe was noted near the vertex, 

and also within the deep white matter of both cerebral 

hemispheres and the lateral right basal ganglia. Stealth MR 

imaging-guided stereotactic brain biopsy was performed and 

sections from the brain biopsy specimen demonstrated a 

macrophage-rich lesion with an associated perivascular 

chronic inflammatory infiltrate. Numerous enlarged oligodendrial 

cells were present, some of which contained 

ground-glass-like viral inclusions and marginated chromatin. 

Several of these nuclei were positive when stained with antibodies 

directed against SV40 virus, confirming the diagnosis 

of PML. Cerebrospinal fluid was not available for evaluation. 

After the diagnosis of PML all immunosuppression was 

discontinued. The neurologic symptoms gradually improved 

and finally resolved over a period of 4 weeks. Renal function 

slowly deteriorated and the patient returned to hemodialysis 

3 months after the diagnosis of PML. After 43 months the 

patient continues to be well and has no residual neurologic 

deficiencies or symptoms. Radiologic resolution of the 

lesions with very minimal residual signal abnormality and no 

evidence for enhancement was observed in follow-up imaging 

studies. 

RESULTS 

Initial MR imaging showed foci of demyelination. Abnormal 

increased signal within the subcortical white matter of the 

right parietal lobe was noted near the vertex, and also within 

the deep white matter of both cerebral hemispheres and 

the lateral right basal ganglia. Follow-up MR imaging

approximately 2.5 years later demonstrated resolution of the 

lesion with very minimal residual signal abnormality and no 

evidence of enhancement. 

CONCLUSION 

Progressive multifocal leukoencephalopathy, a rare subacute 

demyelinating disease usually seen in the immunocompromised 

patient due to infection with the JC virus (papovavirus). 

This disease has a poor prognosis, at least in part 

due to the inherent nature of the patient’s immunocompromised 

status. We present a transplant patient with imaging 

findings suggestive of PML and pathologic confirmation, 

that had resolution of this disease when his immunosuppression 

was discontinued. 

KEY WORDS: Demyelination, transplant 


3:00 PM - 4:33 PM 

Room 107 

(51c) SPINE: Functional and 

Degenerative 




Miscellaneous 





Moderators: Victor M. Haughton, MD 



MR Imaging of the Spinal Cord at 3 T: Evaluation of T2 

FLAIR and Diffusion-Weighted Imaging 

Ririe, D. · Nesbit, G. · Wang, P. · Anderson, J. · Hamilton, B. 


Portland, OR 

PURPOSE 

High-field (3 T) MR imaging is a powerful new clinical tool, 

and one that presents many technical challenges when used 

for imaging of the spine (1-2). MR imaging of the spinal 

cord has been limited due to its small size and the lack of 

useful adjunctive imaging such as fluid-attenuated inversion 

recovery (FLAIR) and diffusion-weighted imaging (DWI). 

This paper summarizes our initial experience of 3 T imaging 

of the spinal cord in patients with clinical myelopathy with 

attention to T2 FLAIR and DWI imaging. 

163 


Over 3000 spine examinations have been performed using 

Philips Intera 3.0 T and a phased array receive spine coil. We 

recently have incorporated T2 FLAIR (11000/100/TI 2600, 

ETL 8), DWI using single and multishot techniques (3000 or 

gated 2 beats/ 84-89/160 2 matrix, b = 600) sequences into 

spine imaging at 3 T. Patients with clinical suspicion for 

myelopathy that had imaging at 3 T using DWI and T2 

FLAIR were included in this study. The studies were evaluated 

by three neuroradiologists blinded to the clinical history 

and diagnosis. Evaluation included the routine 3 T images 

(T1, T2 TSE, gadolinium) without FLAIR and DWI and then 

including these sequences sequentially to determine the utility 

of the added sequences. 

RESULTS 

Twenty-one patients with myelopathy having 3 T imaging 

and FLAIR and/or DWI have, to date, been imaged. Eight 

patients had normal MR imaging examinations, 13 had 

lesions on MR imaging. Diagnoses included: multiple sclerosis, 

syrinx, contusion, myelomalacia, infarction, dural 

AVM, and cords tumor. Diffusion-weighted imaging altered 

the impression in 10%, supported the impression in 31%, 

and found to not be useful in 59%, either due to the diagnosis 

or the motion artifact. FLAIR altered the impression in 

16%, supported the impression in 60%, and found to not be 

useful in 24%. Restricted diffusion was demonstrated in an 

acute multiple sclerosis lesion and acute spinal cord 

ischemia. FLAIR has shown good differentiation of cystic 

versus edematous changes in the spinal cord. 

CONCLUSION 

This initial experience suggests FLAIR and DWI at 3 T provides 

useful diagnostic information beyond standard T1 and 

T2 imaging in select patients with myelopathy. Further study 

is needed to determine the specific utility of these techniques 

in the diagnosis of demyelinating disease, ischemia, and 

trauma. Further optimization of these techniques, especially 

DWI, also may provide improved imaging. 

REFERENCES 

1. Keiper MD, Grossman RI, Brunson JC, Schnall MD. Diffusion 

imaging of the human spinal cord and the vertebral column. Top 

Magn Reson Imag 2003;14(6):461-476 

2. JR Korzan, M Gorassini, D Emery, ZA Taher, C Beaulieu. In 

Vivo magnetic resonance imaging of the human cervical spinal 

cord at 3.0 Tesla. JMRI 2002;16:21-27 

KEY WORDS: spinal cord, diffusion-weighted imaging, 

FLAIR 


Rapidly Growing Giant Bone Island of the Thoracic 

Spine: Imaging and Histologic Features 

Hayek, R. A. 1 · Riley, G. T. 2 · Abdu, W. A. 1 · Mamourian, A. C. 1 

1 Dartmouth-Hitchcock Medical Center, Lebanon, NH, 

2 Walter Reed Army Medical Center, Washington, DC 

PURPOSE 

To review the radiographic, nuclear medicine bone scan, 

MR, and CT features of a biopsy proven giant thoracic spine 

enostosis in a 35-year-old female with severe back pain. This 

Wednesday

Wednesday 

case was unusual because of the comprehensive temporal 

imaging evaluation and documentation of growth over a 5year 

period. 


The patient was 35-year-old female referred to our Spine 

Center for evaluation of back pain, weight loss, fatigue, and 

an incidentally noted sclerotic T11 lesion on CXR. CXR 

obtained in August 2001 demonstrated a 2 cm sclerotic 

lesion under the superior endplate of T11, not present on 

CXR 4 years prior. The lesion was thought to be a bone 

island (enostosis) on the basis of the imaging features. A 

repeat CXR 2 years later (2003) demonstrated interval 

enlargement of the lesion. Due to the patient’s persistent 

back pain and constitutional symptoms, further imaging 

evaluation of the lesion including CT, bone scan, and 

gadolinium-enhanced MR imaging, was obtained. As no 

other findings could explain the patient’s persistent back 

pain and constitutional symptoms and the lesion had demonstrated 

remarkable growth over time, a CT-guided biopsy of 

the T11 lesion was performed. 

RESULTS 

On plain film radiologic and CT examination the lesion was 

sclerotic with a narrow zone of transition. Enostosis is 

described classically as having a spiculated appearance 

“thorny radiation” which represents the transition from the 

lamellar bone to normal trabeculated bone as demonstrated 

in this case (Fig.1). Bone scan demonstrated no radiopharmaceutical 

uptake, a typical finding for an enostosis. On MR 

examination the lesion was nonenhancing (Fig. 2), and uniformly 

low signal on T1- and T2-weighted imaging consistent 

with compact lamellar bone. Histology of the biopsy 

sample demonstrated compact lamellar bone with haversian 

canals consistent with an enostosis. 

CONCLUSION 

Enostosis or bone island, was first described in 1905 by Dr. 

Stieda. These benign lesions generally are more evident in 

adults and are most common in the axial skeleton. In the vertebral 

bodies the lesion usually lies just beneath the cortex, 

and are seen usually in the thoracic or lumbar spine. The 

lesions are thought to be developmental in nature and their 

natural history is variable. Lesions are generally stable in 

size but both regression and progression have been described 

in approximately 1/3 of cases. We have demonstrated the 

classic imaging and histologic findings in this unusual case 

of an enlarging giant enostosis of thoracic spine. 

KEY WORDS: Bone island, enostosis, thoracic spine 

164 


T1 FLAIR versus T1 FSE: The Optimal T1 Pulse 

Sequence for Clinical Spine Imaging at 3 T 

Shapiro, M. D. · Ramnath, R. · Williams, D. · Magee, T. · 

Hartker, R. · Wasudev, N. · Lees, J. 

Neuroskeletal Imaging of Winter Park 

Winter Park, FL 

PURPOSE 

Limited clinical study at 1.5 T has demonstrated that T1 

FLAIR is superior to T1 FSE in detecting bone marrow and 

soft tissue pathology and in differentiating normal interfaces 

between intervertebal disk, cerebrospinal fluid (CSF), and 

spinal cord for spine imaging. At 3 T, T1 relaxation times are 

prolonged compared to 1.5 T which has a deleterious effect 

on distinguishing both between these normal interfaces and 

in differentiating normal soft tissues (including less dark 

CSF) and normal bone marrow from abnormal soft tissue 

and abnormal marrow. 


A prospective, double-blinded study of 100 patient’s spine 

MR images (50 cervical, 50 lumbar) without and with contrast 

was obtained with both T1 FSE and T1 FLAIR on a GE 

3 T Excite MR imaging unit for the purpose of determining 

the better T1 pulse sequence for spine imaging at the higher 

field strength. The following parameters were evaluated: 1) 

normal CSF-cord, cauda equina interfaces; 2) normal disk- 

CSF interface; 3) differentiation of normal-abnormal 

intramedullary signal; 4) differentiation between normal and 

abnormal bone marrow signal; 5) conspicuity of contrast 

enhancement in abnormal soft tissue and bone marrow. 

RESULTS 

T1 FLAIR was clearly superior in the 3 categories, minimally 

better in one and equal to FSE in conspicuity of contrast 

enhancement. 

CONCLUSION 

T1 FLAIR is a better T1-weighted pulse sequence for spine 

imaging at 3 T than T1 FSE. 

KEY WORDS: T1 FLAIR, spine, 3 T 


Consequence of the Misuse of the Term “Annular Tear” 

in Patients Involved in Litigation 

Rothman, S. L. G. 


Los Angeles 


PURPOSE 

The use of the phrase “annular tear” to describe a high intensity 

zone in the posterior annulus seems quite innocuous. 

The ASSR and NASS position paper on definitions strongly 

suggest that the term should not be used. They feared that 

this phrase would be misconstrued to suggest acute trauma to 

the disk. They were correct. In the last 6 months alone, more 

than 20 patients with high intensity zones in the annulus who

had been involved in accidents claimed that their “annular 

tears” were the direct consequence of trauma and therefore 

the cause of the patients posttraumatic low back pain. Both 

of these allegations are insupportable on MR scans and 

should never be made. 


The literature regarding “annular tear” and “high intensity 

zone” is reviewed and summarized. MR scans were 

reviewed on patients claiming annular tear due to trauma to 

determine if there were any criteria to suggest acute disk 

trauma. 

RESULTS 

In none of the reviewed cases was there evidence of acute 

anatomical injury to the intervertebral disk. 

CONCLUSION 

The term “annular tear” should never be used in an MR 

report. The inadvertent or intentional misinterpretation of the 

MR reports causes great harm in the medico-legal arena. 

KEY WORDS: Spine, annular tear 


Are there MR Imaging Features that Predict Regression 

of Lumbar Disk Herniation? 

Erly, W. K. 1 · Munoz, D. 2 · Beaton, R. 1 

1 The University of Arizona, Tucson, AZ, 2 Inland Imaging, 

Spokane, WA 

PURPOSE 

To assess whether MR imaging characteristics of lumbar 

disk herniations can predict subsequent disk regression. 


Medical and radiology records from 1999-2003 of the 

Tucson VA Medical Center were reviewed and 123 patients 

were identified who had more than one lumbar MR imaging 

during the study interval. Of these, 42 patients had at least 

one disk herniation (protrusion or extrusion) identified on 

their first examinaion. Six of these patients were not included 

because of prior lumbar surgery, or inadequate examinations. 

The remaining 36 patients had a total of 7 examinations 

to evaluate 44 disk herniations. The herniated disks 

were evaluated by two CAQ neuroradiologists for size, morphology, 

and a qualitative assessment of the T2 signal. 

RESULTS 

Twenty-five of 45 (57%) herniated disks decreased in size, 

17 (38%) were unchanged, and two increased between the 

first and last examination. Of the disks that decreased in size, 

15 (63%) had an area of increased signal on T2-weighted 

imaging compared to the parent disk on the initial study. Of 

the disks that were unchanged, 6 (35%) had increased signal 

on the T2-weighted images. The mean size of the disks that 

regressed was significantly larger than those that were 

unchanged (8.6 mm vs . 6 mm, p = .001). On average, the 

disks decreased 3.2 mm (37%). The mean age of the patients 

who improved was 56 years, and of those that showed no 

165 

improvement was 63 years (p > 0.05). Nine of 11 (82%) 

extruded disks improved, while 16 of 34 (47%) protrusions 

improved. 

CONCLUSION 

Fifty-six percent of herniated disks in this study group 

decreased in size over time. Larger herniations and extrusions 

were more likely to regress than smaller herniations or 

protrusions. Disks that regressed were more likely to have 

some high signal on T2-weighted images than those that 

were stable. 

KEY WORDS: MR imaging, lumbar spine, herniation 

This paper was selected by the Western Neuroradiology 

Society (WNRS) to receive the Gabriel H. Wilson Award for 

the best paper presented at its 36th Annual Meeting held 

September 30 - October 2, 2004. 


Can Diffusion Tensor Imaging Predict Spinal Cord 

Function in Patients with Cervical Spondylosis? 

Hesseltine, S. · Law, M. · Johnson, G. 


New York, NY 

PURPOSE 

Neurosurgical triage of patients with cervical spondylosis is 

highly dependent on demonstrating abnormal spinal cord 

signal at conventional MR imaging; however, changes in 

spinal cord function and clinical symptomatology precede 

findings of abnormal cord signal. The purpose of this study 

is to spatially compare diffusion tensor metrics of white matter 

tracts in the cervical spinal cord in normal volunteers 

with metrics in patients with varying degrees of cervical 

spondylosis. 


Diffusion tensor imaging was performed in 41 patients with 

varying degrees of spondylosis and 9 healthy volunteers, 

using a pulsed gradient, double spin-echo, echo-planar imaging 

(2000/74; 128 x 128 matrix; 140 x 140 mm FOV; 10 contiguous 

4 mm slices; b = 1000 s/mm2) at 1.5 T. Using a 

method described by Singh et al (1), a score corresponding 

to overall severity of cervical spinal canal stenosis was calculated 

for each patient. We evaluated the DTI characteristics 

of the spinal cord at the C2-3 intervertebral level. At this 

level, fractional anisotropy (FA) and mean diffusivity (MD) 

were calculated within regions of interest at the anterior, lateral, 

and posterior regions of the spinal cord, with separate 

bilateral regions of interest at each of these positions. 

RESULTS 

The spondylosis score for our 41 patients averaged 2.0 ± 1.8 

(average ± standard deviation). The C2-3 level was remote 

from the most superior level of pathology in all of our 

patients, being 3.2 ± 1.8 cervical levels cephalad on average. 

The average age of our patients was 60.0 ± 18.6 (average ± 

standard deviation) years vs 34.3 ± 14.5 years for the control 

group (p < 0.0001); however, there was no significant correlation 

of measured FA or MD with age (R = -0.06 and 0.03, 

respectively). The MD values in the anterior spinal cord 

Wednesday

Wednesday 

were significantly higher in the patients with spondylosis 

compared to controls (0.992 ± 0.234 vs 0.840 ± 0.154, p = 

0.002). There was no significant difference in MD in the lateral 

or posterior regions of the spinal cord, nor was there a 

significant difference in FA in any region of the spinal cord 

comparing disease group to controls (see Table). The anterior 

FA values were significantly lower than the posterior FA 

values in both the patients with spondylosis (0.493 ± 0.079 

vs 0.618 ± 0.102, p < 0.0001) and controls (0.486 ± 0.094 vs 

0.635 ± 0.051, p < 0.0001). 

CONCLUSION 

The study suggests that changes in MD are present in the 

anterior spinal cord, even remote from the site of spondylosis. 

Regional differences in FA in the cervical spinal cord 

also are demonstrated, with significantly lower FA in the 

anterior compartment. This may relate to either increased 

mechanical stress in the ventral spinal cord in spondylosis or 

relative increased gray matter in the anterior regions of interest. 

These findings may be valuable in future investigations 

of DTI characteristics at the site of pathology in patients with 

spondylosis as well as help in the neurosurgical triage of 

patients with degenerative cervical spine disease. 

REFERENCES 

1. J Neurosurg-Spine 2001;94:189-198 

KEY WORDS: Degenerative, diffusion tensor imaging, spine 


Changes in Cervical Spinal Cord Anisotropy in Multiple 

Sclerosis Measured by Diffusion Tensor Imaging 

Hesseltine, S. · Law, M. · Rad, M. · Ge, Y. · Johnson, G. · 

Grossman, R. 


New York, NY 

PURPOSE 

Primary demyelination continues to pose a diagnostic challenge 

in clinical practice; findings on conventional MR 

imaging can sometimes be nonspecific. Diffusion tensor 

imaging (DTI) and fiber tractography have been shown to 

provide physiologic information regarding cerebral white 

matter disorganization, which precedes abnormalities seen 

on conventional imaging. Changes in DTI metrics have been 

demonstrated within focal spinal cord multiple sclerosis 

lesions. The purpose of this study is to compare diffusion 

tensor metrics spatially within both the gray matter and 

white matter tracts of the cervical spinal cord in normal volunteers 

and in patients with multiple sclerosis, in regions of 

the cord that appear normal on conventional MR imaging. 


Diffusion tensor imaging of the upper cervical spine was 

performed in 13 patients with relapsing-remitting multiple 

sclerosis and 9 healthy volunteers, using a pulsed-gradient, 

double spin-echo, echo-planar imaging (2000/74; 128 x 128 

matrix; 140 x 140 mm FOV; 10 contiguous 4 mm slices; b = 

1000 s/mm2) at 1.5 T. At the C2-3 level, fractional 

anisotropy (FA) and mean diffusivity (MD) were calculated 

within regions of interest at the anterior, lateral, and posterior 

regions of the spinal cord, with separate bilateral regions 

of interest at each of these positions. 

166 

RESULTS 

The average age of the multiple sclerosis patients was 43.0 ± 

7.0 years (average ± standard deviation) vs 34.3 ± 14.5 years 

in the control group (p = 0.13). Significantly decreased fractional 

anisotropy was present in the multiple sclerosis 

patients at the lateral and posterior regions of the spinal cord 

(see Table), with no significant difference detected at the 

anterior regions of interest. Although there was no significant 

difference in mean diffusivity between multiple sclerosis 

patients and controls, there was a trend towards slightly 

increased mean diffusivity in the diseased patients. 

DTI metrics at the C2-C3 spinal level in patients with multiple sclerosis 

and normal volunteers 

Fractional Anisotropy 

Anterior Anterior Lateral Lateral Posterior Posterior 

Left Right Left Right Left Right 

Cases 0.52±0.11 0.52±0.13 0.45±0.07 0.45±0.07 0.54±0.09 0.52±0.11 

(n-13) 

Controls 0.49±0.10 0.49±0.10 0.59±0.11 0.62±0.09 0.63±0.05 0.64±0.06 

(n=9) 

P-values 0.42 0.54 0.0007 < 0.001 0.01 0.004 

Mean Diffusivity (x 10 -3 mm -2 s -1 ) 

Anterior Anterior Lateral Lateral Posterior Posterior 

Left Right Left Right Left Right 

Cases 0.91±0.25 0.99±0.36 0.92±0.24 0.92±0.29 0.90 ±0.33 0.94±0.33 

Controls 0.82±0.15 0.86±0.17 0.89±0.13 0.92±0.18 0.86±0.18 0.84±0.21 

P-values 0.27 0.30 0.74 0.99 0.74 0.37 

CONCLUSION 

The study demonstrates a significantly decreased fractional 

anisotropy in the lateral and posterior tracts in the cervical 

spinal cord of multiple sclerosis patients, in the absence of 

spinal cord signal abnormality at conventional MR examination. 

These findings may be clinically more sensitive than 

conventional MR imaging in differentiating primary from 

secondary causes of demyelination. 


spinal cord 


Cord Compression: Which MR Imaging Sequences Are 

Necessary? 

Johnson, A. J. 1 · Ying, J. 1 · El Gammal, T. 2 · Kim, R. Y. 2 · 

Timmerman, R. D. 1 · Littenberg, B. 3 

1 Indiana University, Indianapolis, IN, 2 University of 

Alabama at Birmingham, Birmingham, AL, 3 University of 

Vermont, Burlington, VT 

PURPOSE 

To determine which MR sequences are necessary to evaluate 

the spinal axis for possible metastasis in patients with primary 

non-CNS malignancy and clinical suspicion of spinal 

metastasis. 


This was an observational case series, where cases were 

acquired retrospectively but hypothetical MR interpretations 

and management plans were made prospectively. Subjects 

were consecutive adult patients from one large tertiary care 

medical center with clinical suspicion of metastatic disease 

to the bony spinal axis, for whom spine MR imaging had 

been performed. For the study, standardized protocols of the 

MR scans were prospectively independently interpreted by 

two neuroradiologists. MR protocol types varied: 1) T1weighted 

images only, 2) T1-weighted and T2-weighted 

images, 3) T1-weighted and postcontrast T1-weighted

images, and 4) T1-weighted, T2-weighted, and postcontrast 

T1-weighted images. Based on history and physical findings 

at presentation and the neuroradiologists’ report, two radiation 

oncologists were asked independently to provide a 

hypothetical management plan for each patient. The logit 

model was used to investigate the effects of MR protocol 

type on the probability of planning radiation therapy (RT). 

Pearson correlation coefficients were used to assess the concordance 

of median spinal levels suggested for RT by protocol 

type. A mixed model was used to compare the total number 

of levels for which RT was planned by protocol type. 

RESULTS 

A total of 31 subjects were evaluated, with multiple scan 

interpretations by protocol type, neuroradiologist, and oncologist 

for each subject. The logit model showed that MR protocol 

type did not have a significant effect on the probability 

of the oncologist recommending RT (all p > 0.39). 

Pearson correlation coefficients for median levels of planned 

RT: protocol 1 vs 4 = 0.88; protocol 2 vs 4 = 0.93; protocol 

3 vs 4 = 0.87. The number of levels for which RT was 

planned did not vary significantly by MR protocol type (all 

p > 0.05). 

CONCLUSION 

While MR imaging is known to be the most useful diagnostic 

test in cases of suspected spinal cord compression, it is 

unclear which particular MR sequences are necessary in this 

clinical scenario. Compared to T1-weighted images alone, 

the additional use of T2-weighted and/or postcontrast T1weighted 

sequences did not significantly affect the probability 

that RT would be suggested or what levels would be chosen 

for RT in our study. Our data suggest that unenhanced 

T1-weighted images may be sufficient for evaluation of possible 

cord compression. 

KEY WORDS: Spine MR imaging, technology assessment, 

cord compression 


Subependymomas of the Spinal Cord: MR Imaging and 

the “Ribbon Sign” 

Silvera, V. M. 1 · Jallo, G. I. 2 · Lefton, D. R. 3 · Pinto, R. 3 

1 Children’s Hospital, Boston, MA, 2 The Johns Hopkins 

University, Baltimore, MD, 3 Roosevelt Hospital, New York, 

NY 

PURPOSE 

Subependymomas are nonaggressive tumors, which are most 

often clinically silent and incidentally encountered within 

the ventricles at autopsy. Rarely, these tumors occur within 

the spinal cord and most commonly these tumors are presumed 

preoperatively to represent astrocytomas. We present 

the MR imaging appearance of 8 cases of spinal cord 

subependymomas, 5 of which demonstrated a distinctive 

“ribbon-like” appearance of the affected spinal cord. In our 

experience, this is an imaging appearance unique to this type 

of spinal cord tumor. 

167 


We reviewed the MR studies of eight patients with spinal 

cord subependymomas and correlated them with the clinical 

data, histopathology, and operative reports. 

RESULTS 

Eight patients, 3 women and 5 men presented between the 

ages of 20-51 years with spinal cord subependymomas. 

Three tumors were located within the cervical and 5 tumors 

were located within the thoracic cord. All tumors were 

sharply marginated. Two tumors demonstrated faint 

enhancement and 1 demonstrated equivocal enhancement. 

No patients had associated spinal cord edema or syrinx formation. 

The length of the tumors varied from 1-8 vertebral 

body segments. Three tumors had associated bony remodeling 

of the spinal canal. Five tumors demonstrated a distinctive 

radiologic appearance with a sharp, wavy margination 

of the spinal cord/tumor interface, which we have coined “ 

the ribbon sign.” The “ribbon” represents the compressed 

normal spinal cord. In our review of the literature and our 

clinical database, which consists of 383 spinal cord tumors, 

this imaging appearance is unique to subependymomas of 

the spinal cord. 

CONCLUSION 

Subependymomas of the spinal cord are rare tumors, which 

may present with a distinct imaging appearance that we have 

termed “ the ribbon sign.” Appreciation of this unique 

appearance should allow for a more accurate preoperative 

assessment of this entity. 

KEY WORDS: Spinal cord, neoplasm 

Wednesday

Wednesday 


Fetal Spine Development Assessed by Postmortem MR 

Imaging 

Widjaja, E. 1 · Whitby, E. 2 · Griffiths, P. D. 2 



PURPOSE 

There is an increasing interest in utilizing postmortem MR 

imaging as an alternative to autopsy. Prior to evaluating any 

spinal pathology on postmortem MR imaging, knowledge of 

the normal appearance of the spine at different gestational 

age is crucial. The aim of this study is to evaluate the normal 

development of fetal spine with postmortem MR imaging. 


Postmortem MR imaging was performed on 30 fetuses from 

14 to 41 weeks gestation (mean gestation 24.6 weeks). There 

was no structural abnormality of the spine in these fetuses 

either on MR imaging or at autopsy. Fast spin-echo T2weighted 

MR imaging of the lumbar spine was performed in 

the coronal plane in all cases and sagittal and axial plane in 

some cases. The following parameters were measured: 

height of the L1/2 disk and L2 vertebral body, area of ossification 

center in L2 vertebral body as well as the overall size 

of the vertebral body. The signal of the disk space and the 

vertebral body was assessed also. 

RESULTS 

The height and area of vertebral body increases with gestational 

age (p < 0.01). The increase in disk space is greater 

than the increase in vertebral body height as the gestation 

increases (p < 0.01). The disk space appears as a linear low 

signal area in some fetus up to 21 weeks gestation. High signal 

is seen in the disk of all fetuses beyond 21 weeks. The 

ossification center of the vertebral body increases with gestational 

age (p < 0.01) and the ratio of ossification center of 

the vertebral body to the overall size of the vertebral body 

also increases with gestational age (p < 0.01). The ossification 

center of the posterior elements is first seen medially at 

the base of the lamina. Ossification of the posterior elements 

proceed posteromedially, anteriorly, and laterally as the fetus 

grows. 

CONCLUSION 

Understanding the normal growth and signal characteristics 

of the fetal spine on post-mortem MR imaging is essential 

prior to studying abnormal fetal spine. 

KEY WORDS: Fetal, spine, development 


Postmortem MR Imaging of Fetal Spinal Malformations 

Widjaja, E. 1 · Whitby, E. 2 · Griffiths, P. D. 2 



PURPOSE 

The purpose of this study was to determine the postmortem 

MR appearances of spinal malformations in fetuses and to 

compare this with the autopsy findings. 

168 


Nine women who had fetuses with developmental spinal 

anomalies had termination of pregnancy. Postmortem MR 

imaging was performed on 1.5 T clinical scanner, using 

either the wrist or knee coil. Brain and spine imaging was 

performed in all fetuses. MR imaging of the brain consisted 

of fast spin-echo (FSE) T2 sequence in three orthogonal 

planes (TR = 11460 msec, TE = 92 msec, turbo factor = 32, 

FOV = 11cm, slice thickness = 2 mm, matrix = 256 x 256, 

scan time = 6 min for each plane) and the spine imaging consisted 

of sagittal and axial FSE T2 sequence (TR = 11460 

msec, TE = 92 msec, turbo factor = 32, FOV = 10 cm, slice 

thickness = 2 mm, matrix = 256 x 256, scan time = 6 min for 

each plane). The postmortem MR findings were compared 

with the autopsy findings as well as the in utero imaging. 

RESULTS 

Eight fetuses had open spinal defects, either myelocele or 

myelomeningocele (five at the thoracic level and three at the 

lumbo-sacral level) and one fetus had dorsal myeloschisis. 

All of the fetuses with myelocele or myelomeningocele had 

Arnold-Chiari II malformation and the fetus with dorsal 

myeloschisis did not. Postmortem MR imaging confirmed 

the antenatal ultrasound findings or in utero MR diagnosis in 

all eight cases of myelocele or myelomeningocele. The fetus 

with dorsal myeloschisis had in utero MR imaging at 20 and 

26 gestational week and an incorrect diagnosis of thoracic 

meningocele that was expanding was made on the in utero 

MR imaging. Postmortem MR imaging demonstrated deformity 

of the thoracic cord with neural tissue extending 

through the small bony spina bifida at the mid thoracic level 

into the collapsed cyst. Syringohydromyelia was not seen in 

the in utero MR or postmortem MR imaging in the case of 

dorsal myeloschisis. Autopsy findings were in agreement 

with the postmortem MR findings. 

CONCLUSION 

Postmortem MR imaging was helpful in clarifying the nature 

of abnormality seen either on antenatal ultrasound or in utero 

MR imaging. With the reduction in the number of autopsies 

performed, alternative noninvasive means of confirming the 

abnormalities seen on antenatal scans is crucial. There is a 

potential role for postmortem MR imaging as an adjunct or 

alternative to autopsy. 

KEY WORDS: Spine, malformations, postmortem 


MR Assessment of the Spine following in Utero 

Myelomeningocele Repair 

Zarnow, D. M. · Simon, E. M. · Bilaniuk, L. T. · Danzer, E. · 

Sutton, L. N. · Johnson, M. P. · Zimmerman, R. A. · Adzick, 

N. S. 

Children’s Hospital of Philadelphia 


PURPOSE 

Evaluate MR appearance of the spine following in utero 

myelomeningocele (MMC) repair and correlate imaging 

findings with surgical repair data.


MR imaging of the spine was evaluated retrospectively in 44 

patients following in utero MMC repair (24 female, 20 male; 

age range at examinations: 32 weeks - 4 years, 11 months). 

The number, appearance, and alignment of vertebral bodies 

and levels of posterior element dysplasia and frank deficiency 

were determined by counting from the craniocervical 

junction. The placode was classified as globular in shape, 

flat, or split around a cyst or dermoid, and if cerebrospinal 

fluid could be seen posterior to the placode. Intraspinal 

lesions were noted, including fat and dermoids. Surgically 

proven dermoids diagnosed on outside imaging were included 

also. Dilation of the central canal of the spinal cord or 

frank syringohydromyelia was noted. The spinal findings 

were correlated with hindbrain herniation, gestational age at 

time of repair, and type of surgical closure. 

RESULTS 

Forty of 44 (90%) patients had a full complement of vertebrae, 

with 3 having only four lumbar-type bodies, and one 

having six. Only one patient had a vertebral body anomaly, 

which resulted in focal kyphosis. The frank defect of the 

MMC ranged from the T10 through S1 levels [T10 (1), T12 

(2), L1 (3), L2 (1), L3 (3), L4 (12), L5 (16), S1 (6)]. All 

spines had at least one level of posterior element dysplasia 

immediately cephalad to the defect (range:1-4 levels, mean: 

1.68). Placode morphology was typically globular (50%) but 

CSF could be seen only dorsal to the placode in 11 cases 

(25%). Dermoids were present or known to have been 

resected surgically in 12 (27%) subjects, and were associated 

nearly always with L4 or L5 defects. There was no correlation 

of closure type and dermoid formation. On the first 

postnatal MR imaging, persistent hindbrain herniation of 

greater than 5 mm was seen in 6 subjects, all with L4 or L5 

lesions, and 1 with an L2 lesion (missed diastematomyelia 

causing tethering). Persistent hindbrain herniation was seen 

most commonly (71%) with alloderm closure (25% of all 

alloderm-closed patients). Worsening hindbrain herniation 

on follow-up MR imaging (seen in 35%) was more common 

with L5 and S1 lesions. Dermoids were not associated with 

persistent or increasing hindbrain herniation. Sixteen cases 

(36%) showed dilatation of the central canal, including one 

with frank syringohydromyelia; however these could not be 

correlated with lesion level, repair type, placode morphology, 

dermoid formation, or hindbrain herniation. There was 

no correlation of age at repair with dermoid formation, placode 

morphology, or hindbrain herniation. 

CONCLUSION 

The postnatal appearance of the spine following in utero 

MMC repair is heterogeneous with few consistent features to 

indicate spinal cord retethering. Posterior element dysplasia 

about the level of the frank defect is identified universally 

with detailed postnatal MR evaluation. The true incidence of 

dermoids and the factors leading to them remain unclear and 

should be determined by the randomized trial now in 

progress. Correlation of MR findings with prenatal ultrasound 

results and clinical status, including lower extremity 

function, is also ongoing. 

KEY WORDS: Myelomeningocele, fetal, MR imaging 

169 


3:00 PM - 4:30 PM 

Room 205 

(51d) ADULT BRAIN: New Imaging 

and Postprocessing Techniques 




Miscellaneous 




Moderators: Michael L. Lipton, MD 

Patrick A. Turski, MD 


Comparative Analysis between Perfusion CT and Spect 

Imaging in Cerebral Vasospasm 

Cohen, W. A. · Sviri, G. · Lewis, D. · Newell, D. · Britz, G. · 

Douville, C. 

Harborview Medical Center, University of Washington 

Seattle, WA 

PURPOSE 

The purpose of this study was to compare cerebral blood 

flow (CBF) and mean transit time (MTT) measured using 

CT perfusion (CTP) with CBF estimated by Tc-99m singlephoton 

emission CT (SPECT) in patients with cerebral 

vasospasm after aneurysmal subarachnoid hemorrhage. We 

hypothesized that CTP-derived CBF would correlate directly 

and that CTP-derived MTT would correlate inversely with 

SPECT measurements of CBF. 


Twenty-four patients admitted for subarachnoid hemorrhage 

had 35 CTP and 35 SPECT examinations performed. 

Nineteen studies were obtained because of worsening neurologic 

examinations and 16 because severe vasospasm in the 

anterior circulation was suggested by transcranial doppler. 

No more than 3 hours separated the CTP and SPECT exams. 

Eleven patients had 2 examinations. In each CTP study 

twelve regions of the brain were identified. Cerebral blood 

flow and MTT was measured in each of these regions. Ratios 

were calculated between CBF and MTT measured in regions 

within the anterior circulation to CBF and MTT in measured 

in normal appearing occipital lobe. These ratios were compared 

to similar qualitative estimations of regional CBF by 

Wednesday

Wednesday 

SPECT. In the SPECT studies CBF was estimated by comparing 

the regional flow to flow in the cerebellum. 

Subgroups were compared by ANOVA and alternate T test. 

RESULTS 

A high correlation was found between measurements of 

regional CBF and regional MTT by CTP and by SPECT. 

Cerebral blood flow was graded by SPECT as normal blood 

flow, mild hypoperfusion (CBF > 70-85%), moderatehypoperfusion 

(CBF = 50-70%), or severe hypoperfusion (CBF < 

50%). The corresponding mean, regional CTP-derived CBF 

ratios were: 1.01 ± 0.08, 0.82 ± 0.22, 0.6 ± 0.15 and 0.32 ± 

0.08, respectively (p < 0.0001). Corresponding mean, 

regional MTT ratios were 1.04 ± 0.14, 1.4 ± 0.31, 2.16 ± 

0.46 and 3.3 ± 0.54, respectively (p < 0.0001). Twenty-nine 

of 30 regions with severe hypoperfusion on SPECT had 

regional CBF ratios lower than 0.55 and regional MTT ratios 

higher than 2.5. Of the 323 brain regions with normal perfusion 

or mild hypoperfusion by SPECT, only 35 had regional 

CBF ratios lower than 0.75 and only 21 had regional MTT 

ratios higher then 1.75. 

CONCLUSION 

Regional measurements of CBF and MTT by CTP in patients 

with vasospasm after SAH correlate highly with estimated 

CBF on SPECT imaging. CT perfusion is an accurate tool 

for evaluation of post-SAH cerebral perfusion. CT perfusion 

should prove highly useful in the population with vasospasm 

because of scanner accessibility, speed of study acquisition, 

and the high anatomical resolution available with CT. 

KEY WORDS: Cerebral blood flow, vasospasm, CT 


Manual versus Automatic Selection of Arterial Input and 

Venous Output in Postprocessing of Dynamic Perfusion 

CT Data: Which One Is the More Reliable? 

Liu, S. · Wintermark, M. · Lu, Y. · Dillon, W. P. 



PURPOSE 

To compare the influence of manual and automatic selection 

of reference arterial and venous regions of interest (ROIs) on 

the dynamic perfusion CT results. 


We retrospectively identified 12 patients (5 males, 7 females; 

age range: 29~88) who underwent dynamic perfusion CT of 

the brain, and whose work up ruled out hemispheric stroke 

and intracranial hemorrhage. Each dynamic CT raw dataset 

was processed at GE AW 4.0 (CT perfusion 2.0), using four 

different methods differing only by the arterial input and the 

venous output: A) Arterial input (1 pixel) and venous output 

(1 pixel) were manually selected within the anterior cerebral 

artery (ACA) lumen and the superior sagittal sinus (SSS) as 

the pixel with the maximal enhancement in these structures; 

B) Same approach but with arterial and venous ROIs of 6 

pixels; C) Arterial input (11 pixels) in the ACA and venous 

output (18-65 pixels) in the SSS were selected automatically 

by the used software; D) Manual selection of the arterial 

input (1 pixel) and automatic selection of the venous output 

170 

(18-65 pixels) by the used software. Perfusion CT regional 

cerebral blood flow (rCBF) and volume (rCBV) and mean 

transit time (MTT) parameters were measured in the whole 

brain, the basal ganglia, the middle cerebral artery territory 

and the white matter, in identical ROIs for the four methods. 

The time required to process the dataset according to the four 

methods was recorded. Measurements were compared using 

ANOVA, and paired t-test analysis. 

RESULTS 

Absolute values of rCBF, rCBV, and MTT measured in the 

above-mentioned structures were not significantly different 

(ANOVA p values on 4 ROIs > 0.700, > 0.500, and > 0.990, 

for rCBF, rCBV, and MTT, respectively). We observed a 

trend for rCBV and rCBF values to be increased with methods 

using automatic selection of the venous output (p-value 

of paired t-test between Method A and D on each ROI: rCBV 

< 0.05, rCBF < 0.05, but MTT > 0.30 ). This was due to the 

relatively large size of the automatic venous ROI (18-65 pixels), 

generating partial volume averaging effect and 

decreased area under the time-enhancing curve for the 

venous output ROI. The time required to postprocess the 

data was: 7.0 min, 8.0 min, 3.0 min, and 4.0 min, for methods 

A, B, C, and D, respectively. 

CONCLUSION 

Using automatic selection of arterial input and venous output 

for perfusion CT data postprocessing is time saving and reliable. 

However, manufacturers must use moderate-sized 

ROIs in their automatic algorithms in order for the results to 

remain accurate. 

KEY WORDS: Perfusion, CT, postprocessing


Identification of Ischemic Thresholds for the Infarct 

Core and the Ischemic Penumbra Utilizing Quantitative 

Cerebral Blood Flow Measurements by MR Imaging 

Shaibani, A. · Shin, W. · Khawar, S. · Sakaie, K. · Walker, M. 

T. · Cashen, T. · Bernstein, R. · Carroll, T. R. 

Northwestern University Feinberg School 

Chicago, IL 

PURPOSE 

We have developed a unique quantitative measurement of 

cerebral blood flow utilizing MR imaging. We applied this 

technique in a prospective manner to those patients presenting 

with acute stroke, in whom MR evaluation was requested. 

Ischemic thresholds for irreversible infarction, and 

ischemic penumbra were obtained by comparing the quantitative 

cerebral blood flow (QCBF) values with the diffusionweighted 

images and subsequent cross-sectional imaging. 


We have developed a unique method for the quantification of 

cerebral blood flow parameters (CBF, CBV) with MR imaging, 

which can be performed and calculated on a patient-bypatient 

basis, without the need to rely on population-averaged 

values of flow. Our quantitative CBF imaging is based 

on combining a steady-state free precession MR imaging 

pulse sequence with our standard clinical MR gradient-echo 

EPI perfusion scan. The true-fisp sequence is performed 

before and after the standard perfusion sequence, in “bookend” 

fashion, and gives information critical for the calculation 

of absolute CBV and therefore CBF. We have validated 

previously the technique in normal volunteers, and in animal 

studies (compared with fluorescent microspheres). To date 6 

patients (accrual is continuing) with acute or early subacute 

stroke were evaluated with standard MR imaging, diffusionweighted 

imaging, and quantitative MR perfusion. Cerebral 

blood flow values were calculated and compared with diffusion-weighted 

images and final stroke volume based on 

anatomical (T2) imaging. 

RESULTS 

In all 6 patients, the QCBF values for the final infarct volume 

were consistent with known ischemic thresholds (15 +/- 4 ml/100g/min. 

KEY WORDS: Quantitative cerebral blood flow, stroke, 

ischemic thresholds 


Inversion-Recovery versus Gradient-Echo T1-Weighted 

3 T Brain MR Imaging 

Wang, P. Y. 1 · Howieson, J. 1 · Hamilton, B. E. 2 

1Oregon Health and Science University, Portland, OR, 

2University of Utah Health Sciences Center, Salt Lake City, 

UT 

PURPOSE 

3 T brain gadolinium-enhanced T1 turbo spin-echo MR 

imaging is degraded significantly by posterior fossa venous 

flow-induced artifacts. We implemented an inversion-recovery 

pulse sequence for T1-weighted 3 T MR imaging 

addressing the problems of arterial enhancement and suboptimal 

gray-white matter differentiation seen with gradientecho 

sequences. Comparison then was made between a 

spoiled gradient-echo sequence (T1 FFE) and those from this 

inversion-recovery sequence using an inversion prepulse 

with a turbo spin-echo readout (T1 FLAIR). 


Fifteen patients (11 male, 4 female, ages 18 months-77 

years) underwent routine MR imaging on a Philips Intera 3 

T MR unit with sagittal T1 TSE, axial proton-density, axial 

T2, axial T1 FFE, postgadolinium axial and coronal T1 FFE 

with either a conventional or SENSE acceleration factor 1.7. 

Each also received T1 FLAIR imaging with either pre (3/15) 

or post (13/15) gadolinium enhancement. T1 FFE parameters 

were TR/TE/flip 200-240/2.3 ms/80, SLT 4 skip 1 mm 

(duration 1:21). T1 FLAIR parameters were TR/TE/TI/flip 

2300/10/1950/90, SLT 4 skip 1 mm, echo train length 6, conventional 

or with SENSE acceleration factor 1.5 

(duration:conventional 4:30, SENSE 2:45). Assessment of 

the appearance of gray-white matter contrast, fat-soft tissue 

contrast, vessel enhancement, conspicuity of gadolinium 

enhancement, and chronic brain infarction were made by 3 

neuroradiologists. 

Wednesday

Wednesday 

RESULTS 

Inversion-recovery turbo spin-echo (T1 FLAIR) had excellent 

gray-white matter contrast though with slight posterior 

fossa venous flow artifact and slightly lower resolution. 

There was no high arterial signal intensity pre or postcontrast 

and the images were free of CSF pulsation artifacts. SENSE 

did contribute slightly to posterior fossa flow artifact. T1 

gradient-echo images (T1 FFE) were free of posterior fossa 

venous artifacts but had several findings not seen in postgadolinium 

T1 spin-echo images familiar to radiologists. 

There was consistently suboptimal gray-white matter contrast 

and arterial high-signal intensity on precontrast images 

which were further accentuated by gadolinium administration. 

Cerebrospinal fluid pulsation artifacts also were 

encountered often in the basal cisterns, and there was variable 

fat-soft tissue contrast on postgadolinium images. 

CONCLUSION 

Compared to T1-weighted gradient-echo sequences, T1 

FLAIR had significantly better gray-white matter contrast 

and was free of CSF and arterial motion artifacts. However, 

there was slightly less image resolution and mild posterior 

fossa venous flow phase. Inversion-recovery turbo spin-echo 

was found to be motion-tolerant and gave outstanding T1 

contrast. 

REFERENCES 

1. Melhem ER, et al. Fast inversion-recovery MR: the effect of 

hybrid RARE readout on the null points of fat and cerebrospinal 

fluid. AJNR Am J Neuroradiol 1997;18:1627-1633 

KEY WORDS: Brain, 3 T, T1 


Clinical Relevance of FLAIR with PROPELLER in 

Evaluating the Brain at 3 T 

Shapiro, M. D. · Ramnath, R. · Hartker, R. · Williams, D. · 

Magee, T. · Wasudev, N. · Lees, J. 

Neuroskeletal Imaging of Winter Park 

Winter Park, FL 

PURPOSE 

Periodically rotated overlapping parallel lines with enhanced 

reconstruction (PROPELLER) is a relatively new method of 

collecting image data with the filling of k space in radial 

lines (blades) which rotate in sequence until the image is 

acquired completely. This technique has been described previously 

at 1.5 T with diffusion imaging in the brain and T2 

FSE imaging in the brain and C spine. PROPELLER imaging 

has proven to be extremely successful in decreasing both 

patient motion and magnetic susceptibility artifacts. 


We prospectively scanned 100 clinical patients’ (85 adult 

and 15 children) brains on a GE 3 T Excite with HD scanner 

using both routine axial 2D FLAIR FSE and an axial 2D 

FLAIR FSE with PROPELLER (both with TR -2875, TE -8, 

IR -1100, 3 mm slice thickness (in addition to our routine T1, 

T2 and diffusion pulse sequences). 

172 

RESULTS 

Although the study was designed to be blinded for the two 

neuroradiologists, it was blatantly obvious (related to the 

marked disparity in magnetic in the size of susceptibility 

artifact from the sinuses and skull base at 3 T) which images 

were obtained with PROPELLER since they obviously were 

superior. In addition in both children and adults with excessive 

head motion during the MR scan the difference in quality 

between the two pulse sequences was obvious. 

CONCLUSION 

At 3 T, 2D FLAIR with PROPELLER obviously is superior 

for clinical brain imaging (in both adults and children) to 

routine 2D FLAIR without PROPELLER due to the marked 

reduction in both magnetic susceptibility and patient motion 

artifact. Limitations of PROPELLER at the present time 

include: 1) The inability to image in the coronal or sagittal 

planes. Currently only axial or up to 40 degrees off-axial is 

possible. 2) It is impossible to currently obtain any T1weighted 

2D pulse sequence or any 3D pulse sequence with 

PROPELLER. Diffusion with PROPELLER at 3 T has the 

same advantages already well described at 1.5 T but with 

significant prolongation of scan time. 

KEY WORDS: PROPELLER, FLAIR, 3 T 


Safety of Monthly Triple-Dose Gadopentetate 

Dimeglumine: Preliminary Results of the BECOME 

Trial 

Wolansky, L. J. · Cadavid, D. · Haghighi, M. H. · Liu, J. · 

Tulloch, K. · Joseph, G. · Sethi, N. · Sevdalis, E. · Cook, S. D. 

New Jersey Medical School/University of Medicine & 

Dentistry of New Jersey 

Newark, NJ 

PURPOSE 

Triple-dose gadolinium (1) with postinjection delay has 

shown value in evaluating multiple sclerosis by increasing 

conspicuity of contrast enhancement. The “BECOME” study 

(Phase IV, rater-blinded, randomized study, comparing the 

effects of 250 mg of Betaseron with 20 mg of Copaxone in 

patients with the relapsing-remitting or clinically isolated 

forms of multiple sclerosis using 3 T MR imaging with 

triple-dose gadolinium) is a clinical trial now underway at 

our institution. To monitor relative drug efficacy, monthly 

triple-dose (0.3 mmol/kg) gadopentetate dimeglumine 

(Magnevist, Berlex, Montville, NJ) is being used for MR 

images on up to 14 consecutive occasions in up to 80 subjects. 

No prior study had employed triple-dose on a monthly 

basis for more than four consecutive monthly doses. While 

the efficacy data from the trial will be complete in December 

2005, preliminary safety data are available now. 


Monthly scan questionnaires were filled out at the time of 

MR imaging visits, specifically addressing the question of 

adverse reactions (AEs). The treating physicians also specifically 

inquired about possible AEs at the time of follow-up 

visits. In addition to reviewing these records, details of AEs 

were obtained through follow-up phone calls. At the time of 

this writing, 514 monthly triple-doses of gadopentetate

dimeglumine had been administered on at least two consecutive 

occasions in 57 subjects. Forty-five patients had 

received monthly triple-dose at least five times. One hundred 

and seventy total triple-doses had been administered in 36 

subjects on more than eight consecutive monthly occasions. 

Fifteen subjects had completed the protocol, receiving 14 

monthly triple-doses. 

RESULTS 

Of the total 514 doses, there were only 24 adverse events 

(4.7%) associated with 17 of the injections. The most common 

of these were injection site reaction, seen in 10 of 514 

injections, and headache, seen in 7 of 514 injections. The 

rate of AEs during the first 7 months was 22 of 344 injections 

(6.4%). Interestingly, the rate of AEs after 7 months 

was only 2 of 170 injections (1.2%). There was one serious 

adverse event (0.2%), in which iv fluid and an indeterminate 

amount of contrast extravasated, leading to persisting paresthesia 

involving part of the hand. 

CONCLUSION 

The most commonly encountered AEs were injection site 

reactions (1.9%) and headaches (1.4%), seen on the same 

order of magnitude to those reported by the manufacturer in 

clinical trials for single-dose injections on a single occasion 

(2.3% and 4.8%, respectively). The preliminary evidence 

suggests that monthly triple-dosing with gadopentetate 

dimeglumine is safe and should prove valuable for clinical 

trials. 

REFERENCES 

1. Wolansky L, Bardini J, Cook S, et al. Triple-dose versus singledose 

gadoteridol in multiple sclerosis patients. J Neuroimag 

1994;4:141-145 

KEY WORDS: Gadolinium, high-dose, safety 


Novel Technique for the Estimation of Cerebral Cortical 

Thickness 

Jackson, A. · Scott, M. L. J. · Thacker, N. A. 

University of Manchester 

Manchester, UNITED KINGDOM 

PURPOSE 

To assess change in regional human cerebral cortical thickness 

with normal aging, using a novel technique applied to 

MR images. 


T2 inversion-recovery MR scans (1.5 T, TI/TR/TE = 

300/6850/18 ms, pixel size = 1.8 x 1.8 mm, 51 slices, thickness 

= 3.0 or 4.0 mm) were taken of the entire cerebrum in 

119 normal volunteers (52 male, mean age = 70.3 years, 

range = 19-86 years). The MR image volumes were preprocessed 

into a form suitable for cortical thickness estimation. 

The mean and standard deviation of the image graylevel 

values of the gray matter (GM), white matter (WM) 

and CSF were determined and used to up-interpolate the data 

to half the thickness in the through-plane direction. This volume 

then was registered to the Talairach atlas, defining 31 

sublobar regions for which cortical thickness was later 

obtained. The GM in the images was segmented, producing 

173 

a volume of GM probability maps. The GM thickness was 

estimated as follows. The GM/WM boundary in the registered 

images was found using an iso-contour detector which 

identified voxels consistent with the average of the GM and 

WM mean values. The 3D surface normal to the boundary 

was determined at each edge voxel, and a search performed 

in this direction (through the GM) on the segmented GM 

images until an opposing GM edge was found. The distance 

traversed was recorded. The edge may be a true edge, a transition 

across the 50% GM probability or due to partial 

voluming of GM with other tissue. In both cases, the edge 

may be with either WM or CSF. If it was WM, it was 

assumed that 2 gyral banks had been traversed, and the 

length duly halved. A robust estimate of regional thickness 

was obtained by taking the median of the lengths in each cortical 

region. The statistical error on the majority of regions 

was approximately 0.25 mm. ANOVAs were performed for 

each region over the entire dataset, correlating cortical thickness 

against sex, head size (using a surrogate for intracranial 

volume) and age. 

RESULTS 

There were no significant correlations between age and sex 

or head size. Sex was highly significant in predicting head 

size (p

Wednesday 


Seven right-handed patients affected by drug resistant 

Parkinson’s disease (Hoehn & Yahr score ≥ 3), were studied 

with BOLD-based functional MR imaging (fMRI). Patients 

underwent bilateral implantation of electrodes in the STN. 

Daily therapy was not changed the day of fMRI, performed 

6 days after surgical implantation and before connection to 

subcutaneous pacemaker. Left electrode was used for STN 

stimulation during fMRI in 4 cases, right electrode in 2 cases 

and both electrodes in 1 case. In 2 cases motor task was performed 

also with the hand ipsilateral to the activated electrode. 

Finger tapping task was performed using the contralateral 

hand. An external pacemaker set for therapeutic 

effects without collateral problems was used, connected with 

intracerebral electrode with a 10 mt wire. Functional MR 

imaging was divided into 3 phases: 1) pacemaker on and off 

(DBS-on/off) without voluntary movements; 2) left electrode 

not active (DBS-off) with 6 periods of rest and selfpaced 

right hand movements; 3) activation of left electrode 

(DBS-on) with 6 periods of rest and self-paced movements. 

Tasks were performed with a 1.5 T scanner (64 MHz; 25 

mT/m), using EPI sequences extended from the base to the 

vertex positioned parallel to the commissural line (transmit/receive 

head coil). Statistical parametric mapping 

(SPM99) implemented in MatLab 6 was used for statistical 

analysis using Student’s t-test (correct p value = 0.05 and 

uncorrect p value = 0.001/0.01) for false positive. 

RESULTS 

One patient interrupted fMRI exam and 1 patient was not 

valuable for motion artifacts. Five patients completed all 

tasks: 1) no side effect was observed during and after MR 

imaging protocol; 2) with DBS on/off tasks activations distant 

from the STN were observed, in the limbic cortex and in 

frontal cortex (suppl. motor area BA 6); 3) comparison of 

DBS-off/DBS-on data obtained during self-paced movements 

showed reduction of activated cortical areas when performed 

with DBS-on. This suggested a selection and focalization 

of neural activity especially in primary motor cortex 

(BA 3/4) induced by NST stimulation. 

CONCLUSION 

In this specific condition BOLD-based fMRI demonstrated 

to be safe even with DBS electrodes in the STN. Our preliminary 

experience in this small group of patients showed 

that DBS should facilitate neural circuits through the SMA. 

Moreover DBS should facilitate voluntary movements 

reducing activations of associative corticies and focusing 

neural activity in primary motor cortex. 

KEY WORDS: Functional MR imaging, Parkinson’s disease, 

deep brain stimulation 

174 


Assessment of Diffusion Characteristics for Motor 

Pathway Fiber Tracks in Multiple Sclerosis Using 

Diffusion Tensor Imaging 

Borg, B. 1 · Phillips, M. D. 1 · Marrie, R. 1 · Stone, L. 1 · 

Bhattacharyya, P. 1 · Hirsch, J. 2 · Gass, A. 2 · Lowe, M. J. 1 

1 2 Cleveland Clinic Foundation, Cleveland, OH, Basel 

Universitaetspital, Basel, SWITZERLAND 

PURPOSE 

To test the diffusion characteristics of individual fiber tracts 

in patients with multiple sclerosis (MS) using diffusion tensor 

imaging (DTI). 


Six patients with early stage MS (median EDSS = 3 ± 2) and 

six age- and gender-matched control subjects were recruited 

to participate in this study. All subjects participated in a scan 

session using a Siemens Allegra 3 T MR scanner that included 

a high-resolution T1-weighted anatomical scan, an fMRI 

activation study acquired during the bilateral finger tapping 

and a six direction whole-brain diffusion tensor imaging 

study utilizing 2 mm isotropic voxels. Functional MR imaging 

activation maps were used to produce seed regions in the 

white matter near the bilateral supplementary motor area 

(SMA). Diffusion tensor imaging data were processed using 

a multistep procedure consisting of eddy-current correction, 

motion correction, followed by production of diffusion tensor 

eigenvalues. For each voxel, the eigenvalues are used to 

produce volume ratio index (VRI), a measure of white matter 

anisotropy. The calculated diffusion tensor is used to 

track fibers in the white matter pathways using the method 

of Mori, et al. Threshold values used to terminate a given 

track were VRI < 0.05, and deflection angle > 80 o . These 

parameters were observed to allow tracking to progress 

through some MS lesions. Using SMA seed regions tracks 

passing through the body of the corpus callosum were selected 

for further analysis. Mean VRI, mean principal eigenvalue 

(L1), and mean quadrature sum of nonprincipal eigenvalues 

(L2) were calculated for each callosal track in each subject. 

RESULTS 

Mean VRI values demonstrated reduced anisotropy in MS 

subjects (0.348 ± 0.047) in comparison to controls (0.424 ± 

-0.03)( p < 0.07). Mean L2 values showed a corresponding 

increase in patients with MS (0.783 x 10 -3 mm 2 /s ± 0.073) in 

comparison to controls (0.664 x 10 x 10 -3 mm 2 /s ± 0.039) (p 

< 0.07). In contrast, mean L1 values were unchanged 

between MS subjects (1.348 x 10 -3 mm 2 /s ± 0.050) and controls 

(1.362 x 10 -3 mm 2 /s ± 0.039) (p < 0.9). Findings suggest 

that reported decrease in white matter anisotropy in multiple 

sclerosis is largely due to an increase in the diffusivity of 

water perpendicular to fiber track axons while diffusivity 

parallel to the direction of axons is apparently unaffected by 

MS. 

CONCLUSION 

Incorporating a novel fMRI-based selection of cortical seed 

regions for the DTI-based fiber tracking of inhemispheric 

motor pathways the present study demonstrates a reduction 

in diffusivity perpendicular to fiber tracts with preservation 

of diffusivity parallel to fiber tracts. Findings suggest that the

eduction in fractional anisotropy observed in the white matter 

of MS patients is largely due to an increase in transverse 

diffusivity. 

KEY WORDS: Functional MR imaging, diffusion tensor 

imaging, multiple sclerosis 


Cerebral Cortex Plasticity in Ischemic Stroke Patient: 

Evaluated by Active and Passive Clenching Task 

Functional MR Imaging in 3 T 

Liu, X. 1 · Li, L. 2 · Dai, J. 2 · Westesson, P. 1 

1 University of Rochester School of Medicine and Dentistry, 

Rochester, NY, 2 Beijing Tiantan Hospital, Beijing, CHINA 

PURPOSE 

To analyze cerebral cortex compensation after ischemic 

stroke, by combining analysis of passive and active clenching 

task functional MR imaging (fMRI) in 3 T. 


Eleven patients (10 males, 1 female), who recovered to 

Twitchell-Brunstorm stage Vor VI after MCA territory ischemic 

stroke and 16 volunteers (9 males and 7 females) were analyzed. 

Active and passive clenching task fMRI(multi phase EPI, 

TR/TE = 3000/30) was performed in GE Signa 3 T. The fMRI 

data were analyzed by SPM99, p < 0.05. 

RESULTS 

The patient passive clenching fMRI result shows not only 

activation in Brodmann 1, 2, 3 area of primary sensory cortex, 

but also activation in Brodmann 4, 6 of first somatic 

motor area. The patient active clenching fMRI result shows 

more activation in the 2, 3 than the primary motor cortex. 

Compared with the active task result of volunteers by SPM 

group analysis, there are differences in bilateral Brodmann 3, 

5 and 40 area of patient (Fig. 1 active clenching activation 

difference between patient and volunteer group). 

CONCLUSION 

The passive clenching task could trigger motor cortex activation, 

which may be the mechanism that explains why passive 

rehabilitation treatment is useful for patient motor function 

recovery. The result, that more activation in the 

Brodmann 2, 3, 5, and 40 area during active clenching task 

of patient, suggests that both sensory cortex and postparietal 

cortex compensate motor function. These two areas are 

important for the cerebral cortex plasticity after ischemic 

stroke. 

KEY WORDS: Functional MR imaging, plasticity, ischemia 

175 


High Angular Resolution Diffusion Imaging of 

Intravoxel Crossing Fiber Tracts in the Human 

Brainstem Using Parallel Image Acquisition at 3 T 

Mukherjee, P. 1 · Hess, C. P. 1 · Han, E. T. 2 · Xu, D. 1 · Vigneron, 

D. B. 1 

1 University of California San Francisco, San Francisco, CA, 

2 General Electric Applied Sciences Lab West, Menlo Park, 

CA 

PURPOSE 

Diffusion tensor imaging (DTI) cannot resolve multiple fiber 

orientations within a single voxel in regions of complex 

white matter architecture. High angular resolution diffusion 

imaging (HARDI) overcomes this limitation through the 

acquisition of large numbers of diffusion directional measurements 

at diffusion-weighting factors (b values) greatly 

exceeding 1000 s/mm 2 . In this study, 3 T HARDI with parallel 

imaging using an 8-channel head coil is employed to 

visualize the complex white matter architecture of the human 

brainstem, including intravoxel crossing fibers at the transverse 

pontine tracts and the pyramidal decussation. 


High angular resolution diffusion imaging was performed on 

three adult volunteers using a 3 T EXCITE scanner (General 

Electric, Milwaukee, WI). The 8-channel EXCITE head coil 

was employed for parallel imaging using the array spatial 

sensitivity encoding technique (ASSET) with an acceleration 

factor of 2. A multislice interleaved axial HARDI acquisition 

encompassing the entire brainstem was performed 

with a single-shot echo-planar spin-echo pulse sequence (TR 

= 6 s, TE = 93 ms, NEX = 1) with 131 diffusion-encoding 

directions at b = 3000 s/mm 2 and 1.6 mm isotropic spatial 

resolution. In two of the subjects, an additional HARDI 

acquisition covering only the medulla was performed with 

282 diffusion-encoding directions at b = 3000 s/mm 2 and 1.5 

mm isotropic spatial resolution. The orientation distribution 

function (ODF) was constructed for each voxel using Q-ball 

(1) and spherical deconvolution (2) methods. The resulting 

ODFs were compared with DTI analysis, including directionally 

encoded color fractional anisotropy (FA) maps. 

RESULTS 

Parallel imaging at 3 T using a multichannel phased-array 

head coil provided unprecedented spatial resolution for diffusion 

imaging in the human brainstem, without the strong 

warping artifacts usually encountered in the posterior fossa 

with single-shot echo-planar imaging at high-field and ultrahigh 

diffusion-weighting factors. In all subjects, both the Qball 

and spherical deconvolution ODFs generated reconstructions 

of intravoxel crossing fiber tracts in the brainstem 

that are consistent with known white matter anatomy. These 

included (1) craniocaudally oriented projection fibers of the 

pyramidal tract that intersect with transversely oriented pontocerebellar 

fibers in the pons; and (2) crossing fibers of the 

pyramidal decussation in the core of the medulla (Fig.). 

Diffusion tensor imaging analysis of this HARDI data provided 

no useful fiber orientation information in regions of 

complex white matter architecture. 

Wednesday

Wednesday 

CONCLUSION 

3 T HARDI with parallel imaging enables visualization of 

the white matter architecture of the human brainstem in 

unprecedented detail, including intravoxel crossing fibers 

where DTI fails. 

REFERENCES 

1. Tuch DS, et al. Neuron 2003;40:885-895 

2. Tournier JD, et al. NeuroImage 2004;23:1176-1218 

KEY WORDS: Diffusion, tensor, tractography 


3:00 PM - 4:30 PM 

Room 103 


PowerPoint 

— David S. Martin, MD 

176 


4:40 PM - 5:40 PM 

Room 205 

(53) ELC Lecture F: fMRI 

Postprocessing 


4:40 PM - 6:10 PM 

Room 107 

(54) Pediatric Case-Based Studies 

(ASPNR) (ARS)* 

Panelists: 

— Ashok Panigrahy, MD 

— Blaise V. Jones, MD 

— P. Ellen Grant, MD 

— Gary L. Hedlund, DO 

— Jill V. Hunter, MD 

— Charles M. Glasier, MD 

— Yutaka Sato, MD 

— Lisa H. Lowe, MD 

Moderators: Charles M. Glasier, MD 

Gary L. Hedlund, DO 


— Timothy P.L. Roberts, PhD


4:40 PM - 6:10 PM 

Room 105/106 

(55) Neuroradiology Core Curriculum 

(General) 

(332) Head Trauma 

Head Trauma 

Eric J. Russell, MD, FACR 

— Eric J. Russell, MD, FACR 

(333) Demyelinating Disease 

— Robert I. Grossman, MD 

(334) Intracranial Infection 

— Robert D. Zimmerman, MD 

Moderator: Eric J. Russell, MD, FACR 




able to: 

1) Demonstrate through case-based review a rational imaging 

approach to the evaluation of the adult patient with traumatic 

brain injury. 

2) Discuss optimal CT and MR-based imaging techniques to 

detect and categorize pathologic brain anatomy and function 

following trauma. 


Cerebral trauma is a devastating disease of the young, with a 

peak incidence in the second and third decades (15-24 

years). Rapid imaging characterization combined with modern 

neurosurgical and medical therapies can minimize longterm 

sequelae. Computed tomography (CT) can rapidly 

detect and characterize acute traumatic lesions, and aid in the 

selection of patients requiring emergency surgery. Magnetic 

resonance (MR) imaging is more sensitive to most parenchymal 

posttraumatic lesions, including nonhemorrhagic contusion, 

diffuse axonal injury (DAI) and brainstem injuries. MR 

imaging is, therefore, the procedure of choice for the investigation 

of the subacute and chronic sequelae of trauma and 

is an adjunct to CT for acute assessment. MR imaging also 

is useful for predicting the long-term deficits in cognitive 

function that follow closed head injury. Technological 

advances in both modalities have expanded the role imaging 

plays in the management of the head injured patient. Rapid 

assessment of associated vascular injuries is afforded by 

177 

CTA and MRA. Subtle changes in the brain related to posttraumatic 

clinical deficits and prognosis may not be 

detectable with anatomical imaging techniques. The recent 

development of functional imaging techniques such as diffusion-weighted 

MR imaging, MR spectroscopy (MRS) and 

CT and MR perfusion imaging provide additional information 

in the correct clinical setting. Accurate interpretation of 

the key anatomical features of the brain affected by trauma 

requires familiarity with normal cross-sectional anatomy of 

the brain and its dural coverings, and recognition of the 

evolving appearance of hemorrhage when imaged by CT and 

MR imaging. Gradient-echo T2* weighted and FLAIR 

sequences are particularly valuable for detecting parenchymal 

and subarachnoid hemorrhage respectively. Ultrafast 

echo-planar scans can be used to freeze patient motion and 

obtain critical information with MR imaging in uncooperative 

patients. Diffusion tensor imaging can delineate subtle 

changes on directional diffusion in white matter that may 

indicate brain damage when other studies are normal, and 

provide maps of white matter useful for assessing traumatic 

sequelae. Pathologic entitles reviewed include primary 

(direct result of trauma) and secondary (effects following 

initial injury) events: 

Traumatic subarachnoid hemorrhage 

Shearing injuries (Diffuse axonal injury) 

Parenchymal (Intraaxial) Lesions: Cerebral contusion, 

Intracerebral hematoma 

Extraaxial lesions: Subdural hematoma, Epidural 

hematoma, Herniation and infarction 

Vascular injuries 

Chronic sequelae of trauma: Atrophy, Encephalomalacia. 

REFERENCES 

1. Gean AD. Imaging of Head Trauma. New York, Raven 

Press;1995 

2. George AE, Russell EJ, Kricheff II. White matter buckling: 

CT sign of extra-axial intracranial mass. AJNR Am J 

Neuroradiol 1980;1:425-430 . 

3. MacKenzie JD, Siddiqi F, Babb JS, Bagley LJ, Mannon LJ, 

Sinson GP, Grossman RI. Brain atrophy in mild or moderate 

traumatic brain injury: A longitudinal quantitative analysis. 


4. Huisman TAGM, Schwamm LH, Schaefer PW, Koroshetz WJ, 

Shetty-Alva N, Ozsunar Y, Wu O, et al. Diffusion tensor imaging 

as potential biomarker of white matter injury in diffuse 

axonal injury. AJNR Am J Neuroradiol 2004;25:370-376 

5. Arfanakis K, Haughton VM, Carew JD, et al. Diffusion tensor 

MR imaging in diffuse axonal injury. AJNR Am J Neuroradiol 

2002;23:794-802 

Demyelinating Disease 




able to: 

1) To understand the imaging features of MS and differentiate 

it from other white matter diseases. 

2) To apply new techniques in the understanding of the 

pathology of MS. 

3) To gain knowledge of the imaging characteristics and 

clinical presentation of common white matter diseases. 

Wednesday

Wednesday 


White matter diseases are among the most common that neuroradiologists 

confront. The discussion will begin with the 

diagnosis of multiple sclerosis (MS). Multiple sclerosis has 

a variety of presentations in the brain and spinal cord. There 

are a few helpful features that aid in distinguishing MS from 

other lesions. They include: (1) the morphology of the 

lesions (ovoid); (2) its perivenous distribution; (3) involvement 

at the callosal-septal interface; (4) u-fiber lesions; (5) 

spinal cord involvement of less than two vertebral body segments; 

and (5) peripherally located spinal cord lesions. It 

will offer insights into the pathobiology of the disease based 

upon new technologies including perfusion, MRS, and iron 

imaging and propose a vascular hypothesis regarding MS. 

The next segment of the lecture will center on the differential 

diagnosis of MS. Schemes to distinguish MS from conditions 

that can mimic it on MR imaging will be elucidated. 

These mimics include tumufactive lesions, vascular lesions, 

and other unusual entities. To correctly interpret white matter 

lesions requires the combination of history and knowledge 

of particular diseases. The last segment of the lecture 

will feature those diseases with known etiologies. The MR 

and clinical findings for these entities will be highlighted. 

Intracranial Infection 




able to: 

1) Identify the pathoanatomical, temporal, and physiologic 

imaging features of intracranial infectious and inflammatory 

diseases that allow for early and accurate diagnosis. 

2) Understand how organism type, mode of spread and host 

response determine the imaging features of infectious and 

inflammatory lesions. 

3) Identify the unique DWI and MRS features of purulent 

fluid (pyogenic infections). 

4) Identify the imaging features of CNS HIV infection and 

it’s complications. 


It is the purpose of this presentation to review the imaging 

features of infectious and inflammatory lesions of the central 

nervous system (CNS). The advent of noninvasive imaging 

techniques has had a profound effect on prognosis and outcome 

in these disorders. Effective treatments exist for many 

infectious lesions but prognosis and outcome are affected 

directly by accuracy and timeliness of diagnosis. 

Inflammatory lesions are less common than neoplastic and 

ischemic processes and each type of inflammatory lesion has 

unique features. Therefore, it has taken time for distinctive 

patterns of imaging features to become apparent. More 

importantly, advances in imaging technology have lead to an 

enhanced ability to characterize infectious processes. 

Improvements in “morphologic” imaging have made it possible 

to detect lesions in the brain, spinal cord, and the 

meninges and to precisely identify their anatomical location 

and distribution. Physiologic imaging techniques such as diffusion-weighted 

imaging (DWI) and MR spectroscopy 

(MRS) have been applied to inflammatory lesions with great 

success. Many inflammatory processes have restricted diffu- 

178 

sion (increased signal on DWI) and/or have distinct MRS 

“signatures” and these findings allow for reliable differentiation 

from processes that are morphologically similar. The 

appearance of inflammatory lesions is a reflection of multiple 

factors including type of organism, mode of spread, host 

response, and histopathology. Most infections spread to the 

CNS hematogenously. Direct spread from adjacent structures 

such as the sinuses, nasopharynx or mastoid air cells, 

formerly common, has become less frequent since the introduction 

of antibiotics. Under normal circumstances, the 

brain is well protected by the calvarium, dura, and bloodbrain 

barrier. Disruption of the covering of the brain by trauma, 

surgery, congenital abnormalities, tumors or 

osteomyelitis commonly leads to the development of CNS 

infection. Retrograde axoplasmic flow along cranial or 

peripheral nerves may lead to invasion of the CNS by viral 

agents. Once infection is established, damage to the CNS is 

common because of unique features including absence of 

brain lymphatics, subarachnoid space capillaries and CSF 

immunoglobulin. The imaging features of CNS infections 

can be classified by location of the lesion, host response, 

and/or organism. No single classification scheme allows for 

optimal understanding of these features but by combining 

them the best mix of general and specific imaging findings 

can be presented. 


4:45 PM - 6:15 PM 

Theatre 


Spectroscopy 

(335) MR Spectroscopy: Basics and Techniques 

— J. Michael Tyszka, PhD 

(336) MR Spectroscopy: Adult Applications 

— Mauricio Castillo, MD 

(337) MR Spectroscopy: Pediatric Applications 

— Robert A. Zimmerman, MD 

Moderator: Howard A. Rowley, MD 


MR Spectroscopy: Basics and Techniques 

J. Michael Tyszka, PhD 

Dr. Tyszka received an honors BA in physics (1987) and a 

PhD (1993) in magnetic resonance imaging from Cambridge 

University, England. He also holds an MSc (1988) in infor-

mation technology from Aberdeen University, Scotland. He 

moved to California in 1992 and has since worked in clinical 

magnetic resonance research at the University of 

Southern California, Cedars-Sinai Medical Center, and City 

of Hope National Medical Center. He is currently Director of 

Magnetic Resonance Physics for the Caltech Brain Imaging 

Center and a consultant to Children's Hospital Los Angeles. 

His current research interests range from ultrahigh resolution 

MR microscopy of mass transport in living systems to 

functional connectivity mapping of the human brain. 



able to: 

1) Identify the physical phenomena underlying magnetic resonance 

spectroscopy. 

2) Review the pulse sequences for volume localization and 

unwanted signal suppression in general use for clinical 

MRS. 

3) Discuss the most important obstacles to successful in vivo 

MRS and methods for overcoming these obstacles. 

4) Judge the importance of spectral quantitation in the analysis 

and interpretation of clinical MRS results. 


This presentation covers the essential background to the 

practice of magnetic resonance spectroscopy (MRS) in clinical 

neuroradiology. The fundamental phenomena underlying 

magnetic resonance including nuclear spin paramagnetism, 

chemical shift, scalar coupling and relaxation will be 

reviewed. The popular pulse sequences available for in vivo 

MRS will be presented including pulse-acquire, spin echo, 

volume localization methods such as PRESS and STEAM, 

and suppression sequences for unwanted resonances such as 

CHESS and outer-volume saturation. The basic principles of 

the most widely implemented MRS method, single-voxel 1H 

spectroscopy, will be covered in detail, with special attention 

to real-world issues such as water and lipid suppression, 

magnetic field homogeneity, sensitivity, metabolite visibility 

and relaxation times. Approaches to spectral quantitation 

will be introduced including model and basis spectrum fitting 

in the time and frequency domain. An overview of 

MRS-specific artifacts and their minimization will be presented. 

Methods for generating magnetic resonance spectroscopic 

imaging (MRSI) including echo-planar and spiral 

sequences will be reviewed. Basic signal processing and 

spectral quantitation of 1H MRS and MRSI data will be 

described. Finally, an overview of MRS using other biologically 

relevant nuclei, such as 13C and 31P, will be presented 

with examples relevant to neuroradiology. The presentation 

will be illustrated with relevant examples from both research 

and clinical magnetic resonance spectroscopy at low and 

high magnetic fields. 

REFERENCES 

1. Ross B, Bluml S. Magnetic resonance spectroscopy of the 

human brain. Anatomical Record, 2001;265:54-84 

2. Gadian DG. NMR and Its Applications to Living Systems, 2nd 

ed. Oxford, United Kingdom: Oxford Science Publications, 

1995 

179 

MR Spectroscopy: Adult Applications 




able to: 

1) Understand the typical MR spectra of brain tumors. 

2) Understand the caveats of MRS in the evaluation of brain 

tumors. 

3) Be familiar with the role of MRS in some neurodegenerative 

disorders. 


This presentation will concentrate on the use of MR spectroscopy 

(MRS) in the evaluation of lesions involving the 

adult brain. The role of MRS in the assessment of primary 

and secondary brain tumors will be discussed. Although the 

MR spectra of brain tumors is well known it is clear that 

some lesions of lower histologic grade present spectra that 

are typical but less familiar including elevation of creatine, 

myoinositol and macrolipids/lactate while the typical elevation 

of choline is not present. There is early evidence that 

visualization of polyunsaturated fatty acids on MRS may be 

a marker of cellular apoptosis. MR spectroscopy changes 

also may detect evolving malignancies and histologic grade 

transformations. The utility of examining the surrounding 

edema and neighboring normal-appearing tissues with 3D 

techniques will be emphasized. Presence of elevated choline 

in these regions may help to differentiate a primary malignant 

brain tumor from those of lower malignancy, metastases 

and tumefactive demyelinating lesions. Additionally, the 

concomitant use of other techniques such as perfusion MR 

imaging aid as a means of sampling the regions of tumor that 

are the most representative of their histology. Color metabolite 

maps also provide information regarding biopsy target 

regions that may yield the most representative histology. 

Lastly, the potential use of MRS in some adult neurodegenerative 

disorders such as dementias and normal pressure 

hydrocephalus and the relationship of metabolites such as 

myoinositol and lactate to these entities will be addressed. 

MR Spectroscopy: Pediatric Applications 

Robert A. Zimmerman, MD 



able to: 

1) Learn that abnormal cerebral metabolism, in which lactate 

is produced, is an indication for brain proton spectroscopy. 

2) Learn that measuring elevated choline levels relative to 

other metabolites is a method of gauging how malignant a 

pediatric brain tumor is. 

3) Learn that certain metabolic disorders affecting the brain 

can be characterized by elevated levels of metabolites, not 

normally seen in the brain, and thus their response to therapy 

can be monitored. 


Proton spectroscopy is utilized to provide additional information 

when MR imaging is positive. The purpose of this is 

to add specificity when MR imaging is not positive. This has 

Wednesday

Wednesday 

been particularly valuable in metabolic diseases, leukodystrophies, 

brain tumors, and in hypoxic ischemic 

encephalopathy of the newborn. Proton spectroscopy also is 

valuable to identify abnormalities when the MR imaging is 

negative. MRspectroscopy is abnormal under the following 

circumstances. Metabolites not normally detected in the 

brain accumulate to such a degree that they are seen. The 

second is when the concentration of normal metabolites is 

either increased or decreased beyond its normal values. An 

example of an abnormal metabolite that accumulates to such 

a degree that it can be seen readily is that of galactitol at 3.67 

and 3.75 ppm found in untreated newborns and in older children 

with galactosemia. An example of a normal metabolite 

that is found in excessive amounts is Naa at 2.0 ppm, in the 

leukodystrophy Canavan’s disease, where the enzyme is 

deficient. Technical considerations include the choice of 

voxel(s), their positioning and the TE for the sequence. 

Single voxel of a sufficiently small size works for focal 

lesions, such as small tumors, hippocampal abnormalities, 

etc. Multivoxel techniques, either 2D CSI or 3D CSI, allow 

greater coverage of white and gray matter and provide information 

regarding regional or focal abnormalities such as in 

metabolic diseases where regional patterns of injury may 

help to characterize the disease; Leigh’s disease with injury 

to the putamen and globus pallidus being one example. 

Longer TEs such as 135-144 msec are useful because of Jcoupling 

effects that invert lactate at 1.3 ppm and alanine at 

1.4 ppm allowing their separation from lipid and other 

metabolites that do not invert with J coupling at the same 

ppm. Shorter echo times (e.g., 20 msec) give many more 

metabolites than at TE 135 msec, but have a more difficult 

base line to deal with and require more careful shimming. 

Many tumors look alike. In cerebellar astrocytomas, the 

choline is elevated and the NAA is decreased. A ratio of 

choline to NAA is approximately twice normal in astrocytomas. 

Lactate can be elevated, but is more characteristic of 

cysts and necrosis than of the type of tumor. Primitive neuroectodermal 

tumors (PNET) or medulloblastomas have 

higher elevations of choline than do astrocytomas. By doing 

extracts of 20 tumors and then comparing PNET to astrocytoma 

we found that the choline was at least twice the level in 

the PNET than the astrocytoma. In using choline to NAA, 

the more malignant the tumor the higher the ratio. By using 

a bimodal distribution of NAA to choline and creatine to 

choline we were able to separate low-grade astrocytomas 

from ependymomas from PNET. Ependymomas have more 

creatine. By taking this information along with some basic 

MR imaging data and clinical age, using a neural network, 

we could correctly histologically predict 95% of posterior 

fossa tumors. This has been useful particularly in incidental 

T2 bright lesions in the posterior fossa for identifying 

whether they are or are not tumors. We have used the same 

in evaluating supratentorial tumors. Beyond brain tumors, 

spectroscopy's major role is in inborn errors of metabolism 

in pediatric patients. N-acetylaspartate builds up in the 

brains of patients with Canavan's disease because of the 

absence of aspartoacylase. These children have large brains 

with absent myelination. The highest incidence is in 

European Jews and Saudi Arabians. The role of spectroscopy 

is to diagnose Canavan's, follow the course of treatment, recognize 

other diseases that might mimic Canavan's and evaluate 

the degree of response to genetic therapy. When a 

Canavan's patient study is compared to a control, it shows 

elevated NAA and myoinositol and depressed levels of 

choline and creatine. Canavan's is suitable for genetic inter- 

180 

vention because of the loss of function, recessive mutation, 

the brain is the only site of involvement, there is no known 

treatment, while progression is unrelenting and the outcome 

is uniformly fatal. Together with a multinational group we 

have utilized a lysosome polymer DNA complex for replacing 

the missing enzyme. The results show increased myelination 

posttreatment and increased function, but ultimately a 

poor outcome. Patients with pseudo Canavan’s have a 

decreased level of NAA, inheritance is on the male side, it is 

a dominant and there is a greater degree of cerebellar 

involvement that produces hydrocephalus; however, these 

patients live, grow up, and reproduce. Alexander's disease, 

another cause of macrocephaly and white matter abnormality, 

does not have elevated NAA but decreased NAA and during 

its active breakdown phase has lactate within the brain 

tissue. In summary, spectroscopy is moving from 1.5 T to 3 

T. This will decrease acquisition times and increase resolution, 

allowing smaller voxels and multiple voxels to be 

obtained. What makes spectroscopy interesting, is the ability 

to provide a biochemical understanding of the images that 

help to characterize the tissue and increase our diagnostic 

capability, while, at the same time, providing a management 

tool for understanding whether our therapy is or is not effective. 

REFERENCES 

1. Arle JE, Morriss C, Wang ZJ, et al. Prediction of posterior 

fossa tumor type in children by means of magnetic resonance 

image properties, spectroscopy, and neural networks. J 

Neurosurg 1997;86:755-761 

2. Girard N, Wang ZJ, Erbetta A, et al. Prognostic value of proton 

MR spectroscopy of cerebral hemisphere tumors in children. 

Neuroradiology 1998;40(2):121-125 

3. Gonen O, Wang ZJ, Viswanathan AK, et al. Three-dimensional 

multivoxel proton MR spectroscopy of the brain in children 

with neurofibromatosis type I. AJNR Am J Neuroradiol 

1999;20:1333-1341 

4. Wang Z, Sutton LN, Cnaan A, et al. Proton magnetic resonance 

spectroscopy of pediatric cerebellar tumors. AJNR Am J 

Neuroradiol 1995;16:1821-1833 

5. Wang ZJ, Zimmerman RA. The value of proton MR spectroscopy 

in pediatric metabolic brain disease. AJNR Am J 

Neuroradiol 1997;18:1872-1879 


5:00 PM - 6:00 PM 

Room 103 


(NLM) Workshop: PubMed ® /Medline 

Plus Advanced Tips and Tricks 

— Linda Milgrom


5:40 PM - 6:10 PM 

Room 205 

(58) ELC Lecture G: PDAs and the 

Radiologist 


181 

Wednesday

Wednesday 

Notes:

Thursday Morning 

8:00 AM - 9:00 AM 

Room 205 

(59) ELC Lecture H: Digital Teaching 

Files: Overview for Beginners 


8:00 AM - 9:30 AM 

Theatre 

(60) Updates on “PVL” (ASPNR) 

(341) Epidemiology of White Matter Damage in 

Premature and Term Infants 

⎯ Alan Leviton, MD, MS 

(342) Neuropathology of White Matter Damage in 

Premature and Term Infants 

⎯ Floyd H. Gilles, MD 

(343) Neuroimaging of Perinatal White Matter 

Injury: Past, Present, and Future 

⎯ Ashok Panigrahy, MD 

Moderators: Gary L. Hedlund, DO 

Kevin R. Moore, MD 

⎯ Gregory L. Katzman, MD 

Epidemiology of White Matter Damage in Premature 

and Term Infants 

Alan Leviton, MD, MS 

Clinically trained in internal medicine, infectious diseases, 

and neurology, Alan Leviton also has research training in 

epidemiology. The epidemiology of white matter damage in 

183 





newborns has been one of his passions for 35 years. He has 

been at Children’s Hospital, Boston since he arrived in 1969 

to study neuropathology with Floyd Gilles. 



able to: 

1) Identify the relative contribution of prenatal and postnatal 

factors to the development of image identified white matter 

damage in the preterm newborn. 

2) Evaluate the evidence in support of ischemic and inflammatory 

hypotheses of white matter damage in the newborn. 

3) Describe the later clinical manifestations of white matter 

damage imaged in the neonatal intensive care unit. 


The "ischemic" hypothesis of neonatal white matter damage 

(NWMD) pathogenesis postulates that the vulnerable 

preterm newborn has an incompletely developed vascular 

supply to the cerebral white matter, and a maturationdependent 

impairment in regulation of cerebral blood flow 

(1). The "inflammation" hypothesis postulates that the 

inflammatory phenomena that lead to, or accompany, 

preterm labor and prelabor preterm rupture of membranes 

contribute to the white matter damage (2). Both acknowledge 

the vulnerability of the oligodendrocyte precursor, 

which is prevalent in the preterm brain. Additional aspects of 

vulnerability appear to be related to the inability of the fetus 

to make substances that protect the already vulnerable oligodendrocyte 

precursor. For example, infants with low thyroxine 

levels shortly after birth are at increased risk of white 

matter damage and cerebral palsy (3), just as are infants who 

were not exposed to a course of antenatal corticosteroids. 

The ischemic hypothesis would be supported by findings 

that infants with the lowest blood pressures (or such correlates 

as increased need for volume expanders, etc.) were at 

increased risk of NWMD. Only one of 15 studies that examined 

this, however, has found an association between systemic 

hypotension and white-matter damage (4, 5). On the 

other hand, meta-analyses document that the risk of NWMD 

is associated with placenta inflammation (6), especially if 

the chorionic plate or umbilical cord is involved (7). Other 

studies show an association with biomarkers of inflammation 

in amniotic fluid (2, 7). These findings will be synthesized 

into a cohesive picture that explains why some infants 

are at increased risk of NWMD. 

REFERENCES 

1. Volpe JJ. Neurobiology of periventricular leukomalacia in 

the premature infant. Pediatr Res 2001;50:553-562 

2. Dammann O, Leviton A. Infection remote from the brain, 

neonatal white matter damage, and cerebral palsy in the 

preterm infant. Semin Pediatr Neurol 1998;5:190-201 

3. Dammann O, Leviton A. Brain damage in preterm newborns: 

might enhancement of developmentally regulated endogenous 

protection open a door for prevention? Pediatrics 

1999;104:541-550 

Thursday

Thursday 

4. Dammann O, Allred EN, Kuban KC, et al. Systemic hypotension 

and white-matter damage in preterm infants. Dev Med 

Child Neurol 2002;44:82-90 

5. Martens SE, Rijken M, Stoelhorst GM, et al. Leiden Follow-Up 

Project on Prematurity, The Netherlands. Is hypotension a 

major risk factor for neurological morbidity at term age in 

very preterm infants? Early Hum Dev 2003;75:79-89 

6. Wu YW, Colford JM Jr. Chorioamnionitis as a risk factor for 

cerebral palsy: A meta-analysis. JAMA 2000;284:1417-1424 

7. Dammann O, Leviton A. Role of the fetus in perinatal infection 

and neonatal brain damage. Curr Opin Pediatr 

2000;12:99-104 

Neuropathology of White Matter Damage in Premature 

and Term Infants 

Floyd H. Gilles, MD 



able to: 

1) Understand the pathology and history of focal necroses in 

white matter. 

2) Understand some of the risk factors of focal necroses. 

3) Learn the pathology of diffuse white matter astrocytosis. 

4) Understand some of the risk factors of astrocytosis. 


Two histologic components of neonatal white matter damage 

are focal necroses and widely spread hypertrophic astrocytes. 

The first description of focal necroses in 1850 was followed 

by seven studies from 1862-1904 (1). All described 

multiple small necroses scattered from ventricle to gyral 

white matter. Uncontrolled studies identified multiple putative 

associations (e.g., omphalitis, hyperbilirubinemia, 

hypoxia, hypotension). Their distribution does not support a 

borderzone hypothesis. Measured clinical associations identified 

sepsis as a significant antecedent (2) and endotoxemia 

was confirmed as an antecedent in neonatal animals (3). 

Most putative associations are systemic (e.g., endotoxemia, 

elevated cytokines). However, systemic abnormalities do not 

account for their focality. During the second half of gestation, 

telencephalic vascular channels consist only of 

endothelium-lined sinusoids extending from pia to ventricular 

wall (4, 5). Focal necroses are better explained perhaps 

by a hematogenous component that attaches to endothelium 

and focally damages surrounding white matter. Hypertrophic 

astrocytes are not found in myelinating or premyelination 

hemispheral white matter. White matter containing diffuse 

hypertrophic astrocytes is about three-fold as frequent as 

focal necroses and clusters separately from focal necroses 

(6). Both abnormalities may be found together but have differing 

risk factor profiles. The long-term effect of diffuse 

white matter astrocytosis is widespread fibrillary gliosis 

without gray matter abnormality, as recognized by multiple 

authors from the middle of the 20th century, and is associated 

with hypoplasia of hemispheral white matter or with 

delayed myelination. 

REFERENCES 

1. Schmorl, CG. Zur kenntniss des ikterus neonatorum, 

inbesondere der dabei auftretenden gehirnveranderungen. 

Verhandl deutsch Path Gesellsch 1904;6:109-115 

184 

2. Leviton, A, Gilles, FH. An epidemiologic study of perinatal 

telencephalic leucoencephalopathy. J Neurol Sci 1973;18:53- 

66 

3. Gilles, FH, Leviton, A, Kerr, CS. Endotoxin leukoencephalopathy 

in the telencephalon of the newborn kitten. J 

Neurol Sci 1976;27:183-191 

4. Kuban, KCK, Gilles, FH. Human telencephalic angiogenesis. 

Ann Neurol 1985;17:539-548 

5. Nelson, MD, Gonzalez-Gomez, I, Gilles, FH. The search for 

human telencephalic ventriculofugal arteries. AJNR Am J 

Neuroradiol 1991;12:215-222 

6. Gilles, FH, Leviton, A Dooling, EC. Developing Human Brain: 

Growth and Epidemiologic Neuropathology. Boston: John 

Wright-PSG Publishing Co. 1983;113-117 

Neuroimaging of Perinatal White Matter Injury: Past, 

Present, and Future 

Ashok Panigrahy, MD 

Dr. Ashok Panigrahy is currently an Assistant Professor of 

Radiology at Children's Hospital Los Angeles. He received 

his medical degree from Boston University School of 

Medicine and completed a residency in Radiology and fellowship 

in Neuroradiology at UCLA Medical Center. He 

also completed a fellowship in Pediatric Neuroradiology at 

Children's Hospital Los Angeles. 



able to: 

1) Describe how the epidemiology and imaging characteristics 

of periventricular leukomalacia has changed in the last 

decade. 

2) Review the role of conventional MR imaging, diffusionweighted 

and diffusion tensor MR imaging in the characterization 

of both focal and diffuse perinatal white matter 

injury. 

3) Understand how the use of a neonatal head coil and incubator 

can improve imaging of perinatal white matter injury. 


Improvement in neonatal intensive care has resulted in 

decreased incidence of cystic periventricular leukomalacia 

(PVL) , which is the focal form of perinatal white matter 

injury. However, recent studies have demonstrated that the 

incidence of the diffuse form of perinatal white matter injury 

in the neonate has increased resulting in more cognitive and 

behavioral deficits compared to the classic forms of cerebral 

palsy. The more prevalent diffuse form of perinatal white 

matter injury is difficult to diagnose with conventional MR 

imaging. The pathogenesis of both the focal (PVL) and diffuse 

forms of perinatal white matter injury currently is 

thought to be related to interactions between maternal/fetal 

infection, cytokines and hypoxia- ischemia which results in 

both the generation of reactive oxygen specific agents 

(oxidative stress) and apoptotic oligodendrocyte cell death. 

Perinatal white matter injury can be classified by both conventional 

MR imaging and diffusion weighted/tensor imaging. 

Focal white matter injury is defined by hyperintense T1 

signal foci in the perventricular white matter (correlating 

with coagulation necrosis) with or without cavitation. 

Diffusion-weighted imaging can be used to determine the 

acuity of focal white matter injury. Diffuse white matter 

injury (acute phase) is more difficult to diagnose with con-

ventional MR imaging and can be identified using diffusionweighted 

imaging (decreased ADC). Diffuse white matter 

injury (chronic phase) can be defined by ventriclomegaly 

and thinning of the corpus callosum (conventional imaging) 

and both increased ADC and reduced fractional anisotropy in 

the periventricular white matter. “Normal” age-dependent 

curves of quantitative ADC and FA measurements are sometimes 

needed to make assessment of more subtle diffuse 

periventricular white matter injury. Quantitative short echo 

proton spectroscopy of the periventricular white matter also 

can help with characterization and timing of perinatal white 

matter injury with respect to not only lactate, but also glutamate/glutamine 

and myo-inositol . “Normal” age-dependent 

quantitative curves of metabolite concentration are needed to 

properly interpret data. Future research tools include using 

both arterial spin labeling perfusion imaging and functional 

MR imaging to further characterize both focal and diffuse 

perinatal white matter injury. Future work also is geared 

toward understanding the long-term neurodevelopmental 

sequelae of these MR abnormalities and correlating abnormal 

MR parameters with clinical markers (cytokine levels). 

REFERENCES 

1. Volpe JJ. Cerebral white matter injury of the premature 

infant-more common than you think. Pediatrics 

2003;112:176-180 

2. Leviton A, Gilles FH. Ventriculomegaly, delayed myelination, 

white matter hypoplasia, and “periventricular” leukomalacia. 

How are they related? Ped Neurol 1996;15:127-136 

3. Miller SP, Vigneron DB, Henry RG, Bohland MA, Ceppi-Cozzio 

C, Hoffman C, Newton N, et al. Serial quantitative diffusion 

tensor MRI of the premature brain: development in newborns 

with and without injury. J Magn Reson Imag 

2002;16(6):621-632 

4. Panigrahy A, Barnes PD, Robertson RL, Back SA, Sleeper LA, 

Sayre JW, Kinney HC, et al. Volumetric brain differences in 

children with periventricular T2-signal hyperintensities: a 

grouping by gestational age at birth. AJR Am J Roentgenol 

2001;177(3):695-702 

5. Bluml S, Friedlich P, Erberich S, Wood JC, Seri I, Nelson MD. 

MR imaging of newborns by using an MR-compatible incubator 

with integrated radiofrequency coils: initial experience. 

Radiology 2004;231:594-601 

185 


8:00 AM - 9:30 AM 

Room 105/106 

(61) CT vs MR in the Spine (ASSR) 

(344) Multislice CT: New Avenues in Spinal 

Imaging 

— Alyssa T. Watanabe, MD 

(345) CT Angiography in the Spine 

— Jeffrey S. Ross, MD 

(346) MR Angiography in the Spine: Latest 

Techniques 

— Walter H. Backes, PhD 

Moderators: Jeffrey S. Ross, MD 

Brian C. Bowen, MD, PhD 

Multislice CT: New Avenues in Spinal Imaging 




able to: 

1) Discuss benefits of multislice CT in the spine. 

2) Review technological advances in multislice CT. 


Dramatic technological improvements in computed tomography 

(CT) have occurred in the past few years. Multislice 

CT now permits higher resolution and more rapid scanning 

compared with earlier generation scanners. Multislice CT 

also allows better evaluation of patients with implanted hardware 

due to reduced metallic artifacts. In addition, exquisite 

image reformations can be obtained in multiple planes without 

any significant loss in image quality due to the thinner 

slice acquisitions and near isotropic, volumetric imaging 

techniques. Multislice CT is becoming the standard of care 

in imaging high-risk cervical trauma patients, supplanting 

cervical spine X-rays in initial work up in the emergency 

department. Three-dimensional CT techniques can help 

define complex pathology and assist in surgical planning. 

Applications and limitations of this technology in the spine 


Thursday

Thursday 

CT Angiography in the Spine 

Jeffrey S. Ross, MD 



able to: 

1) Define the role of CT angiography in the evaluation of 

spinal vascular malformations. 

2) Define parameters for utilizing CT angiography in the 

spine. 

3) Define the limitations of CT angiography in the work up 

of vascular spinal lesions. 


Evaluation of complex spinal vascular malformations 

includes MR imaging, contrast-enhanced MR angiography, 

and spinal catheter angiography. With multidetector CT, CT 

angiography is also a component of the imaging armamentarium. 

While various nomenclatures exist for vascular malformations, 

they will be considered as Type I (AV fistula), II 

(intramedullary glomus malformations), III (extensive juvenile 

malformation) and IV (intradural perimedullary fistula). 

CT angiography utilizing 16+ detector arrays is capable of 

imaging the cervical through sacral regions in one pass, with 

scan times under 40 seconds. 0.7 mm collimation with 1 mm 

thick reconstructions should be utilized, and 75-100 cc of 

contrast injected at 4 cc per second. Triggering off the aortic 

arch is adequate with a 5 sec delay. Postprocessing is mandatory, 

with the ability to do multiplanar orthogonal and curved 

reformats. Evaluation of type I fistulas requires meticulous 

evaluation of the neural foramina, looking for subtle vascular 

prominence reflecting the peripheral nidus. Primary 

intramedullary AVMs are identified easily due to the focal 

serpentine vessels. Precise utilization of multiplanar reformats 

can trace draining vessels. Disadvantages of CTA 

include the lack of dynamic phase, timing problems if there 

is diminished cardiac output, and lack of resolution for small 

normal vessels such as anterior spinal artery. 

MR Angiography in the Spine: Latest Techniques 

Walter H. Backes, PhD 

Dr. Walter Backes is a Medical Physicist in the field of radiologic 

imaging and is registered by the Dutch Society of 

Medical Physicists. Since 1998 he is employed as a principal 

investigator at the radiology department of the Maastricht 

University Hospital and is also an associate professor and lecturer 

in neuroradiologic MR imaging at the Eindhoven 

University of Technology, both in The Netherlands. In the past 

5 years he has written more than 25 articles which either deal 

with functional MR imaging or small vessel angiography. 



able to: 

1) Discuss the difficulties of imaging intradural arteries and 

veins of the spinal cord. 

2) Utilize the optimized technique of gadolinium-enhanced 

MRA applied to the spine. 

3) Recognize the possibilities and limitations for clinical 

applications. 

186 


Imaging the ultra-small spinal arteries and their blood supplying 

pathways is one of the remaining challenges to MR 

angiography. At present selective spinal cord digital subtraction 

angiography (DSA) is the method of choice to image the 

arteries and veins of the spinal cord. Contrast-enhanced MR 

angiography (CE MRA) however, featuring intravenous 

injection of gadolinium-based contrast medium, shows great 

promise and improvements for imaging spinal cord vessels. 

Spinal arteries are usually smaller in caliber than their 

accompanying veins but have similar morphology. Therefore 

the only noninvasive way in which a reliable differentiation 

of spinal cord arteries and veins can be expected, is an angiographic 

technique both with sufficient spatial resolution and 

a time resolution that clearly separates arterial and venous 

phases. The current possibilities and limitations will be 

addressed of high-dose CE MRA to image spinal arteries and 

blood suppliers of the thoracolumbar region of the spinal 

cord. To this end, we will first elaborate on the details (arterial 

concentration, k-space filling, large field of view, and 

bolus arrival timing) of a successful CE MRA method, show 

its results in representative cases and elaborate on the mechanism 

to separate arterial from venous phase. Recent optimizations 

of CE MRA allow the visualization of 1) the anterior 

spinal artery and the Adamkiewicz artery and its connections 

the aorta and 2) the supplying radicular artery of 

dural arteriovenous fistulae, which are validated with DSA 

in patients with and without vascular abnormalities. In 

patients with thoracoabdominal aorta aneurysms (TAAA) 

identification of the Adamkiewicz artery may be of importance 

to guide and adapt surgical decisions in order to preserve 

the spinal cord blood supply. Examples will be shown 

how TAAA surgery benefits from presurgical MR imaging 

information of the blood supplying territories that are crucial 

to prevent spinal cord ischemia. 

REFERENCES 

1. Pattany PM, Saraf-Lavi E, Bowen BC. MR Angiography of the 

spine and spinal cord. Top Magn Reson Imag 2003;23:444-460 

2. Nijenhuis RJ, Leiner T, Cornips E, Wilmink JT, Jacobs MJ, Van 

Engelshoven JMA, Backes WH. MR angiography of the spinal 

cord feeding arteries in thoracoscopic spinal surgery: a feasibility 

study and implications for surgical approach. 

Radiology 2004;233:541-547 

3. Yamada N, Takamiya M, Kuribayashi S, Okita Y, Minatoya K, 

Tanaka R. MRA of the Adamkiewicz artery: a preoperative 

study for thoracic aortic aneurysm. J Comput Assist Tomogr 

2000;24:362-368 

4. Farb RI, Kim JK, Willinsky RA, Montanera WJ, Ter Brugge K, 

Derbyshire JA, Van Dijk JMC, et al. Spinal dural arteriovenous 

fistula localization with a technique of first-pass 

gadolinium-enhanced MR angiography: initial experience. 

Radiology 2002;222:843-850


8:00 AM - 9:30 AM 

Room 107 

(62) Brain Tumors (ASFNR) 

(347) Imaging of Angiogenesis 

— James M. Provenzale, MD 

(348) Functional Diffusion Mapping (fDM): A 

Novel Approach for Prediction of Therapeutic 

Outcome in Brain Tumors 

— Brian D. Ross, MD 

(349) Mouse Models of Brain Tumors 

— Eric C. Holland, MD 

Moderator: William G. Bradley, MD, FACR, MBA 

Imaging of Angiogenesis 




able to: 

1) Explain the physiologic principles underlying angiogenesis. 

2) Show that a number of angiogenesis factors are potential 

targets for imaging and therapy. 

3) Depict methods for imaging changes in vessel growth and 

vessel permeability. 


The development of novel therapies for treatment of neoplasms 

has led to the need for a new role for the radiologist. 

These new therapies, many of which are designed to alter the 

genetic structure or physiology of tumor cells, call for new 

ways to evaluate tumors. Previously, it was sufficient for 

radiologists to simply answer the question: “How much did 

the tumor change in size?” This simple approach is no longer 

valuable for a number of reasons: (1) many physiologic 

changes occur before tumors change in size, (2) physiologic 

changes can occur independently of tumor size changes, (3) 

the methods for studying physiologic changes are very different 

from those designed to studying changes in size. The 

purpose of this lecture is to provide information regarding 

physiologic imaging of brain tumors, with specific reference 

to evaluation of antiangiogenesis agents. Almost all of the 

principles outlined in this presentation can be generalized to 

non-CNS tumors and novel agents other than antiangiogenesis 

therapies. Angiogenesis refers to the process of generation 

of new vessels from preexisting vessels. In the past three 

187 

decades the critical importance of angiogenesis in tumor 

growth has been well studied and delineated. Angiogenesis 

is a prerequisite for tumor growth; without angiogenesis, 

tumor growth is limited markedly. Furthermore, effective 

measures to retard angiogenesis decrease the rate of tumor 

growth. Angiogenesis is controlled by a number of factors 

including naturally occurring proteins such as vascular 

endothelial growth factor (VEGF), epidermal growth factor, 

and many others. Many therapies are designed to decrease 

the local levels of these factors or to block their effects. A 

wide variety of physiologic forms of imaging of brain 

tumors are available or being developed. In this presentation, 

a number of models for imaging angiogenesis in animal and 

human models will be discussed. One of the important 

hemodynamic parameters to be studied in evaluating antiangiogenesis 

therapies is vessel permeability due to the fact 

that one of the major angiogenesis factors, VEGF, is a very 

potent permeability promoter. Therefore, agents that inhibit 

effects of VEGF theoretically should decrease permeability. 

In fact, trials of a VEGF inhibitor at our institution have 

found such decreases. Other therapeutic targets, such as epidermal 

growth factor receptor (EGFR) also will be discussed. 

In this presentation, various models for assessing 

permeability using standard contrast agents and novel agents 

(such as liposomes) will be discussed. One of the other 

major hemodynamic parameters is relative cerebral blood 

volume (rCBV) which is effectively a measure of capillary 

density. This parameter can be measured using T2*-weighted 

techniques (either gradient-echo or spin-echo techniques). 

The role of rCBV measurements in brain tumor imaging will 

be discussed and results of rCBV measurements in a VEGF 

inhibitor trial will be present. 

Functional Diffusion Mapping (fDM): A Novel Approach 

for Prediction of Therapeutic Outcome in Brain Tumors 

Brian D. Ross, MD 

Dr. Brian D. Ross is currently a Professor of Radiology and 

Biological Chemistry at the University of Michigan School 

of Medicine and Co-Director of the Center for Molecular 

Imaging. He received his PhD in Biophysics from the 

University of California at Davis and completed an ABTA 

brain tumor fellowship at the University of Minnesota. He is 

Editor of the international cancer journal Neoplasia and 

Executive Editor of the official journal of the Society for 

Molecular Imaging for which he is also one of the founding 

members and currently holds the position of President-Elect. 

He has authored over 70 articles and 4 book chapters and 

has been a leader in the application of diffusion MR imaging 

for cancer treatment monitoring. 



able to: 

1) Demonstrate the use of functional diffusion mapping as an 

early surrogate marker for treatment response. 

2) Summarize the findings on applying diffusion MR imaging 

for cancer treatment prediction. 


Assessment of the efficacy of radiation and chemotherapy 

for brain cancer patients is accomplished traditionally by 

Thursday

Thursday 

measuring changes in tumor size weeks after therapy has 

been completed. The ability to use noninvasive imaging during 

early stages of fractionated therapy to determine if a particular 

treatment will be effective would provide an opportunity 

to optimize individual patient management and avoid 

unnecessary systemic toxicity and expense. We investigated 

whether analysis of diffusion MR imaging data of tumor tissue 

could be used to provide a surrogate marker for early 

prediction of treatment response in brain cancer patients. 

Brain tumor patients were examined by standard and diffusion 

MR imaging before initiation of treatment. Additional 

images were acquired at 3 weeks postinitiation of chemo 

and/or radiotherapy. Images were coregistered to pretreatment 

scans, and tumor water diffusion values were calculated 

using functional diffusion maps (fDM) and correlated 

with subsequent response defined by change in tumor size on 

MR imaging by standard radiographic criteria. Changes in 

tumor diffusion values were found to predict patient 

response at 3 weeks from the start of treatment revealing that 

early changes in diffusion values could be used as a prognostic 

indicator of subsequent volumetric tumor response. 

Diffusion MR imaging provides an early surrogate marker 

for predicting treatment response in patients with brain 

tumors. 

Mouse Models of Brain Tumors 

Eric C. Holland, MD 


8:00 AM - 9:30 AM 

Room 103 

(63) ELC Workshop D: Adobe 

Photoshop and Elements 

— Richard M. Berger, MD 

188 


10:00 AM - 11:36 AM 

Room 105/106 

(64a) SPINAL CORD AND 

PERIPHERAL NERVES: Functional 

and Advanced Imaging Techniques 



(64b) ADULT BRAIN: Metabolic and Miscellaneous 

(64c) ADULT BRAIN: Trauma and Vascular Lesions 

(64d) Great Cases and Excerptas 

Moderators: Michael I. Rothman, MD 

Alan L. William, MD, FACR, MBA 


Cervical Spinal Cord 3 T Diffusion-Weighted MR 

Imaging 

Wang, P. Y. 1 · Wilson, G. J. 2 · Tjauw, I. 1 

1 Oregon Health and Science University, Portland, OR, 

2 Philips Medical Systems, Seattle, WA 

PURPOSE 

Diffusion-weighted MR imaging may be of value in evaluating 

cervical spinal cord pathologies (infarction, demyelination/inflammation, 

tumor) but has not been a routine 1.5 T 

clinical sequence due to low SNR and susceptibility image 

degradation. We compared cervical spinal cord echo-planar 

diffusion-weighted imaging with and without sensitivity 

encoding (SENSE) to determine those appropriate to attain 

clinically useful images. With SENSE multicoils, k-space 

encoding is reduced, resulting in shortened TE, thereby 

increasing SNR and reducing susceptibility artifacts. Similar 

results are achieved by multishot EPI also allows partial kspatial 

encoding during each TR. 


Sagittal and axial echo-planar (EPI) diffusion-weighted imaging 

of the cervical spinal cord was performed (b-value 600 

s/mm2) on a Philips Intera 3 T MR imaging unit. Separate 

phase, frequency, and slice diffusion gradient direction diffusion-weighted 

images were created in addition to standard 

isotropic diffusion-weighted images. Apparent diffusion coefficient 

(ADC) values also were measured. Single-shot EPI was 

performed with SENSE factor 2, TR/TE 3 s/84 ms, 6 NEX, partial 

fourier acquisition of 0.67, with 1.25 x 1.25 x 3 mm sections. 

Scan time was 2:12. Multishot EPI diffusion-weighted 

imaging with echo train length 7 which was performed without 

SENSE with peripheral pulse gating, navigator-phase correction, 

TR 2 heartbeats, TE 80 ms, and b-value 600 s/mm2. Scan 

time was 2:38. Fifteen cases were reviewed by 3 radiologists.

RESULTS 

Sagittal diffusion-weighted imaging was found to be best 

with multishot EPI yielding undistorted images, whereas single-shot 

SENSE EPI was degraded by susceptibility distortion. 

In contradistinction, axial diffusion-weighted imaging 

was found to be superior with single-shot SENSE EPI with 

higher SNR and less geometric distortion compared to multishot 

EPI. Anisotropy of cord diffusion was seen clearly 

between in-plane (ADC 0.4 x 10-8 mm2/s) as opposed to 

through-plane (ADC 1.7 x 10-8 mm2/s) diffusion gradient 

images. Artifacts encountered were most frequently in the 

sagittal plane due to patient motion (though distorted singleshot 

diffusion-weighted images were still of diagnostic 

value) and incomplete fat-suppression (2/15). Cases of MS 

demyelination, and cord infarction and tumor will be presented. 

CONCLUSION 

Artifacts in spinal cord diffusion-weighted imaging at 3 T 

are due mainly to motion and susceptibility distortion and 

still limit this technique. Diagnostic spinal cord diffusionweighted 

imaging was successful with sagittal multishot EPI 

and axial single-shot SENSE EPI within 2-3 minutes. 

However in cases of moving patients, single short SENSE 

EPI in the sagittal plane yielded diagnostic though esthetically 

distorted images. Ongoing work is to characterize and 

address the source of artifacts. 

REFERENCES 

1. Cercignani M, et al. Sensitivity-encoded diffusion tensor MR 

imaging of the cervical cord. AJNR Am J Neuroradiol 

2003;24:1254-1256 

2. Demir A, et al. Diffusion-weighted MR imaging with apparent 

diffusion coefficient and apparent diffusion tensor maps 

in cervical spondylotic myelopathy. Radiology 

2003;229(1):37-43 

KEY WORDS: Spinal cord, diffusion-weighted imaging, 3 T 


Diffusion-Weighted MR Imaging of the Spinal Cord in 

Multiple Sclerosis 

Dirisamer, A. 1 · Bammer, R. 2 · Thurnher, M. M. 1 

1 2 University Hospital Vienna, Vienna, AUSTRIA, Lucas 

MRS/I Center, Stanford University, Stanford, CA 

PURPOSE 

The purpose of the study was to evaluate diffusion characteristics 

of multiple sclerosis (MS) lesions in the spinal cord, 

and compare the results with findings on conventional MR 

sequences. 


Nine patients with proven MS and lesions in the spinal cord 

were included in the study. MR studies were performed on 

1.5 T clinical scanner with following imaging protocol: T2weighted, 

T1-weighted pre and postcontrast, and diffusionweighted 

imaging (DWI) in sagittal and axial planes. For 

DWI multi-shot IEPI with fat suppression and max b factor 

of 600 was used. Slice thickness was 5 mm, fold-over direction 

AP and scanning time was 3:05 min. For axial images 

performed with a surface coil (14 x 17 cm) single-shot EPI 

with 40 slices and 4 mm slice thickness was used. Signal 

189 

intensities on DWI were recorded and ADC measurements 

were performed in all patients. 

RESULTS 

All MS lesions showed high signal on T2-weighted imaging, 

marked enhancement was present in 3/9 patients, faint 

enhancement was observed in 3/9 patients. In 3 patients no 

enhancement was present on postcontrast scans. In 8 of 9 

patients, high signal was observed on DWI with low ADC 

values consistent with restricted diffusion. One patient did 

not show diffusion abnormality despite present enhancement 

on postcontrast image. 

Fig 1A. Sagittal T2-weighted MR imaging shows high signal 

intensity lesion in cervical spinal cord with edema. 

Thursday

Thursday 

Fig 1B. On axial diffusion-weighted MR image high signal 

was shown in dorsal part of cord indicating restricted diffusion 

in MS plaque. Lesion was nonenhancing on postcontrast 

T1-weighted MR images. 

CONCLUSION 

Diffusion-weighted imaging may provide new structural 

information in relation to spinal cord pathology in MS. 

Further work with comparison between DWI/ADC and signal 

intensities on T1-weighted imaging, presence or absence 

of enhancement, MTR is needed. 

KEY WORDS: Multiple sclerosis, spinal cord, diffusionweighted 

imaging 


MR Spectroscopy of the Healthy and Tumorous Cervical 

Spinal Cord 

Dydak, U. 1 · Kollias, S. S. 2 · Boesiger, P. 1 

1 University and ETH Zurich, Zurich, SWITZERLAND, 

2 University Hospital Zurich, Zurich, SWITZERLAND 

PURPOSE 

Proton MR spectroscopy shows a great potential to noninvasively 

add valuable diagnostic information in the evaluation 

of spinal cord lesions, just as has been shown in the brain. 

However MRS of the spinal cord is particularly challenging 

due to its small size, field inhomogeneities, and physiologic 

motion. Cooke et al. (1) have demonstrated how good quality 

spectra can be obtained in the healthy cerebral spinal cord 

at 2 T. In this work we show spectra acquired at 3 T from a 

primary spinal cord tumor. 


Single voxel spectroscopy was performed in the cerebral 

spinal cord of two healthy volunteers and in a patient with a 

primary tumor of the spinal cord. Measurements were carried 

out on a 3 T Philips Achieva system with a dedicated 

spine coil. PRESS localization with voxel sizes between 1.6 

ml and 4 ml was used. ECG triggering was applied to 

improve linewidths of the peaks, and the number of averages 

190 

ranged between 624 and 1024. Acquisition parameters 

included: TR ≥ 850 ms and TE = 43 ms, 60 ms, or 288 ms 

for lactate detection. Water peaks were shimmed to approximately 

10 Hz. All volunteers and the patient gave written 

informed consent prior to participating in the study. 

RESULTS 

Figure 1 shows the typical voxel placement (1a) and spectra 

(1b and 1c) from two healthy volunteers. Choline and creatine 

peaks show similar amplitudes and NAA is clearly the 

highest peak, similar to those published in reference 1.The 

peaks at 3.6 ppm may be attributed to myo-inositol. Figures 

1d-f show the spectra acquired from the spinal cord tumor 

(1e: TE = 60 ms, 1d: TE = 288 ms). Metabolite levels are 

generally lower than in healthy tissue and correspond to the 

spectral pattern typically seen in high-grade brain tumors, 

with low NAA level, high choline level, and increased lactate 

and lipid levels. The high peak at 3.6 ppm in Fig.1e 

might be attributed to elevated inositol levels. 

CONCLUSION 

We have shown that MRS of the cervical spinal cord is possible 

both in the healthy spinal cord as well as in tumors in 

the spinal cord. Spectral quality is limited by low sensitivity 

due to small voxel sizes and inhomogeneities along the cord. 

However spectra from spinal cord tumors show similar 

information content as spectra from brain tumors. 

REFERENCES 

1. Cooke FJ, et al. MRM 2004;51:1122-1128 

KEY WORDS: MR spectroscopy, spinal cord, tumor 


Optimized Degenerative Cervical Spine MR Imaging at 3 T 

Wang, P. Y. · Tjauw, I. · Nesbit, G. · Anderson, J. C. 

Oregon Health and Science University 

Portland, OR 

PURPOSE 

1.5 T MR imaging of cervical degenerative disk disease is 

often suboptimal, gradient-echo images often degraded by 

low tissue contrast, low signal-to-noise, motion and susceptibility 

artifacts. 3 T MR imaging with almost 2 x SNR has 

promise both low flip-angle SPGR and balanced fast-field 

echo (B-FFE/TRUEFISP/b-SSFP) 3 T sequences were optimized 

and compared for cervical MR imaging. SPGR is less 

affected by susceptibility distortion than T2* gradient echo. 


3 T axial cervical spine MR imaging (Philips Intera NT

whole body scanner) with a 12-element phase array CTL 

receive spine coil was performed varying the B-FFE flip 

angle (5/10/15 o ) and TR (32, 35, 40 ms) with fixed TE 9.2 ms 

to optimize CSF/cord and foramina visualization with acquisition 

time. T1-FFE had fixed TR/TE/FA of 4.3/1.7/45 o . Both 

sequences employed 21 cm FOV with 70 sections (0.8 x 0.8 

x 1.5 mm). Twenty-five patients were examined with both 

SPGR and B-FFE and assessed for anatomical depiction, tissue 

contrast, CSF/cord myelographic differentiation, degenerative 

pathology, and artifacts. Images were reviewed by 

three experienced neuroradiologists. 

RESULTS 

The clinical need to assess high-resolution anatomy rapidly 

and accurately of a large number of images in a short period, 

requires therefore not only high SNR images but high tissue 

contrast. T1-FFE was found to be optimal with flip angle 5 o , 

TR/TE 32/9.2 ms which required 4:00, while B-FFE 

required 3:16. Disk herniations, osteophytes, foraminal 

stenoses, and cord deformation were well seen. T1-FFE 

showed disk spaces to be bright, had much higher bone/tissue 

contrast and better contrast for fat-containing foramina. 

The vertebra including the marrow space was much darker 

than surrounding soft tissue. However the sequence showed 

more swallowing-motion artifacts requiring careful placement 

of saturation bands anterior to the region of interest. 

Being less T2* weighted there was less susceptibility artifact 

in the foramina. B-FFE had higher intrinsic tissue SNR with 

slightly better anatomical delineation (which did not result in 

a clinically relevant diference) but less bone-tissue contrast, 

requiring more time to review. The disk levels were much 

more difficult to localize. There was frequent Bo inhomogeneity 

off-resonance banding artifact always obscuring the 

surrounding soft tissues and often the spinal cord which 

would require 2 separate acquisitions. Having short TR/TE 

and being a serial slice sequential acquisition, there was less 

motion artifact sensitivity. Degenerative changes were 

equally well seen on both sequences. Metallic susceptibility 

artifacts were generally worse at 3 T than 1.5 T due to much 

shorter T2*. 

CONCLUSION 

T1-FFE is optimized with TR/TE/FlipA 32/9.2/5 rapidly 

acquired in 4 minutes yielding excellent images which are 

reviewed rapidly with better disk, bone, tissue, and foramen 

contrast. However, B-FFE is preferred in the moving subject 

but often suffers from banding artifact in 3:16 unless 

acquired in 2 separate acquisitions which is time-consuming 

to acquire and review. 

KEY WORDS: Spine, degenerative, 3 T 


Multirigid Fusion of CT and MR Images of the Spine: A 

Possible Alternative to Myelography 

Haynor, D. R. · Hu, P. · Jarvik, J. G. · Mirza, S. 


Seattle, WA 

PURPOSE 

CT myelography (CTM) still is performed because of its 

unexcelled ability to depict both bony and soft tissue (ST) 

191 

detail in and around the dural sac. We hypothesized that millimeter-accurate 

image fusion between MR and CT would 

combine the strengths of each modality (bone depiction by 

CT, ST depiction by MR imaging), reducing the need for 

CTM. To achieve the accuracy required, the spine must be 

treated as a collection of independent rigid bodies. 


We modified the level sets method widely used in image segmentation 

to allow simultaneous evolution of multiple competing 

volumes, one per vertebra. This produces satisfactory 

segmentation of the spine into individual vertebrae, even if 

severe degenerative disease is present, with only minimal 

user interaction. With no further interaction required, each 

vertebra then is separately registered with the MR data, and 

the CT-defined bones “painted” onto the MR images. We 

applied this method retrospectively to 50 consecutive cervical 

spine studies in which both CTM and MR imaging were 

performed, creating 50 sets of DICOM-compliant fusion 

images. Six neuroradiologists are comparing the diagnostic 

accuracy of the fusion images with that of CTM and MR 

imaging; we are evaluating both interobserver and intraobserver 

variability, the latter by presenting some studies in 

each category twice. Six surgeons are formulating operative 

plans based on either the fusion images or the CTM; the 

plans also are being compared with respect to inter and intrasurgeon 

variability. 

RESULTS 

Fusion was successful in all cases. Comparison of MR 

images with the fusion images makes it possible to see what 

part of foraminal or canal narrowing is due to ST and what 

to bone. A representative fusion image is attached (Fig). The 

formal evaluation process is underway and is expected to be 

complete by the end of January 2005. Preliminary experience 

suggests improved agreement between observers on the 

cause of degenerative narrowing. 

Fusion image. Vertebra from CT; rest of image from MR 

imaging. Note, for example, that MR-black tissue protruding 

posteriorly from the vertebra is not bone, and must represent 

posterior longitudinal ligament. 

CONCLUSION 

By segmenting and registering each vertebra separately, 

Thursday

Thursday 

accurate fusion of CT and MR spine images can be performed, 

and the separate contributions of bone and ST in 

causing degenerative narrowing can be distinguished. The 

method shows promise for further reducing the need for 

myelography. The effect on intra and interobserver agreement 

for diagnosis and surgical plans is currently being evaluated. 

KEY WORDS: Spine, image fusion, registration 


Predicting the Neurologic Level of Injury with MR 

Imaging following Cervical Spinal Cord Injury 

Sharma, D. K.. · Leypold, B. · Flanders, A. E. 

Thomas Jefferson University Hospital 


PURPOSE 

Spinal cord injury (SCI) produces lesions characterized by 

edema and hemorrhage whose location has been reported 

previously to roughly correspond to the neurologic level of 

injury. The purpose of this study is to determine the degree 

of variability of spinal cord lesion location on MR imaging 

relative to the neurologic level of injury (NLI) as determined 

by clinical examination. 


The degree of damage to the cervical spinal cord parenchyma 

was assessed on the MR images of 67 patients who sustained 

acute traumatic tetraplegia with a clinical assessment 

of neurologic level of injury at C4, C5, or C6. All patients 

underwent MR imaging within 72 hours of injury. Three MR 

parameters were recorded which reflected the most cephald 

extent of spinal cord hemorrhage, edema, and lesion epicenter 

relative to the nearest adjacent vertebral segment. 

Bivariate plots were created comparing each of the five MR 

imaging features to the neurologic levels of injury. Standard 

error was calculated also. 

RESULTS 

The amount of standard error was the least overall when the 

location of the lesion epicenter was compared to the NLI (C4 

- .508, C5 - .559, C6 - .653) while the standard error for the 

upper interface of edema relative to the NLI was (C4 - .748, 

C5 - .792, C6 - .789). The standard error for the anatomical 

location of the upper margin of hemorrhage was (C4 -.398, 

C5 - .706, C6 - .333). While all three locations demonstrated 

a relationship between the anatomical location and NLI, the 

relationship was more linear for lesion epicenter and upper 

level of edema than it was for the upper margin of hemorrhage. 

Standard Error for MR Imaging Measurements 

Relative to Neurologic Level of Injury 

MR imaging parameter C4 C5 C6 

Lesion Epicenter .508 .559 .653 

Upper boundary of hemorrhage .398 .706 .333 

Upper boundary of edema .748 .792 .789 

192 

CONCLUSION 

The anatomical location of the lesion epicenter and the upper 

margin of hemorrhage most closely correspond to the neurologic 

level of injury. MR imaging can be used as an objective 

method for determining the neurologic level of injury 

when the clinical examination is unreliable. 

KEY WORDS: Spinal cord injury, MR imaging, neurologic 

level 


Correlation of MR Imaging Findings with Intraoperative 

Findings after Cervical Spine Trauma 

Goradia, D. 1,2 · Linnau, K. F. 1,2 · Cohen, W. A. 1,2 · Mirza, S. 1,2 

· Hallam, D. K. 1,2 · Meshberg, E. 1 

1 University of Washington, Seattle, WA, 2 Harborview 

Medical Center, Seattle, WA 

PURPOSE 

MR imaging is thought to be superior to other modalities in 

depicting the presence of ligamentous injury of cervical 

spine. However, few data exist comparing the postulated 

high diagnostic accuracy to clinical cases. The current study 

compares acute MR imaging with surgical observations from 

open operation for internal fixation of blunt injury to the cervical 

spine. 


Utilizing the institutional spine registry, we identified retrospectively 

patients with acute traumatic cervical spine 

injuries who underwent preoperative MR imaging and subsequent 

surgical stabilization. All cases were treated between 

1998 and 2001. At surgery the extent of anatomical injury at 

the most severely injured level was recorded on a standardized 

form for either anterior or posterior structures or both 

depending upon the operative approach. MR imaging was 

obtained on a 1.5 T system within 2 days of injury. All MR 

examinations included sagittal T2-weighted sequences with 

fat saturation and/or sagittal STIR in addition to sagittal T1weighted 

sequences. A board certified neuroradiologist and a 

general radiologist retrospectively evaluated the scans blinded 

to surgical findings. Observations were decided by consensus 

based on codified criteria of injury. Structures were 

considered abnormal if they were displaced or disrupted or if 

abnormal signal was present. Radiologic findings then were 

correlated with surgical observations, which were regarded 

as standard of reference.

RESULTS 

Of 31 patients (age range 15-76 years), an anterior surgical 

approach was chosen in 17 and a posterior approach in 13 

patients. In 1 patient anterior and posterior approaches were 

utilized. Epidural hematoma was seen in 42% (13/31) of 

cases and 71% (22/31) had spinal cord injury on MR imaging 

Anterior structures: Greatest agreement between MR 

abnormalities and surgical findings was in the annulus fibrosus 

and posterior longitudinal ligament. Lesser agreement 

was noted for the anterior longitudinal ligament and for vertebral 

body fractures. Posterior structures: Overall agreement 

between MR findings and injury status at surgery was 

high for facet capsules but less for injury to posterior osseous 

elements, ligamentum flavum, and interspinous ligament. 

RESULTS SUMMARY 

Structure Number Number Total Sensitivity Specificity 

abnormal injured number (%) (%) 

at MR at surgery 

Anterior Longitudinal 13 14 18 71 25 

Ligament 

Anterior Annulus 17 15 18 93 0 

Fibrosus 

Posterior Annulus 16 13 18 92 20 

Fibrosus 

Posterior Longitudinal 16 15 18 93 33 

Ligament 

Vertebral Body 15 6 18 100 25 

Posterior Osseous 17 22 77 45 87 

Facet Capsules 21 22 26 86 50 

Ligamentum Flavum 8 6 14 67 50 

Interspinous Ligament 13 6 14 100 12.5 

CONCLUSION 

For the annulus fibrosus, posterior longitudinal ligament, 

facet capsules, and interspinous ligaments, MR findings are 

highly sensitive compared with intraoperative observations 

of injury but have poor specificity. For posterior osseous elements, 

anterior longitudinal ligament and ligamentum 

flavum, MR imaging is less accurate. Future work is needed 

to clarify the most accurate MR findings predictive of spine 

stability in order to allow operative planning based on MR 

imaging. 

KEY WORDS: Spine trauma 


Spinal Cord Diffusion Tensor Imaging and Fiber 

Tracking Can Identify White Matter Tract Disruption 

and Glial Scar Orientation following Partial Hemisection 

Schwartz, E. D. 1 · Duda, J. 1 · Cooper, E. 2 · Shumsky, J. S. 2 · 

Gee, J. C. 1 

1 2 University of Pennsylvania, Philadelphia, PA, Drexel 

University College of Medicine, Philadelphia, PA 

PURPOSE 

Diffusion tensor MR imaging (DTI) provides rotationally 

invariant data concerning water diffusion, which in spinal 

cord white matter is highly anisotropic due to the regular 

structure of axons. From this data, white matter tracts may be 

inferred and connectivity between spinal cord segments may 

be determined. 


We evaluated this potential application by imaging normal 

adult rats and rats that received cervical lateral funiculus par- 

193 

tial hemisections, disrupting the rubrospinal tract. Vitrogen 

and either unmodified fibroblasts or BDNF secreting fibroblasts 

were transplanted at time of injury. At 2 months, animals 

were sacrificed; the spinal cords were dissected out and 

then imaged in a 9.4 T magnet. 

RESULTS 

Diffusion tensor tractography (Fig. 1) demonstrated the disruption 

of the rubrospinal tract axons in the dorsal lateral 

white matter (curved arrow) while indicating preservation of 

ventral (long arrow) and dorsal (short arrow) tracts. A small 

amount of tracking is seen in the dorsal lateral white matter 

and correlated with spared white matter seen on histologic 

images. Additionally, DTI imaging could identify the orientation 

of glial processes in the gray matter adjacent to the site 

of injury. In the injured animals, the reactive astrocytes 

appear to orient themselves perpendicular to white matter 

tracts. 

CONCLUSION 

In summary, DTI identified not only white matter disruption 

following injury, but could distinguish the orientation of the 

accompanying glial scar. 

KEY WORDS: Diffusion tensor imaging, spinal cord injury 


Value of Diffusion-Weighted MR Imaging in the 

Diagnosis of Acute Cervical Cord Injury 

Tsuchiya, K. · Fujikawa, A. · Honya, K. · Nakajima, M. · 

Nitatori, T. 

Kyorin University 

Tokyo, JAPAN 

PURPOSE 

In acute head injury, diffusion-weighted imaging has an 

additional value to conventional MR imaging as it demonstrates 

diffusion abnormalities caused by some conditions 

such as axonal injury. We have effectively employed diffusion-weighted 

imaging using a single-shot fast spin-echo 

(SSFSE) sequence in spine MR examinations. This study 

was performed to analyze findings of acute cervical cord 

injury on diffusion-weighted imaging and to evaluate its 

value in predicting functional prognosis. 

Thursday

Thursday 


Our study group included 14 patients (ten men and four 

women; age range, 2-86 years; mean age: 51 years) who 

underwent MR examination at 1.5 T in the acute phase (2 

hours to 3 days after onset) that included diffusion-weighted 

imaging as well as conventional T1- and T2-weighted scans. 

Diffusion-weighted imaging was performed using an SSFSE 

sequence in the sagittal plane. Scanning parameters were as 

follows: echo time, 80-100 ms; imaging matrix; 128 x 128; 

field of view; 30 x 30 cm; and section thickness; 5 mm. 

Motion probing gradients of b = 400 s/mm2 were applied in 

three (x, y, and z) directions. First, we evaluated diffusionweighted 

imaging findings of a lesion that was hyperintense 

on T2-weighted images. Apparent diffusion coefficient maps 

were assessed also except in one patient. Second, we 

reviewed medical records to check functional outcome of the 

patients at discharge. Third, follow-up MR examination 

obtained in six patients was evaluated also. 

RESULTS 

On diffusion-weighted images, the lesion showed hyperintensity 

in eight patients and a mixture of hyperintensity and 

hypointensity in one patient. In the remaining five patients, 

no abnormal signal was noted. Hyperintense lesions that 

were demonstrated in a total of nine patients showed restricted 

diffusion on ADC maps except in one patient. At discharge, 

among the nine patients whose lesion included 

hyperintensity, six patients were in a state of tetraparesis or 

in need of assistance. However, two of the five patients with 

unremarkable diffusion-weighted findings were also in a 

similar state. Remaining patients (three and three, respectively) 

were unassisted. Follow-up MR imaging revealed 

either myelomalacia or exacerbation in six of the nine 

patients and four of them showed poor functional outcome. 

CONCLUSION 

Diffusion-weighted imaging in the acute cervical cord injury 

often shows restricted diffusion. Although it predicts unfavorable 

functional prognosis, exceptions are not rare. 

KEY WORDS: Spinal cord, injury, diffusion 


Functional Activation of the Human Spinal Cord during 

Vibration Stimulation of Different Dermatomes 

Lawrence, J. 1 · Stroman, P. W. 2 · Kollias, S. 1 

1University Hospital Zürich, Zürich, SWITZERLAND, 

2University of Manitoba, Winnipeg; Queens University 

Kingston, Kingston, MB, CANADA 

PURPOSE 

To demonstrate distinguished patterns of functional activity 

in the cervical and lumbar human spinal cord during vibration 

stimulation (50 Hz) of different dermatomes using functional 

MR imaging (fMRI) methodology. 


The cervical and lumbar spinal cords of 7 healthy human 

subjects were imaged at 1.5 T in separate experiments during 

stimulation of the right bicep, wrist, palm, patella, and 

achilles tendon. Six subjects also received stimulation to the 

left palm during cervical imaging. 

194 

RESULTS 

During stimulation of the right bicep, activity was centered 

mainly around the level of C5/C6 disk. Five of the seven 

subjects demonstrated activity in the right and left dorsal 

horns at this level. Wrist stimulation, elicited activity in the 

right dorsal horn of 4 subjects and left dorsal horn of 5 subjects 

at the level of C7. Stimulation of the right palm induced 

fMRI signal change in the ventral horns of the C7 level in 5 

subjects, and in the right dorsal horn of the C7/T1 level in 4 

subjects. During left palm stimulation 5 of the 6 subjects 

exhibited activity in the ipsilateral dorsal horn at the T1 vertebral 

level. Patellar stimulation elicited activity in the right 

dorsal horn of the T11/T12 disk level in 5 of the 7 subjects. 

Activity was observed in the right and left ventral horns at 

the T12 vertebral level in four subjects during Achilles tendon 

stimulation. During all of the experiments areas of signal 

change were observed in adjacent segments of the spinal 

cord as well; however, they were much more variable 

between subjects. 

CONCLUSION 

This study demonstrates that the physiologically expected 

segmental organization of the human spinal cord can be 

monitored using fMRI, which may prove useful for the noninvasive 

evaluation of spinal cord function in humans. 

KEY WORDS: Spinal cord, functional MR imaging, vibratory 

stimulation 


Neurotractography of the Cervical Nerve Roots and 

Their Distal Fibers by Diffusion Tensor Scan Using 

Parallel Imaging 

Tsuchiya, K. 1 · Fujikawa, A. 1 · Honya, K. 1 · Nakajima, M. 1 · 

Takemoto, S. 2 

1 2 Kyorin University, Tokyo, JAPAN, Toshiba Medical 

Systems, Tokyo, JAPAN 

PURPOSE 

It has been reported that diffusion tensor (DT) tractography 

of the cervical spinal cord can be performed successfully 

using a single-shot echo-planar sequence in combination 

with parallel imaging. We report preliminary results of application 

of a similar technique to depict the cervical nerve 

roots and their distal fibers. 


Diffusion tensor imaging was performed with a 1.5 T MR 

system using a seven-channel receiver coil. We used a single-shot 

echo-planar FLAIR-based sequence with parallel 

imaging (TR/TE/TI = 13400/80/2200 msec, reduction factor 

= 1.8, 40 5 mm sections without gap, FOV = 240 x 350 mm, 

imaging matrix = 128 x 128, NEX = 4) to acquire diffusionweighted 

images in the axial plane with phase encoding in 

the anterior-posterior direction. Motion-probing gradient 

(MPG) was applied in six directions with a b-value of 500 

sec/mm 2 . Scan time was 6 minutes 16 seconds. In postprocessing, 

we performed maximum intensity projection to 

reconstruct images around the spine. Using this technique, 

we obtained images of the cervical spine from eight volunteers 

(six men and two women; age range, 21-45 years; mean 

age; 37 years).

RESULTS 

In each subject, not only the spinal cord but also most nerve 

roots as well as the nerve ganglia and peripheral nerves 

could be identified. The peripheral nerves were visualized as 

far as 3 cm distal to the neural foramen. Image distortion was 

minimal. 

CONCLUSION 

Diffusion tensor imaging using the parallel imaging technique 

and MPG of 500 sec/mm 2 can be a technique to visualize 

the cervical nerve roots and their distal fibers. 

KEY WORDS: Spine, tractography, diffusion 


Diffusion-Weighted Imaging of Peripheral Nerve Disease 

Kim, W. · Mukherjee, P · Chin, C. 



PURPOSE 

MR neurography (MRN) is emerging as a powerful tool in 

demonstrating peripheral nerves and associated pathology. 

The ability to visualize the peripheral nervous system allows 

for the potential to apply diffusion-weighted imaging not 

previously described in peripheral nerve diseases. To evaluate 

peripheral nerve disease with diffusion-weighted imaging. 


Patients with cervical or lumbosacral neuropathy were 

referred by neurologists, neurosurgeons, and oncologists for 

MRN (1/03-7/04). (Philips 1.5 T Gyroscan Intera, the 

Netherlands): Axial and coronal STIR (TR 2200; TE 20; TI 

160; NEX 4; FOV 22; 256 x 192; 3/0.3), T1 pre and postgadolinium 

with fat saturation (TR 500; TE 14; NEX 3). 

Coronal diffusion-weighted imaging was obtained in those 

patients with evidence of a peripheral nerve mass. Diffusionweighted 

imaging was performed in three directions (singleshot 

echo-planar: TR 2 x pulse-pulse interval; TE 70 msec; 

195 

FOV 22; 256 x 144; 5.0/0.5 mm; B-value 400 sec/mm 2 ). 

Apparent diffusion coefficient (ADC) maps were obtained 

and regions of interest were drawn and analyzed for mean 

ADC values using software available on the scanner. 

RESULTS 

Nine patients underwent MRN with diffusion-weighted 

imaging: 4 male, 5 female; mean age 42.3 years (range 6-88 

years). Seven patients were diagnosed with peripheral nerve 

tumors (4 primary nerve sheath tumors; 3 metastatic 

tumors). One patient had radiation plexopathy and one 

patient had an inflammatory neuritis. Diagnoses were confirmed 

pathologically (schwannoma, neurofibroma, 

leukemia, and inflammation) as well as by clinical and radiographic 

correlation. All patients with tumor and the patient 

with inflammatory neuritis demonstrated increased STIR 

signal and enhancement within the affected nerve(s) on 

MRN. The patient with radiation plexopathy demonstrated 

nonenhancing increased STIR signal within the affected 

nerves. On diffusion-weighted imaging, the peripheral nerve 

tumors appeared similar in signal intensity to normal nerve. 

The patients with radiation and inflammatory plexopathy 

demonstrated increased diffusion within affected nerves 

compared to normal nerve. The ADC values obtained from 

peripheral nerve tumors ranged from 0.91 to 2.01 x 10 -3 

mm 2 /sec (mean 1.63 +/- .34 x 10 -3 mm 2 /sec). Radiation therapy-related 

change resulted in an ADC value of 2.78 +/- .26 

x 10 -3 mm 2 /sec and inflammatory neuritis resulted in an ADC 

value of 3.82 +/- .32 x 10 -3 mm 2 /sec. The normal ADC values 

of peripheral nerves ranged from 0.87 to 1.63 x 10 -3 

mm 2 /sec (mean 1.44 +/- .21 x 10 -3 mm 2 /sec.). 

CONCLUSION 

MR neurography has allowed visualization of peripheral 

nerves and peripheral nerve diseases such as tumor, radiation, 

and inflammatory changes with greater detail. 

Diffusion-weighted imaging has not been described in 

peripheral nerve disease and can be applied to peripheral 

nerves and ADC values obtained in tumor, radiation, and 

inflammatory changes affecting peripheral nerves. 

KEY WORDS: diffusion-weighted imaging, MR neurography 

Thursday

Thursday 


10:00 AM - 11:30 AM 

Theatre 

(64b) ADULT BRAIN: Metabolic and 

Miscellaneous 



(64a) SPINAL CORD AND PERIPHERAL NERVES: 




Moderators: David J. Mikulis, MD 



High-Resolution Diffusion Tensor Imaging of the Human 

Hippocampus 

Shepherd, T. M. 1 · Ozarslan, E. 1 · Yachnis, A. T. 1 · Blackband, 

S. J. 1,2 

1 2 University of Florida, Gainesville, FL, National High Field 

Magnet Lab, Tallahassee, FL 

PURPOSE 

At 1.5 T, it is difficult to resolve individual lamina of the hippocampus 

with diffusion tensor MR imaging (DTI). 

However, assessing changes to diffusivity and fractional 

anisotropy for individual hippocampal lamina may improve 

the specificity of DTI for particular hippocampal pathologies 

like Alzheimer’s disease. As a first step towards this goal, 

this study characterizes DTI contrast in the different lamina 

of healthy hippocampus autopsy specimens at 60 µm inplane 

resolution. 


Eight mm coronal segments of hippocampal body were dissected 

from 5 autopsy specimens (mean age 55.6 ± 6.2 

years). These samples were immersion-fixed in 20% formalin 

for 1 week, had short post-mortem intervals to fixation 

(21.2 ± 5.7 hours), and no evidence for neuropathology. 

Hippocampi were washed in PBS for 24 hours, then imaged 

in Fluorinert TM using a 10 mm birdcage coil inside a 14.1 T 

magnet. Diffusion tensor imaging datasets were obtained 

with 60 µm in-plane resolution using a multislice pulsedgradient 

spin-echo sequence (TR/TE = 1500/34 ms, slice 

thickness = 300 µm). An image without diffusion-weighting 

was collected (NEX = 36), then 21 diffusion-weighted 

images were collected with 415 mT/m diffusion gradients 

oriented along different directions (diffusion time = 17 ms, b 

= 1250 s/mm 2 , NEX = 12). These data were used to calculate 

the apparent diffusion tensor at each image voxel. The mean 

diffusivity (), fractional anisotropy (FA) and fiber direc- 

196 

tion then were determined for manually segmented hippocampal 

lamina. 

RESULTS 

Diffusion tensor imaging at 60 µm resolution were obtained 

from 5 hippocampi with signal-to-noise ratios of 31.1 ± 13.0 

at b = 1250 s/mm 2 . M 0, , and FA images demonstrated 

the laminar anatomy of the hippocampus well. Diffusivity 

ranged from 1.21 ± 0.22 x 10 -4 mm 2 /s in the fimbria to 3.48 

± 0.72 x 10 -4 mm 2 /s in granule cells (P < 0.001). The FA of 

several hippocampal lamina were also statistically different 

(P < 0.001). Color fiber maps (Fig.) demonstrated many 

additional microstructural features of the human hippocampus 

(e.g., that principal orientation in stratum radiatum is 

dominated by pyramidal neuron apical dendrites). 

CONCLUSION 

High-resolution DTI provides a wealth of microstructural 

information about specific anatomical lamina of the human 

hippocampus. These data can be used to study diffusion-specific 

changes to particular hippocampal regions to explain 

signal changes observed with current clinical scanners and 

may help develop specific MR imaging surrogate markers 

for neuropathologies such as Alzheimer’s disease or epilepsy. 

In addition, contrary to previous reports, these data suggest 

water diffusion in the human CNS is not homogeneous. 

KEY WORDS: Microscopy, anisotropy, tractography 


Comparison Study of Diffusion Tensor and Diffusion- 

Weighted Imaging in Wallerian Degeneration after 

Cerebral Ischemic Stroke 

Liu, X. 1 · Tian, W. 2 · Westesson, P. 1 

1University of Rochester School of Medicine and Dentistry, 

Rochester, NY, 2Beijing TongRen Hospital, Beijing, CHINA 

PURPOSE 

To evaluate the diffusion tensor imaging (DTI) manifestation 

of wallerian degeneration by cerebral ischemic stroke; and 

compare the ability of DTI and diffusion-weighted imaging 

(DWI) in the diagnoses of wallerian degeneration. 


Sixteen volunteers (9 males and 7 females) and 44 scans of 

29 patients (22 males, 7 females), who ranged 11 hours to 1 

year after MCA territory ischemic stroke, which involved 

cortical spinal tract, were analyzed, including 13 patients

accepted prospective series of DWI (TR/TE = 7000/80.5, b 

value 1000 s/mm 2 ) and DTI examination (EPI sequence 

TR/TE = 8000/85, b value 1500 s/mm 2 ,13 directions) in GE 

Signa 3 T after stroke onset. Fractional anisotropy (FA), 

anistropy index (AI) of DTI and apparent diffusion coefficient 

(ADC) of DWI in regions of interest (ROIs) of cerebral 

peduncle were calculated and compared. Three-dimensional 

tractography of bilateral cortical spinal tract in 22 patients 

were created. 

RESULTS 

According to the time post stroke onset, all data were divided 

into 5 groups. Group 1: scans in the first week after stroke 

onset. Group 2: from 2 weeks to 4 weeks. Group 3: 5-14 

weeks. Group 4: after 14 weeks. Combing with DWI and 

regular T2-weighted and FLAIR images, only 32 scans were 

diagnosed as wallerian degeneration. None of these 32 scans 

was in group 1. But on color FA and AI maps, it is clear to 

see that cortical spinal tract in lesion side was smaller than 

contralateral side. In all groups, FA and AI values of cerebral 

peduncle in lesion side were lower than contralateral side 

cerebral peduncle, P < 0.01 by paired t test. In group 1, ADC 

values of cerebral peduncle in lesion side were lower than 

contralateral side cerebral peduncle, P < 0.01. While in other 

groups, the ADC values between both side cerebral peduncles 

didn’t have significant difference. On 3D tract fiber 

maps of both cortical spinal tracts by “seed” method, the 

fiber tracts in the lesion side were thinner than contralateral 

side. 

CONCLUSION 

Wallerian degeneration in acute, subacute, and chronic phases 

of ischemic stroke could be observed on DTI images. 

Diffusion tensor imaging could detect wallerian degeneration 

earlier and clearer than regular MR imaging and DWI 

by characteristic of decreasing anisotropy diffusion of cortical 

spinal tract. Three-dimensional tract map postprocessed 

from DTI data is useful observing the status of bilateral pyramid 

tracts stereotacticly. Diffusion tensor imaging is the 

optimum imaging method for diagnosing wallerian degeneration 

in vivo. 

KEY WORDS: Diffusion tensor imaging, diffusion-weighted 

imaging, wallerian degeneration 


Need for the Left Hemisphere? Remarkable Clinical 

Outcome after Prenatal Brain Damage in a Case of 

Hemihydranencephaly 

Ulmer, S. 1 · Moeller, F. 1 · Brockmann, M. A. 2 · Kuhtz- 

Buschbeck, J. P. 3 · Stephani, U. 1 · Jansen, O. 1 

1 University Hospital of Schleswig-Holstein, Kiel, 

GERMANY, 2 University Hospital of Schleswig-Holstein, 

Luebeck, GERMANY, 3 Christian Albrecht University, Kiel, 

GERMANY 

PURPOSE 

Occlusion of one carotid artery occurring between weeks 20 

and 27 of gestation (before the last trimenon) - after neural 

migration and before synaptogenesis - is thought to be the 

underlying mechanism of hemihydranencephaly. Only seven 

case reports of this disorder in which the human brain lacks 

197 

one complete hemisphere are published. We present a 36year-old 

man born at term from nonconsanguineous, healthy 

parents. Due to a right-sided hemiparesis he received physical 

therapy during childhood. He completed school and is 

now working in a security department. He reported occasional 

headaches and nuchal pain. 


We performed MR imaging and MR angiography (MRA) on 

a 1.5 T scanner (Magnetom Vision, Siemens, Erlangen, 

Germany) and a variety of standardized tests [Wolf motor 

function test (WMFT), motor-activity-log (MAL), 2 point 

discrimination (2 pd), Purdue pegboard and documentation 

of mirror movements] for quantitative assessment of sensorimotor 

skills and language tests [Aachener aphasia test 

(AAT) including spontaneous speech, repetition, naming, 

comprehension, written language, and the token test]. 

RESULTS 

MR imaging demonstrated nearly complete absence of the 

left hemisphere with exception of a small residual hippocampus, 

a small rim of the occipital cortex, and a small 

cerebral peduncle due to wallerian degeneration. The cerebellum 

was completely normal. Time-of-flight (TOF) 

angiography demonstrated absence of the left internal 

carotid artery (ICA). The patient reported (self-assessment, 

MAL) that he is able to drive a car with standard transmission. 

Although natural prehension movements were almost 

unimpaired, he is unable to open a door with a key, button or 

unbutton clothing, switch between TV programs using a 

remote control, or perform further tasks requiring fine motor 

control. In the WMFT, he reached 75% of all possible points 

because of reduced speed and/or precision of almost all the 

performed tasks and because of impaired fine motor control 

in three tasks. Functionality and quality of movement did not 

differ in terms of the score. The thresholds for the 2-point 

discrimination of the tips of pinprick testing were 10 mm 

apart from each other on the affected side. He was unable to 

perform any required task in the Purdue pegboard. Strong 

mirror movements were found when moving both hands. 

The AAT confirmed that speech was not affected. The token 

test was completed correctly. The patient is strongly lefthanded 

(handedness by Millner). 

CONCLUSION 

Because of the subtotal damage of the “dominant” hemisphere, 

a severe hemiparesis, aphasia or dysphasia or severe 

intellectual impairment would have been expected. 

Disabilities in the clinical tests were caused by deficits in 

fine motor control, strong mirror movements, and reduced 

speed and/or reduced precision of the performed tasks. Fine 

motor control requires unaffected 2-point discrimination. 

Morphology cannot predict clinical impairment. Cortical 

reorganization with remarkable outcome is possible if the 

damage occurs in very early childhood. Even though there 

must be a genetic predisposition to define the precentral 

gyrus as motor cortex, in the immature brain, cortical areas 

may exist that can adopt similar or contralateral function and 

enable cortical plasticity whose protocols are executed via 

the uncrossed corticospinal fibers. 

KEY WORDS: Hemihydranencephaly, MR imaging, outcome 

Thursday

Thursday 


Comparison of High-Resolution 3D Flash Contrast- 

Enhanced MR Angiography Using Parallel Acquisition 

Techniques to Conventional 3D Time-of-Flight MR 

Angiography in Carotid Artery Stenosis 

Sunenshine, P. J. · Pramanik, B. · Law, E. M. · Hecht, E. M. 


New York, NY 

PURPOSE 

Three-dimensional (3D) time-of-flight (TOF) MR angiography 

(MRA) is a well established high-resolution noninvasive 

way to assess carotid artery stenosis. Although accurate, this 

method is prone to artifacts as a result of long acquisition 

times and saturation effects. Contrast-enhanced MRA techniques 

have become increasingly popular to counter the saturation 

effects of slow flow and decrease acquisition times. 

However, spatial resolution of contrast-enhanced MRA is 

relatively low compared to conventional 3D TOF imaging. 

With the advent of parallel acquisition techniques and multichannel 

neck coils, spatial resolution can be increased significantly 

without sacrificing temporal resolution. The purpose 

of this study is to compare conventional 3D TOF MRA 

and 3D FLASH contrast-enhanced MRA with parallel imaging 

to determine if parallel imaging can be used to improve 

the spatial resolution of dynamic contrast-enhanced MRA to 

that comparable with conventional 3D TOF MRA. 


Prospectively, four consecutive patients presenting for 

carotid MRA underwent 3D TOF MRA and 3D contrastenhanced 

MRA with parallel acquisition techniques (GRAP- 

PA, acceleration factor 3) at 1.5 T using a 4-channel phasedarray 

neck coil and a multiple-channel head coil. 

Conventional 3D imaging was performed using a voxel size 

of 0.7 x 0.5 x 0.9 mm, a 288 x 384 matrix, and an acquisition 

time of 6:21 minutes. Dynamic contrast-enhanced MRA 

with parallel imaging was performed using a 3D FLASH 

sequence using a voxel size of 0.8 x 0.8 x 0.8, a 384 x 512 

matrix, and an acquisition time of 24 seconds following 

intravenous injection of gadolinium (0.2 mmol/kg) and a 20 

cc saline bolus via power injector. Timing was determined 

using a time-resolved echo-sharing angiographic technique 

(TREAT) using a 3 cc contrast bolus and 20 cc saline flush. 

Quantitative assessment was performed by calculating signal-to-noise 

ratio (SNR) and contrast-to-noise ratios (CNR) 

in the distal common carotid artery for each MRA sequence. 

Qualitative assessment was performed by two neuroradiologists 

blinded to the imaging method used. Subjective assessment 

of image quality, noise, degree of stenosis using the 

NASCET criteria, and degree of confidence in diagnosis was 

performed using a 4-point scale (1 = excellent, 4 = nondiagnostic). 

RESULTS 

Signal-to-noise ratio and CNR were both increased with contrast-enhanced 

MRA as shown in the table. Qualitative 

parameters such as overall quality was better in the contrastenhanced 

MRA sequences. Image quality of the contrastenhanced 

MRA was rated the same or higher on all patients 

when compared with conventional noncontrast-enhanced 3D 

TOF imaging. 

198 

TABLE 1 3D TOF vs CE MRA with Parallel Imaging 

NC 3D TOF 3D CE MRA P-value 

SNR 106.6 139.4 p ≤ 0.75 

CNR 81.7 121.9 p ≤ 0.5 

Overall Quality 2.2 1.9 p ≤ 0.5 

*Note: Average overall quality on a scale of 1-4; (1 = excellent, 

2 = more than adequate for diagnosis, 3 = adequate for 

diagnosis, 4 = nondiagnostic). 

CONCLUSION 

Although not statistically significant in this preliminary 

small group of patients, dynamic contrast-enhanced MRA 

using parallel imaging to achieve comparable spatial resolution 

demonstrated an overall trend of increased signal to 

noise and better overall quality of image when compared to 

conventional 3D TOF imaging. 

KEY WORDS: MR angiography, carotid, parallel 


Utility of CT Angiography with Virtual Angioscopy versus 

Doppler Ultrasound in the Examination of the 

Stented Carotid Artery 

Orbach, D. B. · Pramanik, B. · Lee, J. · Maldonado, T. · Riles, 

T. · Grossman, R. I. 


New York, NY 

PURPOSE 

Doppler ultrasound (DU) is the imaging modality most commonly 

used for follow up after placement of carotid artery 

stents. However, many studies have reported a not insignificant 

incidence of false-positive DU reports of intrastent 

stenosis, likely due to distortion of the ultrasound beam by 

the stent material. We developed a methodology for accurately 

assessing carotid stents using a multidetector CT to 

perform CT angiography (CTA) and virtual angioscopy. 


We measured luminal stent diameter in a cohort of six 

patients who recently had undergone carotid stenting and in 

whom we had CTA, intraoperative digital subtraction 

angiograms (DSA), and DU, all performed within 1 month. 

Additionally, we generated virtual angioscopic endoluminal 

views inside the stents in order to assess the surface morphology 

of the endoluminal aspect of the stent. The patients 

returned for 6-month follow up, at which time their DU 

results were compared again with the CTA and virtual 

angioscopy. Our methodology for performing the CTA, as 

previously described, involved a specialized reconstruction 

kernel and optimized transparency and color tables for the 

virtual angioscopy. 

RESULTS 

In two cases, the first post-stent DU reported a moderate 

stenosis, while the DSA and CTA showed the stents to be 

patent. In cases in which there was true restenosis requiring 

repeat angioplasty, all three modalities were likely to give an 

accurate depiction. However, only CTA with virtual 

angioscopy allowed for visualization of the morphology of 

the stenotic lesion. In a case with restenosis requiring repeat

angioplasty, virtual angioscopy showed the stenotic lesion to 

lie asymmetrically on one side of the stent lumen, with an 

appearance very different from that of symmetric intimal 

hyperplasia. 

CONCLUSION 

CT angiography is a viable modality for following patients 

after carotid artery stenting. CT angiography may not be susceptible 

to the false-positive findings found not infrequently 

with DU of the stented carotid. Surface morphologic features 

may be useful in stratifying the potential risk presented by 

individual patients’ plaques. 

KEY WORDS: CT angiography, virtual angioscopy, carotid 

stent 


“Idiopathic” Normal Pressure Hydrocephalus: Further 

Evidence That Normal Pressure Hydrocephalus Is a 

Two-Hit Disease-Benign External Hydrocephalus in 

Infancy and Deep White Matter Ischemia in Late 

Adulthood 

Bradley, W. G. · Bahl, G. 

University of California San Diego 

San Diego, CA 

PURPOSE 

Previous studies have shown that the intracranial volumes of 

patients with normal pressure hydrocephalus (NPH) are significantly 

greater than age- and sex-matched controls, suggesting 

that NPH begins in infancy when the sutures can still 

expand. If this is true, then a major route of egress for CSF 

from the ventricles would be through the extracellular space 

of the brain to the subarachnoid space. In this case the apparent 

diffusion coefficient (ADC) should be higher in NPH 

patients compared to normal controls with comparable levels 

of DWMI (which independently raises ADC). If the myelin 

pallor of the DWMI results in a more hydrophilic environment 

and increased resistance to centrifugal migration of 

CSF, there should be a build up of CSF and an increase in 

ADC medial to the DWMI. The purpose of this study was to 

determine if the absolute ADC is higher in the periventricular 

region for comparable levels of DWMI and if it “peaks” 

medial to the DWMI. 


Thirteen patients were referred for evaluation of NPH on the 

basis of clinical symptoms. They had aqueductal CSF stroke 

volumes averaging 177 uL which is more than 4 times normal 

on phase-contrast CSF velocity imaging. Apparent diffusion 

coefficient profiles were compared to 10 age-matched 

controls at an axial level just above the lateral ventricles. 

Comparable sized regions of interest (ROIs) were drawn in 

the centrum semiovale for patients and controls, and the 

average ADC was computed over the ROI. Since DWMI is 

associated with increased ADC, patients and controls were 

chosen to have a similar amount of DWMI, which was graded 

as mild to moderate in severity by comparison of T2weighted 

images. 

RESULTS 

The average ADC in the immediate periventricular region 

199 

for comparable levels of DWMI was 1519 for NPH patients 

compared to1377 for controls (p < .01 using an independent 

t-test). Apparent diffusion coefficient profiles going horizontally 

across the brain showed a peak in ADC in NPH patients 

medial to the observed DWMI which was not seen in controls. 

CONCLUSION 

These findings are consistent with DWMI causing obstruction 

to the extracellular migration of CSF. Since increased 

resistance to the egress of CSF has been shown to cause ventricular 

enlargement, this is additional evidence that DWMI 

is the second hit in NPH. 

KEY WORDS: Dementia, MR imaging, hydrocephalus 


Diagnostic Value of Early Brain MR Imaging in Patients 

with Clinically Isolated Syndromes 

Rovira, A. · Tintoré, M. · Aymerich, X. · Grivé, E. · Nos, C. 

· Huerga, E. · Montalban, X. 

Vall d’Hebron Hospital 


PURPOSE 

According to the criteria proposed by McDonald, a diagnosis 

of multiple sclerosis (MS) in patients with a clinically 

isolated syndrome (CIS) can be achieved with a single brain 

MR image obtained at least 3 months after the onset of the 

clinical event. However, in clinical practice, a diagnostic 

brain MR image is obtained earlier in many instances with 

the purpose of freeing patients from uncertainty on diagnosis. 

The purpose of this study is to analyze the diagnostic 

properties of MR imaging obtained prior to the 3-month 

interval. 


Consecutive patients with a CIS in whom two brain MR 

imaging studies were performed, the first within 8 weeks and 

the second 3 to 5 months after symptoms onset. In both 

exams the presence and number of T2 lesions, number of 

gadolinium-enhancing lesions, number of Barkhof criteria, 

and (only in the second exam) the presence of new T2 

lesions were assessed. The association between the MR 

imaging data obtained with the two exams was calculated 

using the Pearson Chi-square test. In a subgroup of patients 

with a minimum clinical follow up of 24 months, we compared 

the diagnostic properties of the two MR imaging 

exams with regard to conversion to clinically definite MS 

(CDMS) according to the Poser criteria. 

RESULTS 

Eighty-six patients were included in this study (mean age: 

31.1 years; female-to-male ratio:1.86). The first MR imaging 

exam was obtained at a median interval of 20 days and the 

second at a median of 113 days. At the first exam 36% of 

patients had no focal lesions, 42% had at least nine T2 

lesions; 40% at least three Barkhof criteria; 31% at least one 

gadolinium-enhancing lesion and, 27% at least three 

Barkhof criteria and at least one gadolinium-enhancing 

lesion. These percentages changed slightly in the second 

exam (35%, 45%, 48%, 26%, and 24% respectively). As 

Thursday

Thursday 

expected, a significant positive association was demonstrated 

between the data obtained in the first and second MR 

imaging exams (P < .001). Demonstration at least three 

Barkhof criteria and at least one gadolinium-enhancing 

lesion at the first MR imaging predicted the presence of 

these two criteria at the second MR imaging with a high 

specificity (89.4%) and negative predictive value (92.2%). 

Among the subgroup of patients with minimum follow up of 

24 months, 26% presented a clinical relapse, and therefore 

fulfilled the diagnosis of CDMS. Comparison of the diagnostic 

properties between the first and the second MR imaging 

with regard to conversion to CDMS showed very similar 

results in sensitivity (67% vs 67%), specificity (92% vs 

88%) and accuracy (85% vs 82%). 

CONCLUSION 

The findings from this study suggest that a single brain MR 

image may suffice to establish the diagnosis of MS even 

when it is performed within the first 2 months after the onset 

of symptoms. Nevertheless, long-term validation is required 

before the diagnostic criteria proposed by McDonald, et al. 

can be applied during this early period. 

KEY WORDS: Multiple sclerosis, diagnosis, MR imaging 


Investigation into the Voxelwise Comparison of 

Fractional Anisotropy Changes in Multiple Sclerosis 

Patients Using Diffusion Tensor Imaging 

Patel, S. A. 1 · Faro, S. H. 2 · Hum, B. 1 · Gonzalez, C. F. 1 · 

Schwartzmann, R. 1 · Mohamed, F. B. 2 

1 Drexel University, Philadelphia, PA, 2 Temple University, 


PURPOSE 

The purpose of this study was to develop a semiautomated 

method to investigate the fractional anisotropy (FA) changes 

in multiple sclerosis (MS) patients using voxelwise analysis 

and comparison to a normalized FA atlas using a standard 

neuroanatomical space. 


Diffusion tensor imaging was performed on 10 normal controls 

and 6 patients with MS lesions. We acquired DTI by 

collecting diffusion-weighted images using a spin-echo EPI 

sequence. Diffusion gradients were applied in six noncollinear 

directions with a b value of 1000 sec/mm 2 . Twenty 

6 mm axial slices covering the entire brain were imaged 

using a 1.5 T scanner (TR = 6000 ms, TE = 100 ms, FOV = 

240 mm, 98 x 128 and 4 acquisitions). Fractional anisotropy 

images then were created using an in-house modified diffusion 

tensor toolbox of SPM99. The FA maps then were spatially 

normalized into the space of the SPM99 EPI template 

and smoothed. A normal FA atlas was created using the 10 

normal subjects. Statistical maps were generated by voxelwise 

comparison of the FA map of each of the patients with 

the normal FA atlas in a two sample t-test (uncorrected p 

value of 0.05). 

RESULTS 

The statistical maps showing regions of significant FA differences 

were compared with routine T1, T2, and fluid atten- 

200 

uated inversion recovery (FLAIR) images. As can be seen in 

Fig. 1, there were numerous regions (shown in red, orange, 

and yellow) with statistically significant (p = 0.05) FA 

changes. Note that the large lesion on the statistical map in 

the left hemisphere corresponds well in location with the 

lesion shown on the FLAIR image (arrow). It can be seen 

that the center of the lesion exhibits more significant FA 

changes than the peripheral regions. Also, there are numerous 

regions of FA changes throughout the brain in the normal-appearing 

white matter (NAWM) that were not visible 

on the routine MR images. Similar findings were seen on the 

rest of the five MS cases analyzed as well. In two cases, 

however, there were lesions visible on MR imaging that did 

not show significant FA changes. 

CONCLUSION 

Our results show that this semiautomated method significantly 

improved the speed, accuracy, and reproducibility of 

FA map analysis in MS patients. Also, this method demonstrated 

our ability to globally visualize FA changes as well as 

characterize quantitatively the extent of MS lesions and 

abnormal FA in NAWM. Such analysis of FA data using a 

standard anatomical atlas will greatly help in longitudinal 

follow up of these MS patients. These preliminary results are 

very encouraging and warrant further investigation. 

KEY WORDS: Multiple sclerosis, diffusion tensor imaging 


Brain MR Manifestations of Hepatic Encephalopathy 

Tomlin, H. A. 1 · Prayer, D. 2 · Schmid, M. 2 

1 McMaster University, Hamilton, ON, CANADA, 2 Medical 

University of Vienna, Vienna, AUSTRIA 

PURPOSE 

With the high prevalence of liver disease and cirrhosis in 

many populations, hepatic encephalopathy is an important 

diagnostic entity. Although potentially clinically inapparent, 

up to 70% of patients with cirrhosis may have at least minimal 

hepatic encephalopathy (1, 2). Since high T1 signal 

intensity in the basal ganglia and low T2 signal in the putamen 

and cerebellar dentate nucleus have been described in 

the literature in hepatic encephalopathy (3-6) prevalence of 

these changes in patients with advanced liver disease, as well 

as assessment of degree of cerebral and cerebellar atrophy 

and Evans ratio were evaluated. 


Twenty-eight admitted adult patients with advanced stage

liver disease, Child-Pugh stages B and C, underwent standardized 

sequence brain MR imaging including T1 MTC, T2 

and FLAIR. Resulting data and measurements were tabulated 

and prevalence data were calculated. Images were 

reviewed with respect to MR signal intensity in the basal 

ganglia, midbrain, dentate nucleus, and cerebral and cerebellar 

white matter. Atrophy was evaluated using Evan’s ratio in 

the supratentorial region and visibility of cerebellar sulci on 

axial T2-weighted images infratentorially. 

RESULTS 

Results demonstrated high T1 signal in 70.4% of patients in 

the globus palliudus, 81.5% in the putamen, 22.2% in the 

head of the caudate, and 3.7% in the pars compacta. Low T2 

signal was recorded in the putamen in 88.9% and the dentate 

nucleus in 37.0%. Evan’s ratios averaged 0.32 indicating a 

mild degree of cerebral atrophy in the study group overall. 

Cerebral atrophy was seen in 75.0% and cerebellar atrophy 

in 87.5%. Atrophy of the pars compacta of the midbrain was 

identified in 17.9%. 

CONCLUSION 

The majority of patients with advanced liver disease demonstrate 

MR findings of high T1 signal in the globus pallidus, 

high T1 and low T2 signal in the putamen, and cerebral and 

cerebellar atrophy. Approximately one third of the patients 

have low T2 signal in the dentate nucleus, about one-fifth 

show high T1 in the head of the caudate and display pars 

compacta atrophy. The average Evans Ratio showed mild 

atrophy at 0.32. These radiologic statistics are useful in combination 

with the clinical presentation to aid in the diagnosis 

of hepatic encephalopathy. 

REFERENCES 

1. Ong JP, Mullen KD. Hepatic encephalopathy. Eur J 

Gastroenterol Hepatol 2001,13(4):325-334 

2. Das A, Dhiman RK, Saraswat VA, Verma M, Naik SR. 

Prevalence and natural history of subclinical hepatic 

encephalopathy in cirrhosis. Gastroenterol Hepatol 

2001;16(5):531-535 

3. Cakirer S, Galip GM, Beser M. Acquired hepatocerebral 

degeneration. Online reference. 2001 April 08. URL: 

http://www.eurorad.org/case.cfm?UID=1033 

4. Morgan MY. Cerebral magnetic resonance imaging in 

patients with chronic liver disease. Metab Brain Dis 

1998;13(4):273-290 

5. Brunberg JA, Kanal E, Hirsch W, et al. Chronic acquired 

hepatic failure: MR imaging of the brain at 1.5 T. AJNR Am 

J Neuroradiol 1991;12(5):909-914 

6. Jog MS, Lang AE. Chronic acquired hepatocerebral degeneration:case 

reports and new insights. Mov Disord 

1995;10(6):714-722 

KEY WORDS: Encephalopathy, MR imaging, hepatic 

201 


Rabies Encephalomyelitis: Neuroradiologic and 

Pathologic Findings in Four Transplant Recipients 

Gomez, C. E. 1 · Burton, E. C. 1 · El-Feky, W. 2 · Opatowsky, 

M. 1 

1 Baylor University Medical Center, Dallas, TX, 2 Texas 

Neurology, Dallas, TX 

PURPOSE 

Two males and two females aged 18 to 55 years, received 

organs and vascular tissue from a single donor who had died 

from rapid neurologic deterioration and subarachnoid hemorrhage. 

Presented and reviewed are the neuroimaging and 

pathologic findings of these first known confirmed cases of 

rabies encephalitis transmitted via solid organ and vascular 

transplantation. 


Case 1: 53-year-old male liver recipient. Case 2: 50-year-old 

female kidney recipient. Case 3: 18-year-old male kidney 

recipient. Case 4: 55-year-old female iliac artery recipient. 

RESULTS 

Case 1: Initial MR imaging showed diffusely increased 

FLAIR signal involving the leptomeninges and equivocal 

FLAIR signal elevation in the brain stem. Five days later, 

MR imaging revealed profound symmetric abnormalities 

within the frontal and temporal lobes, hippocampi, basal 

ganglia, medulla and visualized portions of the upper cervical 

spinal cord. 

Case 2: MR imaging demonstrated early findings of sulcal 

effacement along the frontoparietal vertex as well as subtle 

FLAIR signal elevation of the surface leptomeninges. Five 

days later, two sequential unenhanced CT studies progressed 

over a 7-hour period from subtle abnormalities to marked 

cerebral edema with ventricular and basal cistern effacement 

as well as herniation. Case 3: Initial MR imaging showed 

symmetric areas of T1-signal shortening within the posterior 

putamina suggestive of microhemorrhages. Ten days later, 

this pattern was seen throughout the corpus striatum, midbrain, 

and hippocampi. In addition, a large left subinsular 

Thursday

Thursday 

infarction had developed in the absence of angiographic evidence 

of vasculitis. Case 4: Baseline senescent brain CT 

findings rapidly progressed to increased FLAIR and T2 signal 

throughout the basal ganglia, hippocampi, brain stem and 

perirolandic gyri. 

CONCLUSION 

Rabies encephalomyelitis is uncommonly encountered in 

clinical practice in the United States. We report the first 

known cases of human-to-human transmission of rabies via 

organ and vascular transplantation. All four patients as well 

as the donor had a fulminant encephalitic course that resulted 

in death. The imaging and pathologic manifestations of 

these four cases show both classically described findings of 

rabies encephalomyelitis as well as some atypical features. 

KEY WORDS: Rabies, encephalomyelitis 


Tectal Lesions as the Only Manifestation of Whipple’s 

Disease: Report of Two Cases 

Cortes, M. D. P. · Tampieri, D. · Melancon, D. 

Montreal Neurological Institute 


PURPOSE 

To report two unusual cases of tectal-enhancing lesions as 

the only manifestation of Whipple’s disease in two males. 


Two males, 24 and 30 years of age presented with neurologic 

deficits. The first patient developed acute symptomatic 

hydrocephalus and the second patient vertical gaze palsy. 

There were no systemic symptoms associated. 

Neuroimaging demonstrated brainstem lesions in both 

patients; eventually, biopsies were necessary due to clinical 

deterioration, enlargement of the lesions and in one of the 

patients lack of response to medical treatment with steroids. 

Histopathologic analysis led to the diagnosis of Whipple’s 

disease. With antimicrobial treatment the first patient has 

remained stable clinical and radiologically. The second 

patient has consistent clinical and imaging improvement. 

RESULTS 

Sequential CT scan and MR imaging revealed a progressively 

enlarging lesion with inhomogeneous enhancement located 

in the upper brainstem of the first patient. In the second 

case, two ring-enhancing lesions surrounded by mild edema 

at the tectum were seen. After treatment, imagenologic follow 

up has demonstrated reduction of the lesions in one of 

the patients and stability in the other. 

CONCLUSION 

The diagnosis of Whipple’s disease is more common nowadays. 

Rarely, this disease can present confined to the CNS. 

In these circumstances, the neuroradiologic diagnosis is particularly 

challenging, since no pathognomonic findings have 

been described and lesions of various locations and appearances 

have been reported. We have presented two cases with 

different midbrain lesions secondary to Whipple’s disease 

that manifested only neurologically. The presence of lesions 

of similar appearance adds to the already broad spectrum of 

202 

imaging findings in CNS Whipple’s disease, and should lead 

the neuroradiologist to include this possibility in the list of 

differential diagnosis. A review of the literature and discussion 

of the topic will be provided. 

KEY WORDS: Whipple’s disease, brainstem, CNS 

Discussion 


10:00 AM - 11:33 AM 

Room 107 

(64c) ADULT BRAIN: Trauma and 

Vascular Lesions 







Moderators: Alisa D. Gean, MD 

Lindell R. Gentry, MD 


Automated Detection of Traumatic White Matter Injury 

Using Voxel-Based Morphometry of Diffusion Tensor 

Images: A 3 T Study with Parallel Imaging 

Le, T. H. 1 · Mukherjee, P. 1 · Manley, G. T. 1 · Meeker, M. 1 · 

Aton, E. 1 · Vigneron, D. B. 1 · Dillon, W. P. 1 · Ghajar, J. 2 · 

Cognitive and Neurobiological Research Consortium- 

Traumatic Brain Injury 

1 University of California San Francisco, San Francisco, CA, 

2 Cornell University Medical College, New York, NY 

PURPOSE 

To perform automated detection of white matter injury in 

individual trauma patients by comparison of diffusion tensor 

images (DTI) to a database of normal volunteers using 

voxel-based morphometry. 


Eleven patients with head trauma, ranging from mild to 

severe, and 15 adult volunteers were imaged with conventional 

MR imaging and DTI performed on a 3 T GE scanner 

with an eight-channel EXCITE head coil. Conventional MR 

imaging sequences included FLAIR, gradient-echo T2*weighted, 

and high-resolution 3D FSPGR T1-weighted and 

3D FSE T2-weighted imaging. Whole-brain DTI was 

acquired in 55 independent diffusion-encoding directions at 

b = 1000 s/mm2 and 1.8 mm isotropic spatial resolution

using an axial single-shot spin-echo echo-planar pulse 

sequence (TR = 14 s, TE = 63 ms, NEX = 1) with ASSET 

parallel imaging (acceleration factor 2). Fractional 

anisotropy (FA) images were spatially transformed and registered 

to a standardized FA template using nonlinear methods 

described in Statistical Parametric Mapping (SPM). The 

normal database was constructed from standardized FA maps 

of the 15 normal volunteer DTI scans that registered within 

two standard deviations of the residual squared difference 

from the FA template. Standardized FA maps from individual 

patients then were compared with this normal FA database 

using automated voxel-by-voxel comparison. Z-score 

and t-score maps were generated. 

RESULTS 

Acute FA reduction within the inferior frontal fasciculi was 

detected in three patients with inferior frontal contusions 

imaged in the first week after injury (example Figure, left). 

The reduction in FA improved on 1-month follow-up imaging 

(Figure, right) and correlated with the patients’ cognitive 

improvement. Fractional anisotropy maps from one patient 

with diffuse axonal injury at 6-month follow up demonstrated 

a diffuse decrease in FA throughout the white matter (Zscores 

Thursday 

intraspinal subdural space, and documents the time course of 

migration by serial imaging studies. 

REFERENCES 

1. Orlin JR, Osen KK, Hovig T. Subdural compartment in pig. 

Anat Rec 1991;230:22-37 

KEY WORDS: Hematoma, subdural, migration 


Perfusion CT Evaluation of Cerebral Vascular 

Autoregulation in Severe Head Trauma Patients 

Wintermark, M. 1,2 · Chiolero, R. 2 · van Melle, G. 3 · Maeder, P. 2 

· Regli, L. 2 · Schnyder, P. 2 · Meuli, R. 2 

1 University of California San Francisco, San Francisco, 

CA, 2 Lausanne University Hospital, Lausanne, 

SWITZERLAND, 3 University of Lausanne, Lausanne, 

SWITZERLAND 

PURPOSE 

In head trauma patients, cerebral vascular autoregulation is 

meant to maintain brain perfusion despite systemic variations 

in blood pressure. Disturbance of autoregulation is key 

to the development of cerebral edema in severe head trauma 

patients. The purpose of this study was to evaluate the ability 

of perfusion CT (PCT) to characterize cerebral vascular 

autoregulation in severe head trauma patients with brain 

edema. 


A total of 80 PCT examinations were obtained in 42 severe 

head trauma patients with features of cerebral edema on conventional 

noncontrast cerebral CT, either on admission or 

during follow up. Perfusion CT results were correlated with 

the mean arterial pressure (MAP) measured at the time of 

each PCT examination. A cluster analysis allowed identification 

of different subgroups of patients. Mean arterial pressure 

values and PCT results in these groups were compared 

using Kruskal-Wallis and Wilcoxon (Mann-Whitney) tests. 

Moreover, the functional outcome of the 42 patients was 

evaluated 3 months after trauma on the basis of the Glasgow 

Outcome Scale (GOS) score. 

RESULTS 

Three main groups of patients were identified: a) 22 PCT 

examinations collected in 13 patients, characterized by high 

rCBV and rCBF values, and by significant dependence of 

PCT rCBV and rCBF results on MAP values (p < 0.001); b) 

23 PCT examinations collected in 19 patients showing PCT 

results similar to control patients; c) 33 PCT collected in 16 

patients, with low rCBV and rCBF values and near independence 

of PCT results with respect to MAP values. The 

first group was interpreted as showing impaired cerebral vascular 

autoregulation. The second group was linked to the 

first group, because eight patients went from one group to 

the other from admission to follow up; it was characterized 

by a better functional outcome. Finally, the third group was 

hypothesized to have preserved or pseudo-autoregulation. 

CONCLUSION 

Perfusion CT provides insight into cerebral vascular autoregulation 

in severe head trauma patients. It shows distinct pat- 

204 

terns possibly correlated with different kinds of traumatic 

brain edema. 

KEY WORDS: Head trauma, autoregulation, perfusion CT 


Contribution of Multidetector CT Angiography to the 

Diagnosis and Management of Vascular Injury in 

Patients with Severe Craniocervical Trauma 

Goldsher, D. 1,2 · Daitzchman, M. 1 · Eran, A. 1 · Abrantes, Y .1 · 

Shreiber, R. 1 

1 Rambam Medical Center, Haifa, ISRAEL, 2 Technion, Israel 


PURPOSE 

To assess the contribution of CT angiography (CTA) to the 

evaluation of vascular insult in patients with craniocervical 

injuries. 


Of 2321 patients admitted to our emergency department 

between January 2002 and March 2003 with craniocervical 

trauma, 73 patients underwent multidetector CTA (MCTA). 

Indications were: penetrating cervical trauma in 26/73, penetrating 

skull injuries with bony and/or metal fragments 

adjacent to major blood vessels in 38/73 and fractures at base 

of skull or at cervical spine involving pathways of major cervical 

or cranial arteries in 9/73 patients. Multidetector CTA 

studies were performed on a multidetector CT system: MX 

8000 or MX 8000 IDT (Philips Ltd) with 4 and 16 multislice 

capability, respectively. Covered volume extended from the 

aortic arch to a level above the circle of Willis in cases of 

cervical trauma, and from the C2 level to the brain convexity 

in brain injury. The images were read and interpreted by 

two neuroradiologists independently. Multidetector CTA 

results and clinical indications lead to performing digital 

subtracted angiography (DSA) in 11 patients. 

RESULTS 

Vascular pathologies related to the acute trauma were found 

in 15 patients: traumatic arterial occlusion by foreign bodies 

in 3/15 patients, arterial dissections in 9/15, and pseudoaneurysms 

in 3/15. All were diagnosed on MCTA and confirmed 

by DSA. In all cases MCTA demonstrated the type of 

the vascular injury, the anatomical location, and the relationship 

to surrounding structures. In patients with penetrating 

injuries, MCTA disclosed the route of the penetrating object 

and the damage caused along the track. 

CONCLUSION 

Multidetector CTA is suggested as a minimally invasive 

imaging modality for accurate and rapid assessment of vascular 

injuries. It is valuable for treatment and planning decision-making 

in severe craniocervical trauma patients. 

KEY WORDS: Trauma, CT angiography, vascular


Dynamic Nature of Willisian Collaterals 

Liebeskind, D. S. 1,2 · Cross, B. J. 2 · Ances, B. M. 2 · Weigele, 

J. B. 2 · Melhem, E. R. 2 · Hurst, R. W. 2 

1University of California Los Angeles, Los Angeles, CA, 

2University of Pennsylvania, Philadelphia, PA 

PURPOSE 

Willisian collaterals, or patent segments at the circle of 

Willis, may compensate rapidly for hemodynamic changes. 

Persistent blood flow in embryologic remnants such as the 

posterior communicating artery (PCOMM) has been associated 

with conflicting results regarding the risk of subsequent 

ischemia, yet serial imaging was omitted. We hypothesized 

that serial MRA may reveal evolving changes in willisian 

collaterals due to vascular events including progressive 

stenoses, occlusion, and recanalization. 


Serial TOF MRA of the circle of Willis (2-6 studies per case) 

was acquired in 64 cases of stenosis or occlusion of the basilar, 

ICA, or MCA (median age 55.5 years, range 17-79 years; 

27 men, 37 women) over varying lengths of time (median 

1.47 years, range 0.03-7.99 years). Digital measurements of 

apparent vessel diameter were calibrated to a standard nonvascular 

reference in each case. Changes in apparent vessel 

diameter were calculated for willisian segments and the 

proximal cerebral arteries. Angiographic changes were analyzed 

based on the initial primary vascular lesion and interval 

clinical or radiographic events. 

RESULTS 

Serial TOF MRA measures of apparent vessel diameter 

revealed that willisian segments and the proximal cerebral 

arteries may normally vary by up to 20%. In basilar occlusion 

(n = 2), P1 may increase by 75% and PCOMM may 

increase by up to 40%. In ICA occlusion (n = 20), progressive 

increases may be seen in ACOMM by up to 172%, ipsilateral 

or contralateral A1 by 158%, P1 by 130%, and 

PCOMM by 78%. In MCA occlusion (n = 9), ipsilateral A1 

may increase by 130%, P1 by 69%, PCOMM by 44%, and 

ACOMM by only 26%. In stenoses of the basilar (n = 5), 

ICA (n = 20), and MCA (n = 8), such increases were less pronounced. 

Willisian segments previously atretic or inapparent 

on MRA were noted to appear and progressively dilate and 

other segments were noted to decrease or even disappear 

with interval vascular events such as stroke, occlusion, worsened 

stenoses, and recanalization after dissection or revascularization. 

CONCLUSION 

Willisian segments have primary and secondary collateral 

function that may be chronicled with serial TOF MRA. 

Ischemic risk related to willisian supply likely requires serial 

evaluation. 

REFERENCES 

1. Schomer DF, Marks MP, Steinberg GK, Johnstone IM, 

Boothroyd DB, Ross MR, Pelc NJ, et al. The anatomy of the 

posterior communicating artery as a risk factor for ischemic 

cerebral infarction. N Engl J Med 1994;330(22):1565-1570 

205 

2. Rutgers DR, Klijn CJ, Kappelle LJ, van der Grond J. Recurrent 

stroke in patients with symptomatic carotid artery occlusion 

is associated with high-volume flow to the brain and 

increased collateral circulation. Stroke 2004;35(6):1345-1349 

3. Rutgers DR, Klijn CJ, Kappelle LJ, van Huffelen AC, van der 

Grond J. A longitudinal study of collateral flow patterns in 

the circle of Willis and the ophthalmic artery in patients with 

a symptomatic internal carotid artery occlusion. Stroke 

2000;31(8):1913-1920 

KEY WORDS: Collateral, angiography, stroke 


Visualization of Hemodynamics in a Silicon Aneurysm 

Model Using Time-Resolved 3D Phase-Contrast MR 

Imaging 

Isoda, H. 1 · Hirano, M. 2 · Yamashita, S. 1 · Inagawa, S. 1 · 

Kosugi, T. 3 · Alley, M. T. 4 · Markl, M. 4 · Pelc, N. J. 4 · 

Sakahara, H. 1 

1Hamamatsu University School of Medicine, Hamamatsu, 

JAPAN, 2GE Yokogawa Medical Systems, Hino, JAPAN, 

3Renaissance of Technology Corporation, Hamamatsu, 

JAPAN, 4Stanford University School of Medicine, Stanford, 

CA 

PURPOSE 

Hemodynamics affects the development and growth of 

intracranial aneurysms. If we could estimate the rupture 

potential of aneurysms based on hemodynamics then only 

patients with aneurysms that show high rupture potential 

would undergo treatment. This would be beneficial for all 

patients and be cost effective. The purpose of our study was 

to visualize hemodynamics in a silicon vascular model with 

a middle cerebral aneurysm using time-resolved 3D phasecontrast 

MR imaging (4D flow) (1). 


We obtained a rotational angiographic data set of the right 

internal carotid artery for a 67-year-old female with an 

unruptured right middle cerebral aneurysm with a fundus 

diameter of 8 mm. Based on these DICOM data, a master 

cast of the vascular lumen was produced with a 3D printer 

using powders and adhesive. From this master cast, a realistic 

silicon model, three times actual size, then was constructed. 

We ran an aqueous solution of glycerol as a flowing fluid 

through the silicon vessel model by a pulsatile pump (one 

cardiac cycle = 2 sec, maximum systolic velocity = 84 cm/s, 

Reynolds number = 810.8, Womersley number = 3.26). The 

4D-flow technique is based on a radiofrequency-spoiled gradient-echo 

sequence and it encodes flow velocity in three 

orthogonal directions. The 4D-flow technique was carried 

out using 1.5 T MR scanner with the following parameters; 

TR/TE/NEX = 5.8/2.1/1, FA = 15, FOV = 140 x 140 x 108 

mm, Matrix = 160 x 160 x 36, VENC = 20 cm/s, 20 phases 

during one cardiac cycle, imaging time = 30 min, axial 

plane, ECG trigger. Two-dimensional velocity vector fields 

on arbitrary planes, 3D stream lines and 3D particle traces 

within the aneurysm and adjacent parent arteries were calculated 

using visualization software (EnSight). 

RESULTS 

Time-resolved images of 3D stream-lines (Fig. 1A) and 3D 

particle traces clearly demonstrated that the aneurysm had 

Thursday

Thursday 

3D complex vortex flows within it during systolic phase and 

that flow streams ran out into the two adjacent M2 branches 

of the middle cerebral artery. Two-dimensional velocity vector 

fields (Fig. 1B) on arbitrary planes clearly showed timeresolved 

flow velocities within the aneurysm. 

Fig 1. Three-dimensional stream lines (A) and 2D velocity 

vector field (B) of the intracranial aneurysm model. M1, M1 

segment of MCA; M2, M2 segment of MCA. 

CONCLUSION 

In our model the 4D-flow technique provided us with timeresolved 

3D hemodynamic information about the intracranial 

aneurysm and the adjacent parent artery. This is a promising 

technique for visualizing and analyzing intraaneurysmal 

hemodynamics. 

REFERENCES 

1. Markl M, et al. Time-resolved three-dimensional phase-contrast 

MRI. J Magn Reson Imag 2003;17:499-506 

KEY WORDS: Aneurysm, hemodynamics, phase-contrast 

MR imaging 


Intracranial Aneurysms in Patients with Nonspecific 

(Takayasu) Aortoarteritis and Moyamoya Disease 

Mishra, N. · Gupta, V. · Gaikwad, S. B. · Garg, A. 

Neurosciences Center 

New Delhi, INDIA 

PURPOSE 

Nonspecific aortoarteritis (NSAA) and moyamoya disease 

are common arteritides and can cause occlusion of major 

cerebral arteries. Since these diseases occur in relatively 

younger age group, the arteries providing collateral flow are 

placed under long-term hemodynamic stress, which may 

result in formation of intracranial aneurysms. This report 

presents supportive evidence in ten such cases. 


We retrospectively evaluated cerebral angiograms done at 

our center in the last 5 years and reviewed clinical and radiologic 

profile of cases in which intracranial aneurysms were 

associated with NSAA or moyamoya disease. 

RESULTS 

Ten aneurysms were seen in 8 patients (male 5, female 3; age 

14 to 47 years). All patients presented with sudden onset 

severe headache [Hunt & Hess grade I (n-1), II (n-3), III (n-2) 

or IV (n-2)]. CT revealed subarachnoid hemorrhage (all 

cases), intraventricular blood (n-5) and adjacent parenchymal 

hematoma (n-2). None of the patients with moyamoya had any 

206 

clinical/radiologic evidence of cerebral ischemia. One patient 

with NSAA had hemiparesis due to previous middle cerebral 

territory infarction. Two patients with NSAA were hypertensive, 

which was well controlled by drugs in one of them. 

Digital subtraction angiography revealed moyamoya (n-4), 

NSAA (n-2) and moyamoya pattern along with aortoarteritis 

(n-1). Evidence of collateral flow from posterior to the anterior 

circulation was seen in all cases. Stenosis/occlusion of aortic 

arch vessels was seen in all cases of NSAA along with 

renal (n-2) and abdominal aorta involvement (n-2). The 

intracranial aneurysms were located in P1 segment of posterior 

cerebral artery (n-4), basilar top (n-2), posterior inferior 

cerebellar artery (n-2), anterior inferior cerebellar artery (n-1), 

and plexal segment of lateral posterior choroidal artery (n-1). 

Six aneurysms in 5 patients were treated by endovascular coiling 

with complete occlusion in 4 and incomplete (> 90%) in 1 

case. No procedure-related complications occurred except one 

where development of thrombus necessitated use of abciximab 

and no clinical consequences ensued. This patient was 

later detected to have heparin-induced thrombocytopenia 

(HIT). Surgical treatment in one case was complicated by 

intraoperative rupture of the aneurysm and transient hemiparesis. 

Two patients were not treated because of poor clinical 

grade and/or inaccessible location. One small, unruptured 

PCA aneurysm also was not treated. Follow-up angiograms in 

3 patients revealed stable occlusion of all treated aneurysms. 

Follow-up MRA in the patient of HIT at 6 months also 

revealed effective obliteration of the treated aneurysms. Two 

patients are on wait for follow-up study. 

CONCLUSION 

Moyamoya and nonspecific aortoarteritis (Takayasu arteritis) 

can be associated with posterior circulation intracranial 

aneurysms. Location of the aneurysms indicates that they are 

possibly related to hemodynamic stress of collateral flow, 

rather than the primary disease process. Most of these 

aneurysms can be treated safely by endovascular techniques. 

Rarely, moyamoya pattern can be seen in patients with nonspecific 

aortoarteritis, as seen in one of our cases. 

KEY WORDS: Aneurysm, aortoarteritis, moyamoya 


In Vivo Use of Optical Coherence Tomography in the 

Assessment of Aneurysm Healing 

Masaryk, T. J. 1 · Thorell, W. 2 · Chow, M. 3 · Huang, D. 4 · 

Prayson, R. 1 · Woo, H. 1 · Rasmussen, P. 1 

1 Cleveland Clinic Foundation, Cleveland, OH, 2 University of 

Nebraska Medical Center, Omaha, NE, 3 Mackenzie Health 

Sciences Centre/University of Alberta, Edmonton, AB, 

CANADA, 4 University of Southern California, Los Angeles, 

CA 

PURPOSE 

Optical coherence tomography (OCT) is an imaging modality 

initially described by Huang, et al. in 1991 utilizing 

backscattered light to produce high-resolution tomography 

of optically accessible biologic tissues such as the eye, blood 

vessels or the GI tract. The purpose of this study was: 1) To 

define the ability of OCT to image an animal model of cerebral 

aneurysms using an in vivo imaging probe and 2) To 

correlate in vivo OCT images with histologic findings.


Four dogs underwent creation of two venous pouch sidewall 

carotid aneurysms with subsequent embolization (Rx) using 

GDC coils for one, and Matrix coils for the other 

(BSCI/Target, Fremont, CA). The animals were sacrificed 2, 

4, 6, and 8 weeks after Rx. In vivo OCT was performed 

immediately after Rx and just prior to sacrifice using an 

imaging wire/probe 0.014” in diameter. Presacrifice 

angiograms were evaluated for: coil compaction and radiolucency 

at the neck. Optical coherence tomography images 

were evaluated for: visualization of the neck, visualization of 

coils and/or coating, and the presence of tissue response. The 

aneurysms were harvested en bloc, imbedded in methylmethacrylate, 

cut, polished, and surface stained with H&E. 

RESULTS 

In vivo OCT demonstrated the neck of experimental 

aneurysms, the pattern of coils at the neck of treated 

aneurysms, evolving thrombus, dissection and progressive 

healing across the neck of Rx aneurysms; both those treated 

with bare platinum and biopolymer coated coils. 

CONCLUSION 

This study demonstrated high-resolution in vivo OCT 

images produced using an 0.014” imaging probe depicting 

the endovascular surface of an aneurysm neck, the pattern of 

coils within the aneurysm sac, and the associated tissue 

growth/healing over time. Optical coherence tomography 

was able to distinguish histologic changes seen at the 

aneurysm neck which were not seen with angiography. 

Further in vivo animal testing may refine these techniques to 

improve image quality and allow comparison to other 

endovascular imaging methods. With further development, 

OCT may be a useful adjunct in the treatment of cerebrovascular 

disease, and in particular, defining the tissue response 

to implantable vascular devices over time. 

KEY WORDS: Optical coherence tomography, aneurysm, 

healing 


Imaging of Arteriovenous Malformations with Phase- 

Contrast VIPR: An Ultrafast Phase-Contrast MR 

Angiography Technique Superior to 3D Time-of-Flight 

Durick, N. · McCue, J. · Turk, A. · Rowley, H. · Turski, P. · 

Gu, T. · Mistretta, C. 


Madison, WI 

PURPOSE 

Phase-contrast VIPR is a novel approach under investigation 

for use in MR angiography (MRA). This technique uses 

undersampling and projection reconstruction with increased 

imaging speed by a factor of 10-30. Using this technique the 

traditional artifacts associated with 3D phase contrast and 

undersampling are reduced with the large volume acquisition 

and 3D technique. The goal of this project is to compare 

image quality and artifacts of PC VIPR with traditional 3D 

time-of-flight (TOF) and 2D phase contrast (PC) MRA 

imaging in the evaluation of patients with arteriovenous malformations 

(AVMs) with digital subtraction angiography 

(DSA) being the standard. 

207 


Nine patients with previously diagnosed AVMs were 

referred for clinical MR imaging to be used in treatment 

planning and staging. Each patient received standard brain 

MR imaging including 3D TOF and 2D PC as well as PC 

VIPR MRA sequences. Arteriovenous malformation grading 

and qualitative assessment was performed by 3 blinded CAQ 

neuroradiologists and assessed for anatomical accuracy and 

artifacts. MR images then were compared with DSA images 

which were available in 6/9 cases. The sequence was implemented 

on a 1.5 T scanner (GE Medical Systems, 

Milwaukee, WI) with a single-receiver head coil so that each 

VIPR projection was excited 4 times, once with no-flow 

encoding gradient and once with encoding in the x, y, and z 

directions. The reference and the three flow-direction excitation 

data were reconstructed for the three flow directions. 

Parameters for the PC VIPR were 0.63 mm x 0.63 mm x 0.63 

mm voxel size compared to 0.86 mm x 0.74 mm x 1.4 mm, 

TE of 7.5 ms for PC VIPR vs 2.4 for 3D TOF. The scan time 

for a 3-slab 3D TOF was 8:30 minutes and for the PC VIPR 

was 5:00 minutes. Informed consent from each patient and 

IRB approval was obtained. 

RESULTS 

In all 9/9 AVM cases PC VIPR technique yielded results 

favorable to conventional 3D TOF and 2D PC MRA imaging. 

Phase-contrast VIPR images were more accurate in 

anatomical depiction of the size and extent of AVMs when 

compared to 3D TOF and 2D PC MRA techniques. In all 

cases PC VIPR provided accurate grading and simulated digital 

subtraction angiographic images. The artifacts resulting 

from severely undersampling in projection acquisition did 

not detract from image quality or interpretation. Phase-contrast 

VIPR acquisition time (5:00 minutes) was significantly 

faster that 3D TOF and 2D PC sequences (8:30 minutes and 

2:30 minutes per plane of acquisition). 

CONCLUSION 

MR PC VIPR technique simulates digital subtraction 

angiography and compares favorably to traditional 3D TOF 

and 2D PC MRA imaging in patients with AVMs. Phase-contrast 

VIPR yields superior anatomical depiction through 

improved detection of high- and low-flow vasculature and 

improved large and small vessel visualization in all cases 

with decreased imaging time. Further evaluation is warranted 

to determine additional clinical applications and potential 

utilization of the anatomical and physiologic data obtainable 

by this method. 

KEY WORDS: Arteriovenous malformations 

Thursday

Thursday 


Visualization of Hemodynamics in Intracranial Arteries 

Using Time-Resolved 3D Phase-Contrast MR Imaging 

Yamashita, S. 1 · Haruo, I. 1 · Hirano, M. 2 · Inagawa, S. 1 · 

Takehara, Y. 1 · Alley, M. T. 3 · Markl, M. 3 · Pelc, N. J. 3 · 




3Stanford University School of Medicine, Stanford, CA 

PURPOSE 

Hemodynamic factors such as flow velocity or shear stress 

of the arterial wall affect the development of various vascular 

lesions, such as aneurysms and atherosclerosis. If we 

could analyze the hemodynamics of intracranial arteries precisely, 

we could predict the occurrence of aneurysms or atherosclerotic 

lesions. The purpose of our study was to visualize 

hemodynamics of intracranial arteries using timeresolved 

3D phase-contrast MR imaging (4D-flow) (1). 


MR examinations for five healthy volunteers (age, 23-47 

years old; 32.4 years old on average) were performed with a 

1.5 T MR unit. The 4D-flow was based on radio frequencyspoiled 

gradient-echo sequence and velocity encoding was 

performed along all three spatial directions. Four-dimensional 

data including time dimension were obtained. 

Measurements were gated retrospectively to the electrocardiogram 

cycle and CINE series of 3D data sets were generated. 

Utilized imaging parameters were as follows; 

TR/TE/NEX = 5.2-5.4/2.3-2.7/1, FA = 15, RBW = 62.5 kHz, 

FOV = 16 cm, Matrix = 160 x 160, slice thickness = 1-2 mm, 

number of slices = 12-40, VENC = 60 cm/sec, acquisition 

time = 20-40 minutes, Slew Rate = 120 T/m/ms. Timeresolved 

images of 3D stream lines, 3D particle traces, and 

2D velocity vector fields on arbitrary planes were calculated 

from 4D data sets by flow visualization software (EnSight). 

RESULTS 

When we generated 3D stream lines originating from the 

bilateral C1-2 segment of internal carotid arteries and the 

basilar artery we were able to see the 3D stream lines from 

the circle of Willis to the bilateral M2 segment of middle 

cerebral arteries (Fig. 1). Time-resolved images of 3D particle 

traces also clearly demonstrated intracranial arterial flow 

dynamics. Two-dimensional velocity vector fields on the 

planes traversing carotid siphon or basilar tip were clearly 

visualized (Fig. 2). These images could be observed as CINE 

images. These results were obtained for all five volunteers. 

208 

CONCLUSION 

The 4D-flow technique helped us understand 3D in vivo 

hemodynamics in the human intracranial arteries. This 

method might be a useful and promising noninvasive method 

for analyzing in vivo hemodynamics of intracranial arteries. 

The hemodynamic information provided by this technique 

hopefully will enable us to predict the risk of aneurysmal 

rupture or occurrence of atherosclerotic plaques. We then 

could provide tailor-made therapies and managements for 

each patient. 

REFERENCES 



KEY WORDS: Intracranial arteries, hemodynamics, phasecontrast 

MR imaging 


Contribution of CT Angiography to the Diagnosis of 

Dissection in the Craniocervical Arteries 

Goldsher, D. 1,2 · Abrantes, Y. 1 · Daitzchman, M. 1 · Eran, A. 1 · 

Shreiber, R. 1 

1 2 Rambam Medical Center, Haifa, ISRAEL, Technion, Israel 


PURPOSE 

To evaluate the ability of multislice CT angiography 

(MCTA) to detect and diagnose dissection of major cervical 

or cranial arteries.


Routine CT angiography studies of the craniocervical arteries 

were performed on 279 patients admitted to our emergency 

department over an 18-month period. The patients 

presented with stroke, acute spontaneous SAH, penetrating 

cervical trauma, fractured base of skull or cervical vertebrae, 

suspected aortic dissection, or pulsating cervical mass. CT 

angiography was performed using MX8000 and MX8000 

IDT (Philips) with 4- and 16-slice capability, respectively. 

The imaged volume extended from the aortic arch to a level 

above the circle of Willis. The diagnosis of arterial dissection 

was confirmed with MR imaging, DSA, or surgery. MR 

imaging studies were performed on 0.5 T (Gyrex) and 1.5 T 

(Signa) MR imaging systems, GE (Elscint). DS angiography 

studies were performed using a Multistar system (Siemens). 

RESULTS 

Seventeen dissections were detected in the craniocervical 

arteries in 15 patients, 9 males and 6 females, 8-74 years of 

age. Eight dissections of carotid arteries were diagnosed in 7 

patients: four were located at the cervical segment of the 

internal carotid artery (ICA) and two were at the carotid 

siphon. The seventh patient had bilateral dissection of the 

common carotid arteries extending from the dissected aortic 

arch. Dissected vertebral arteries were diagnosed in 6 

patients, 3 along extracranial segments: at the C1-2 level in 

two and the C3 level in one. Another 4 dissections (in 3 

patients) were located at intracranial segments of vertebral 

arteries. Four patients had dissected basilar arteries. A dissecting 

aneurysm of the right PICA was demonstrated in 

another patient. In 12 patients, the diagnosis was confirmed 

using DSA or MR imaging. Multislice CT angiography 

demonstrated the narrowed true lumen of the dissected arteries 

and pointed to the exact location of the dissection. The 

intramural hematoma indicating the extension of the dissected 

segment was demonstrated in the affected carotid and 

basilar arteries but was not depicted by MCTA in the vertebral 

arteries in our series. 

CONCLUSION 

Multislice CT angiography is a valuable noninvasive tool for 

the detection and rapid evaluation of dissected major craniocervical 

arteries. It can contribute to treatment planning and 

also used for planning diagnostic and therapeutic DSA when 

necessary. 

KEY WORDS: CT angiography, dissection, cerebral arteries 


Double Origin of the Posterior Inferior Cerebellar 

Artery, Vertebrobasilar Artery Junction Fenestration, 

Trigeminal Artery and Aneurysm: Case Illustration of a 

Previously Unreported Constellation of Rare Anomalies 

and Variants 

Case, R. S. · Lesley, W. S. 

Texas A&M University/Scott & White Clinic 

Temple, TX 

Double origin of the posterior inferior cerebellar artery 

(PICA) has been demonstrated previously by angiography 

only four times in the peer-reviewed literature. None of the 

209 

prior reports describe this PICA variant in conjunction with 

either a trigeminal artery aneurysm or a fenestrated vertebrobasilar 

junction—all of which were present in this 

patient. Overall, a 40% incidence exists for the presence of 

intracranial saccular aneurysm in patients with a double origin 

PICA. 

KEY WORDS: Double origin pica, aneurysm, trigeminal 

artery 


10:00 AM - 11:30 AM 

Room 205 


(Scientific Papers 386 - 400B) 






Moderators: Kelly K. Koeller, MD 

TBD 


Trigeminal and Concurrent Glossopharyngeal Neuralgia 

Secondary to Lateral Medullary Infarction 

Warren, H. G. · Kotsenas, A. L. · Czervionke, L. F. 

Mayo Clinic 

Jacksonville, FL 

PURPOSE 

Most patients presenting with trigeminal neuralgia will have 

disease involving the trigeminal nerve or ganglion or the primary 

sensory nucleus in the pons. We discuss the unusual 

finding of lateral medullary infarction associated with concurrent 

trigeminal and glossopharyngeal neuralgia. This case 

illustrates the need to carefully examine the entire trigeminal 

nerve pathway, including the medulla and upper cervical 

cord, with MR imaging in patients presenting with trigeminal 

neuralgia. 


A 69-year-old woman developed sudden left facial pain in 

her left temple extending into the left ear, cheek, and jaw. 

The pain was intense, burning, and sharp and was exacerbated 

by light touch. Emergency CT and subsequent MR imaging 

(1 week later) performed at an outside institution were 

reported to be normal. The patient was treated empirically 

with valacyclovir hydrochloride for suspected herpes zoster 

without relief. Subsequent trials of gabapentin 600 mg three 

times daily and oxycarbazepine 300 mg twice daily also pro- 

Thursday

Thursday 

vided no relief and the patient was referred to our institution 

for consideration for trigeminal nerve ablation. She underwent 

high-resolution gadolinium-enhanced brain MR imaging 

with diffusion-weighted and 3D-CISS sequences. 

RESULTS 

MR imaging demonstrated linear increased signal intensity 

on T2-weighted image in the lateral left medulla in the left 

spinotrigeminal nucleus and tract (Figure) and left solitary 

nucleus. Diffusion-weighted imaging was normal and there 

was no enhancement with gadolinium. No vascular compression 

of the trigeminal nerve root was identified on 3D- 

CISS images. These findings were consistent with a chronic 

left lateral medullary infarction involving the left spinotrigeminal 

nucleus and tract, nucleus ambiguous, and solitary 

nucleus. 

CONCLUSION 

Trigeminal neuralgia (TN) presenting as a consequence of 

lateral medullary vascular compromise is extremely rare. We 

are aware of only a single previously reported case which 

was not associated with concurrent glossopharyngeal neuralgia. 

TN most often involves the cisternal portions or root 

entry zones of the nerves or the primary sensory nucleus in 

the pons. When a structural lesion is suspected, advanced 

MR imaging plays an important role, identifying vascular 

compression of the root entry zone, the cause of 80-90% of 

cases. This case illustrates the need to image the entire extent 

of the trigeminal neural pathway into the upper cervical 

spinal cord to identify structural causes of TN, especially 

when associated with other cranial nerve symptoms. 

KEY WORDS: Trigeminal neuralgia, glossopharyngeal neuralgia, 

lateral medullary infarction 


Case Report of Multicystic Cavernous Venous 

Malformation (Angioma) of the Cerebellopontine Angle 

and Review of the Literature 

Miller, A. 1 · Aulino, J. M. 1 · Chambers, M. R. 2 · Johnson, M. 

D. 3 · Allen, G. S. 1 

1 Vanderbilt University Medical Center, Nashville, TN, 

2 University of Alabama, Birmingham, AL, 3 University of 

Tennessee Medical Center, Knoxville, TN 

PURPOSE 

Report an unusual case of multicystic cavernous venous malformation 

(angioma) of the cerebellopontine angle (CPA); 

210 

describe the imaging findings that commonly are confused 

with more common tumors, such as vestibular schwannoma; 

and summarize histologic categorization of venous malformations. 

Brief review of the literature is discussed. 


A 57-year-old man presented with 2-month history of profound 

right-sided hearing loss. Patient became progressively 

more unsteady during ambulation with occasional veering 

toward the right. Neurologic symptoms included perioral tingling 

and numbness, decreased sensation of the right tongue, 

and a sensation of coolness on his right chin. The patient 

underwent stereotactic right suboccipital craniotomy using a 

far lateral transcondylar approach, and microsurgical excision 

of the mass. Patient did not suffer any additional loss of 

function of the seventh or eighth cranial nerves during or following 

surgery. Histologic diagnosis was cystic cavernous 

angioma of the CPA. 

RESULTS 

Head CT showed low attenuation extraaxial mass in the right 

CPA which displaced the brainstem toward the left. Small 

flecks of bone or calcium were identified within the mass. 

MR imaging showed a 3.9 cm multilocular cystic mass in the 

right cerebellopontine angle which had curvilinear peripheral 

and internal enhancement. The right internal auditory 

canal was spared. The fourth ventricle was compressed by 

this mass with secondary dilatation of the third and lateral 

ventricles. 

CONCLUSION 

Cavernous venous malformation (angioma) is a rare lesion 

of the CPA. This case describes the atypical imaging appearance 

and clinical presentation of a cavernous angioma, subsequent 

surgical treatment, and review of the literature. The 

classic presentation of a temporal bone cavernous venous 

angioma is described as having rapid worsening of symptoms, 

in particular hearing loss and cranial nerve VII 

deficits, with imaging findings of a small, enhancing internal 

auditory canal mass. However, this “classic” presentation 

may be pertinent only for cavernous venous malformation of 

the internal auditory canal and not the CPA. This patient had 

progressive symptoms for months prior to presentation. The 

more common MR appearance of the cavernous venous malformation 

is hyperintense T1 and T2 signal with heterogeneous 

enhancement. Occasionally a hemosiderin rim is visualized. 

A cystic-appearing cavernous venous malformation is 

rare. The term cavernous hemangioma is an outdated name 

for this entity. The more appropriate histologic term is cavernous 

venous malformation. Temporal bone venous hemangiomas 

have four histologic types: capillary, cavernous, 

mixed capillary-cavernous, and venous. The histologic differentiation 

is based upon the appearance of the vascular 

spaces. 

KEY WORDS: Cavernous angioma, cerebellopontine angle, 

literature review


De Novo Formation of a Cavernous Angioma Adjacent to 

a Developmental Venous Anomaly in a Patient with 

Factor V Leiden Deficiency 

Ervine, S. L. M. · Kim, J. K. 

Oakwood Hospital 

Dearborn, MI 

PURPOSE 

Many theories have been proposed regarding the de novo 

formation of cavernous angiomas (CA). One theory involves 

the formation of CA from developmental venous anomalies 

(DVA), which frequently coexist (1). We describe the development 

of a CA adjacent to a DVA in a patient with factor V 

Leiden deficiency (FVL) and sticky platelet syndrome. 


While at work, a 41-year-old female with a history of FVL 

and sticky platelet syndrome developed a headache with leftsided 

weakness, left facial droop, and blurred vision in her 

left eye. She presented to our emergency department with 

persistence of symptoms for several hours. 

RESULTS 

Noncontrast CT of the head was negative. MR imaging 

revealed a lesion in the right occipital lobe consistent with a 

CA. Adjacent to this focus was a DVA. Although the DVA 

was clearly present on the patient’s prior MR imaging 

obtained 2 years prior, the CA was not present at that time, 

even in retrospect. In this patient with FVL and sticky 

platelet syndrome, we suspect that interval venous thrombosis 

led to formation of a CA. 

211 

CONCLUSION 

Many etiologies have been suggested to explain the frequent 

coexistence of DVAs and CAs. One proposed theory for 

development of CAs from DVAs is that increased venous 

pressure due to venous restrictive disease could lead to hemorrhage 

from a DVA, with subsequent development of a cavernous 

malformation (1). Our patient has a plausible etiology 

for development of venous restriction, in that she has two 

thrombotic disorders, FVL, and sticky platelet syndrome. We 

suspect that in our patient one or more thrombotic events 

occurred related to the DVA, leading to venous stenosis, elevated 

venous pressure, hemorrhage, and subsequent cavernous 

angioma formation. 

REFERENCES 

1. Maedar, et al. Development of a cavernous malformation of 

the brain. AJNR Am J Neuroradiol 1998;19:1141-1145 

KEY WORDS: Cavernous angioma, developmental venous 

anomaly, vascular malformation 


MR Findings in a Sporadic Case of Meningeal 

Angiomatosis 

Shekdar, K. V. · Zimmerman, R. A. · Rorke-Adams, L. B. · 

Storm, P. B. 



PURPOSE 

To describe characteristic MR features of a sporadic case of 

meningeal angiomatosis, in the clinical setting of uncontrollable 

focal seizures. 

Thursday

Thursday 


A four-year-old, previously healthy female, presented with 

new onset left-sided facial seizures. The seizure frequency 

increased in a short period of 2 weeks to intractable focal 

seizures. A MR study was obtained. Based on the MR imaging 

features the following possibilities were considered: a 

focal cortical neoplasm/dysplasia, dural vascular malformation, 

and a focal meningeal angiomatosis. The patient underwent 

a near-total surgical resection. The gross and microscopic 

features were in favor of meningeal angiomatosis. 

Following resection the patient’s seizures are under control. 

RESULTS 

MR study revealed mild expansion within a focal area of the 

right perisylvian cortex. Abnormal stippled hypointense signal 

was seen on the T2 and FLAIR images with a poorly 

defined area of surrounding hyperintensity. There was 

increased susceptibility, on the gradient-echo images, in this 

focally expanded cortex. Prominent vessels were seen adjacent 

to this lesion. No obvious enhancement was appreciated 

in the lesion. However, enhancement was noted in the prominent 

vessels around the lesion and in the adjacent cortex. 

CONCLUSION 

Meningeal angiomatosis, most commonly presents with symptoms 

of uncontrollable partial, focal seizures. Although an association 

with NF-2 has been described, about two thirds of the 

cases are sporadic in nature. Imaging findings are fairly typical 

with primary involvement of the cortical gray matter which 

reflects the pathologic changes of a focal leptomeningeal and 

meningovascular proliferation. The differential diagnoses 

include cortical dysplasia/tumor, dural arteriovenous malformation, 

Sturge-Weber syndrome, and meningiomas. 

KEY WORDS: Meningeal, angiomatosis, epilepsy 


Disorders of Intracranial Pressure: The Dynamic 

Appearance of the Dural Venous Sinuses 

Farb, R. I. · Baryshnik, D. 

Toronto Western Hospital 


PURPOSE 

Idiopathic intracranial hypertension (IIH) and spontaneous 

intracranial hypotension (SIH) represent entities positioned 

at opposite ends of the spectrum of primary disorders of 

intracranial pressure. These syndromes are associated with 

characteristic changes in the appearance of the dural venous 

sinuses (DVS) (1, 2) that are well demonstrated at auto-triggered 

gadolinium-enhanced MR venography (ATECO 

MRV). We present examples indicative of the remarkable 

change in the caliber of the DVS following treatment of 

these disorders and normalization of intracranial pressure. 


Typical case of IIH: A 21-year-old woman with known IIH 

presented with headache, blurred vision, and papilledema 

despite maximal medical therapy. An MR image of the brain 

was normal. An ATECO MRV demonstrated the typical narrowed 

appearance of the distal transverse sinuses characteristic 

of IIH (1). A ventriculo-peritoneal shunt was inserted, 

212 

resulting in improvement in headache and vision. At a follow-up 

exam the CSF opening pressure normalized and 

repeat ATECO MRV revealed normal-appearing DVS with 

resolution of the transverse sinus narrowing. Typical case of 

SIH: A previously healthy 19-year-old woman presented 

with headache and diplopia, nausea, vomiting, photophobia, 

sonophobia, and bilateral abducens palsies. MR imaging 

revealed morphologic findings typical of SIH (3). An 

ATECO MRV also demonstrated marked dilatation of the 

DVS which also has been described in this syndrome (2). 

The patient was treated with an epidural blood patch procedure. 

Her headache improved and she was discharged home. 

Follow-up MR imaging and ATECO MRV revealed resolution 

of the stigmata of intracranial hypotension. Moreover, 

the dural venous sinuses returned to normal caliber. 

RESULTS 

ATECO MRV enables noninvasive visualization of the 

intracranial venous system in greater detail than previously 

has been available (4). The dynamic changes seen in the caliber 

of the DVS in disorders of intracranial pressure can be 

anticipated by the application of the Monro-Kellie doctrine 

governing intracranial compartmental volumes (5). The 

increase in CSF pressure seen in IIH is compensated for by a 

decrease in the intracranial venous volume. The dural venous 

sinuses act as an intracranial volume/pressure reservoir that 

can accommodate these changes in pressure. This is seen on 

MRV by apparent venous sinus narrowing in patients with 

IIH. This sinus narrowing in IIH is not fixed — the appearance 

of the distal transverse sinus returns to normal when the ICP 

normalizes. Conversely, in the syndrome of SIH a decrease in 

CSF volume (and thus decrease in intracranial pressure) is 

accompanied by an increase in the intracranial venous volume 

secondarily. This is seen on MRV as enlargement of the dural 

venous sinuses that also normalizes following treatment. 

CONCLUSION 

There is a direct and dynamic relationship between intracranial 

pressure and the appearance of the DVS. We have found 

this finding to be highly useful in both the diagnosis and the 

management of patients with disordered intracranial pressure. 

REFERENCES 

1. Farb RI, et al. Neurology 2003;60(9):1418-1424 

2. Dillon WP, et al. AJNR Am J Neuroradiol 1998;19(6):1001-1002 

3. Chung SJ, et al. Neurology 2000;55(9):1321-1327 

4. Farb R, et al. Radiology 2003;206(1):203-209 

5. Mokri B. Neurology 2001;56(12):1746-1748 

KEY WORDS: Pseudotumor cerebri, intracranial hypotension, 

MR venography 


Anomalous Migration of the Adenohypophysis 

Grimme, J. D. 1 · Marsot-Dupuch, K. 2 · Smoker, W. R. K. 3 · 

Castillo, M. 1 

1 2 University of North Carolina, Chapel Hill, NC, University 

of Paris South, Paris, FRANCE, 3University of Iowa, Iowa 

City, IA 

PURPOSE 

To present the imaging findings and discuss the possible etiology 

of anomalous migration of the anterior pituitary lobe.


We retrospectively analyzed the CT and MR imaging findings 

in three patients with abnormalities assumed to represent 

anomalous migration of the anterior pituitary lobe. All 

patients received their examinations for reasons other than 

endocrine abnormalities. 

RESULTS 

All three patients (two adults, one newborn, two males) were 

endocrinologically normal. In all, MR imaging showed that 

the intrasellar pituitary gland consisted of only the bright 

neurohypophysis. The sella was small and the stalk was seen 

in all. Within the sphenoid sinus, and surrounded by bone we 

identified a soft tissue area of signal intensity similar to gray 

matter. This presumed adenohypophysis was located within 

an intersphenoid septum in two cases and in a craniopharyngeal 

canal in the newborn. The anterior and posterior lobes 

have different embryologic origins. The anterior lobe is ectodermal, 

and develops from Rathke’s pouch. The posterior 

lobe is of neuroectodermal origin and extends inferiorly as 

the infundibulum from the diencephalon (1). The two 

migrate and fuse to form a single gland. Migrational arrest of 

the adenohypophysis occurs in the sphenoid bone within the 

craniopharyngeal canal, and may present as a small soft tissue 

mass adjacent to the intersphenoidal septum (2). The 

ectopic neurohypophysis may be located anywhere between 

the median eminence and the superior aspect of the pituitary 

gland within the sella turcica (3). While anterior and posterior 

hypophyseal anomalies may be associated with other 

structural or functional anomalies, they may occur in isolation, 

and each may occur in the presence of a normal location 

of the other, lending credence to independent migration 

of the two components. 

CONCLUSION 

Findings of absent intrasellar adenohypophysis, intrasellar or 

ectopic neurohypophysis, small sella, and a small soft tissue 

area within the sphenoid sinus are typical of anomalous 

migration of the pituitary gland. 

REFERENCES 

1. Martin JH. The hypothalamus and the regulation of 

endocrine and visceral functions. In: Dolan JJ, ed. 

Neuroanatomy: Text and Atlas. Stamford: Appleton and 

Lange;1996:422-423 

2. Marsot-Dupuch K, Smoker WRK, and Grauer W. A rare 

expression of neural crest disorders: an intrasphenoidal 

development of the anterior pituitary gland. AJNR Am J 

Neuroradiol 2004;25:285-288 

3. Mitchell LA, Thomas PQ, Zacharin MR, and Scheffer IE. 

Ectopic posterior pituitary lobe and periventricular heterotopia: 

cerebral malformations with the same underlying 

mechanism? AJNR Am J Neuroradiol 2002;23:1475-1481 

KEY WORDS: Adenohypophysis, developmental anomaly 

213 


Petrified Ears in a Case of Keutel Syndrome: Imaging 

Findings 

Parmar, H. A. · Blaser, S. · Unger, S. · Papsin, B. 



PURPOSE 

We present unusual imaging findings of petrified ears in a 

case of Keutel syndrome. 


A 9-year-old girl, born of consanguineous marriage, presented 

with long standing hearing loss. Audiogram showed significant 

mixed, but predominantly conductive hearing loss, 

which was corrected with hearing aids. She had an underprojected 

nasal tip with occlusion of the right nostril and palpable 

hardening of the cartilages in the pinnae, probably due 

to dysplastic cartilage formation. On further evaluation she 

was found to have thick mitral valve leaflets with mild mitral 

and aortic regurgitation and shortening of distal phalanges 

(brachytelephalangism). Peripheral pulmonary stenosis also 

was strongly suspected. 

RESULTS 

CT scan of the petrous temporal bones showed small and 

hypoplastic cochlea bilaterally with small cochlear nerve 

canals. There was also extensive calcification of both auricular 

cartilages, a feature known as “petrified ears.” 

Radiograph of the chest showed extensive calcification of 

the tracheo-bronchial tree. Previously reported literature on 

petrified ears was reviewed and amongst the commonest 

causes mentioned were frost bite or familial insensitivity to 

cold, physical trauma, radiation injury, endocrine abnormalities 

(Addison’s disease, acromegaly, hypothyroidism), 

inflammatory conditions (polychondritis, perichondritis, 

syphilitic perichondritis), radiation injury, insect bites and 

systemic chondromalacia. Rare causes include Keutel syndrome, 

an autosomal recessive disorder characterized by diffuse 

cartilage calcification and more commonly seen in 

patients from the middle-eastern countries. 

Thursday

Thursday 

CONCLUSION 

Keutel syndrome is a rare cause of petrified ears but should 

be considered in the differential diagnosis particularly in 

association with calcification of other cartilage such as tracheo-bronchial 

tree and features of hearing loss, brachytelephalangism 

and peripheral pulmonary stenosis. CT scan of 

the petrous bones helps to delineate the extent of cochlear 

anomaly in children with hearing loss. The pinna calcifications 

in this child were an unexpected finding and led to her 

diagnosis of Keutel syndrome. 

KEY WORDS: Keutel syndrome, pinna calcification, hearing 

loss 


Central Nervous System Extraosseous Ewing’s Sarcoma: 

A Newly Defined Pathologic Entity 

Pekala, J. S. · Gururangan, S. · Mukundan, S. 


Durham, NC 

PURPOSE 

Primary or secondary central nervous system extraosseous 

Ewing’s sarcoma (CNS-EES) is a newly defined pathologic 

entity, previously undescribed in the neuroradiology literature. 

Proper diagnosis of CNS-EES is made via molecular 

genetic analysis using reverse transcriptase polymerase 

chain reaction (RT-PCR) or fluorescence in situ hybridization 

(FISH) of biopsy samples. Although CNS-EES is histologically 

similar to central primitive neuroectodermal 

tumors (c-PNET) such as medulloblastoma, pineoblastoma, 

and supratentorial PNET, the clinical behavior, treatment, 

and prognosis of CNS-EES is vastly different. 


Case 1: An 8-year-old female presented with headache, nausea, 

and vomiting of 4 months duration. Following imaging, patient 

underwent tumor resection with RT-PCR and FISH analysis 

diagnosing CNS-EES. Patient was treated with chemotherapy 

and radiotherapy, and is currently alive with no evidence of disease 

20 months following diagnosis. Case 2: A 7-year-old 

female presented with 1-month history of early morning 

headache, and vomiting. Following imaging, she underwent 

tumor resection, with CNS-EES diagnosis confirmed by FISH. 

Subsequently, a pulmonary parenchymal mass was found, 

resected and proven to be EES. The patient was started on 

chemotherapy, and also will receive radiation therapy. 

RESULTS 

Case 1: MR brain scan revealed an extraaxial enhancing 

mass arising along the tentorium and compressing the cerebellar 

vermis and quadrigeminal plate. Case 2: MR brain 

scan showed 8 cm T1 hyperintense lesion within the 

parenchyma of the right frontal lobe consistent with hemorrhagic 

mass. As this mass appeared to arise in the parenchyma 

and not the dura, a search for a primary lesion commenced. 

CNS metastatic workup was negative, but chest CT 

showed 2.8 cm calcified left lower lobe nodule, abutting the 

pleura but not involving the adjacent rib. A review of both 

our cases and the four reported cases of CNS-EES in the 

pathology literature revealed all lesions to be solitary well 

circumscribed intensely enhancing lobulated masses. In five 

214 

of six cases, the masses appeared to arise from the dura. 

Variable findings included hyperintensity on diffusionweighted 

imaging, as well as the presence of hemorrhagic 

and cystic components. 

CONCLUSION 

Typical PNETs (e.g. medulloblastoma, pineoblastoma, or cerebral 

PNET) found in the CNS exhibit central markers and are 

designated as c-PNETs. Primary or secondary extraosseus 

Ewing’s sarcoma (PNETs with peripheral markers identified 

by FISH or RT-PCR) when found in the CNS are designated 

CNS-EES. Because the clinical and biologic behavior of c- 

PNET is very different from that CNS-EES, correct diagnosis 

must be made to ensure appropriate therapy and prognosis. 

Knowledge of CNS-EES enables the neuroradiologist to suggest 

the proper diagnosis and need for special pathologic tests 

to the referring oncologist. This is most important when a purported 

PNET is observed to abut a dural reflection. 

KEY WORDS: Pediatric brain tumor, Ewing’s sarcoma, 

peripheral PNET 


Report of a Rare Case of High-Grade Leiomyosarcoma 

of the Skull 

Ketkar, M. · Rodriguez, J. · Karim, A. · Fowler, M. · 

Hardjasudarma, M. · Nanda, A. · Gonzalez Toledo, E. 

Louisiana State University Health Sciences Center 

Shreveport, LA 

PURPOSE 

To report a rare case of high-grade leiomyosarcoma of the 

skull. 


A 26-year-old right-handed male presented to the emergency 

department with headache and a 6-month history of growth 

in the right scalp. 

RESULTS 

The CT scan of head and MR imaging of head (Fig. 1) 

revealed an expansile lesion involving the right parietooccipital 

bone associated with a large soft tissue mass and 

extending into the dura. However no significant enhancement 

of the lesion was noted on the postcontrast images. No 

other parenchymal or extraaxial lesions were noted.

CONCLUSION 

The patient underwent a neuronavigation-guided surgical 

resection of the tumor followed by cranioplasty. The resection 

extended to the transverse sinus inferiorly and superior 

sagittal sinus medially. No frank invasion of the venous 

sinuses were noted intraoperatively. The histopathology 

revealed a malignant spindle cell neoplasm consistent with a 

high-grade leiomyosarcoma with myxoid and epitheloid 

areas. Postoperative imaging showed a complete resection of 

the mass with adequate skull reconstruction (Fig. 2). The 

patient received postoperative radiation and has had an otherwise 

uneventful clinical course. 

KEY WORDS: Leiomyosarcoma, skull 


Unusual Central Nervous System Presentation of 

Primary Amyloidosis 

Frank, D. · Bhatia, R. · Gedzdman, E. · Gallo, B. · 

Schumacher, J. 


Miami, FL 

PURPOSE 

Review of a rare and dramatic case of cerebral amyloid 

angiopathy resulting from systemic primary (AL) amyloidosis, 

occurring as the initial manifestation of the disease with 

brief discussion of the pathophysiology as well as other patterns 

of central nervous system (CNS) involvement in different 

forms of amyloidosis. 


A 59-year-old female presented to the emergency department 

with a 2-day history of worsening lethargy, confusion, 

and gait difficulty. She also was noted to have a 2-year history 

of short-term memory loss and progressive decline in 

higher mental functions. Lab tests were significant for elevated 

serum kappa light chains, serum IgG, and CSF IgG elevated 

kappa light chains also were present in the patients 

urine, although electrophoresis of serum, CSF, and urine did 

not demonstrate a monoclonal peak. 

RESULTS 

Initial CT scans of the brain demonstrated diffuse, symmet- 

215 

ric decrease in cerebral attenuation with sulcal effacement; 

findings consistent with diffuse cerebral edema. In addition 

to diffuse, hyperintense T2 and FLAIR signal throughout the 

cerebral and cerebellar white matter, MR imaging demonstrated 

numerous punctate hypointense lesions on gradientecho 

and FLAIR images consistent with hemosiderin deposition. 

There was no radiographic evidence of acute hemorrhage. 

A brain biopsy was obtained to establish definitive 

diagnosis after all other major encephalopathies were 

excluded diagnostically. The specimen revealed positive 

staining for vascular amyloid deposition as well as associated 

perivascular inflammatory changes. 

CONCLUSION 

Amyloidoses are a heterogeneous group of diseases characterized 

pathologically by the parenchymal or perivascular 

deposition of insoluble proteins or protein fragments. They 

are classified by the specific excess or abnormal serum protein 

present, which may be due to neoplastic, inflammatory, 

or genetic causes. Primary (AL) amyloidosis is caused by a 

plasma cell dyscrasia leading to overproduction of kappa or 

lambda light chains. Neurologic involvement in the systemic 

amyloidoses, such as AL, generally takes the form of peripheral 

and autonomic neuropathy. Central nervous system 

involvement, when present, is often primarily vascular with 

hemorrhage as the initial presentation. Our specific case 

illustrates a dramatic nonhemorrhagic presentation of primary 

(AL) amyloidosis with cerebral involvement. The 

overall radiographic features are consistent with those 

described for other amyloidoses that more commonly 

involve the CNS. 

REFERENCES 

1. Caulo M, et al. AJNR Am J Neuroradiol 2001;22:1072-1076 

2. Glusker P, Horoupian DS, Lane B. AJNR Am J Neuroradiol 

1998;19:469-475 

3. Khan MF, Falk RH. Postgrad Med J 2001;77(913):686-693 

4. Rensick AA, et al. Brain Res Rev 2003;43(2):207-223 

KEY WORDS: Cerebral amyloidosis, cerebral amyloid 

angiopathy 

Thursday

Thursday 


Intraparenchymal Central Nervous System Involvement 

of Mycoses Fungoides (Cutaneous T Cell Lymphoma) 

Reyes, M. · Rosel, P. · Foltz, C. · Nedzi, L. 

Tulane University Medical Center 

New Orleans, LA 

PURPOSE 

Mycoses fungoides (MF) is a cutaneous T cell lymphoma that 

rarely metastasizes to the central nervous system (CNS). We 

present two patients with the clinical history of MF that presented 

with mental status changes and upon imaging were 

found to have intraparenchymal CNS disease. A biopsy was 

consistent with cutaneous T cell lymphoma metastases to the 

CNS. 


CT and MR imaging were used to characterize intraparenchymal 

metastatic disease to the brain in two patients 

with MF. 

RESULTS 

Multiple intraaxial masses were identified in the brain of 

both patients with a clinical diagnosis of MF. CT with and 

without iodinated contrast was the initial diagnostic study 

followed by MR imaging with and without gadolinium contrast. 

The size and number of lesions were assessed using 

both modalities. Biopsy was planned based on imaging findings. 

CONCLUSION 

Mycoses fungoides is a cutaneous T cell lymphoma that can 

metastasize to the CNS. Leptomeningeal spread of disease is 

typically seen rather than intraparenchymal spread of disease. 

Only 43 previous cases of CNS MF are reported and 

we present two new patients with intracranial metastatic disease. 

New treatment methods have prolonged survival considerably 

in these two patients. Intraparenchymal rather than 

leptomeningeal metastasis is observed in these patients due 

to the increased time of control of the patients’ local disease. 

KEY WORDS: Central nervous system, mycoses fungoides, 

cutaneous T cell lymphoma 


Central Nervous System Manifestations of Chronic 

Epstein-Barr Virus Infection in X-Linked 

Lymphoproliferative Disease 

Weeks, J. K. · Helton, K. J. · Conley, M. E. · Khan, R. B. 

St. Jude Children’s Research Hospital 

Memphis, TN 

PURPOSE 

The unique central nervous system (CNS) manifestations of 

a patient with X-linked lymphoproliferative disease (XLP) 

will be described. X-linked lymphoproliferative disease is a 

rare disorder characterized by inability to clear Epstein-Barr 

virus (EBV) infection. We are the first to provide MRA evidence 

of diffuse, fusiform aneurysmal dilatation of intracranial 

vessels in XLP. 

216 


A previously healthy 7-year-old male was diagnosed with 

Burkitt’s lymphoma of the intestine in March 1990. Treatment 

with multiagent chemotherapy induced long-term remission. In 

July 2001, he presented with malaise, weight loss, night 

sweats, and fever. Complete blood count contained atypical 

lymphocytes, and EBV DNA was detected in the patient’s 

serum by polymerase chain reaction (> 48,000 copies/mL). 

The patient developed lymphadenopathy, pleural effusions and 

hepatitis. By September 2001, he became disoriented and 

lethargic, with impaired short-term memory. Brain MR imaging 

demonstrated dilatation of multiple intracranial vessels and 

a hyperintense lesion in the right thalamus. Cerebro-spinal 

fluid (CSF) analysis revealed 275/mm 3 white cells (90% lymphocytes), 

1080 mg/dL protein, and 6,000 EBV DNA 

copies/mL. DNA analysis demonstrated a deletion of exon one 

of SH2D1A, the gene responsible for XLP. He improved with 

intravenous immunoglobulins, cyclosporine, acyclovir, and 

high dose steroids. Biopsies of a new frontal lobe lesion and 

pathologic lymph node showed immunoblastic B-cell lymphoma, 

and all tumor cells were EBV DNA positive. He was 

treated with monoclonal antibodies, rituximab, and alemtuzumab. 

The patient received maternal haplo-identical stem 

cell transplant, but developed hepatic dysfunction and venoocclusive 

disease(VOD), confirmed by ultrasound. He succumbed 

to his illness, and an autopsy request was denied. 

RESULTS 

T1- and T2-weighted MR images of the brain demonstrated 

hyperintense foci in bilateral basal ganglia, internal capsules, 

thalami, and medial temporal lobes. MR angiography 

showed fusiform dilatation of the cisternal intracranial vasculature 

bilaterally, including: A1, M1, and M2 segments of 

the anterior and middle cerebral arteries, supraclinoid segments 

of the internal carotid arteries, and distal portion of the 

basilar artery. T1-weighted imaging shows a 4.9 x 4.3 cm 

peripherally enhancing lesion in the right frontal lobe with 

surrounding vasogenic edema. 

CONCLUSION 

This patient’s illness started with an infectious mononucleosis, 

he then developed EBV encephalitis. Terminal phase of 

illness was characterized by neoplastic proliferation of 

infected B-cells, and histology confirmed lymphoma in the 

brain and lymph nodes. Aneurysmal dilatation of intracranial 

vessels seen in this patient is a novel finding, which may 

have resulted from a vasculitic process or direct EBV infection 

of the vessel wall. 

KEY WORDS: X-linked lymphoproliferative disease, 

Epstein-Barr virus, CNS vasculitis


Disseminated Leptomeningeal Juvenile 

Xanthogranulomatosis: Presentation with Raised 

Intracranial Pressure and Absence of Intracranial 

Masses on Imaging 

Gor, D. M. 1 · Shekhar, K. 2 · Pollock, A. N. 2 · Simon, E. M. 2 · 

Feygin, T. 2 · Bilaniuk, L. T. 2 · Zimmerman, R. A. 2 

1Hospital of University of Pennsylvania, Philadelphia, PA, 

2Children’s Hospital of Philadelphia, Philadelphia, PA 

PURPOSE 

Juvenile xanthogranuloma (JXG) is a subgroup of histiocytosis 

characterized by proliferation of cells of monocytemacrophage 

lineage. It is distinguished from the less aggressive 

Langerhan’s cell histiocytosis (LCH) clinically, and histologically 

by absence of Birbeck or Langerhan’s granules in 

the cytoplasm of the proliferating histiocytes (1). JXG is 

most often asymptomatic and discovered during autopsy. 

Fifty percent of cases have a skin lesion. It most often affects 

the choroid plexus or the dura as focal solid masses (2). Few 

reported symptomatic cases have been secondary to obstruction 

to the CSF flow by the solid lesions. 


We present a case of JXG with diffuse cranial and spinal leptomeningeal 

involvement without any solid masses. The 

patient presented with intracranial hypertension. Case: 

Twelve-year-old male with a history of Chiari I malformation 

successfully treated in the past, presented with complaints 

of intermittent occipital headaches for 5 months. 

Ophthalmologic examination revealed hazy optic disks bilaterally, 

remainder of the examination was nonfocal. CT scan 

of the brain was normal. CSF opening pressure was 30 cm of 

water, and symptoms improved after removal of 15 cc of 

CSF. A diagnosis of pseudotumor ceribri was made. The 

patient returned 3 months later with similar complaints. 

Opening pressure was 26.5 cm of water with a closing pressure 

of 13 cm of water after removal of 13 cc of CSF. CSF 

protein was high at 119 mg %, glucose low at 20 mg %, and 

30 Wbcs with a differential of 48 % macrophages and 38 % 

monocytes. Bacterial, viral, and fungal studies were negative. 

MR imaging performed showed diffuse leptomeningeal 

enhancement in the supra and infratentorial compartment, 

enhancement of cranial nerves VII and VIII, diffuse 

enhancement on the surface of the entire spinal cord, and diffuse 

enhancement of most spinal nerve roots. The enhancement 

was non-nodular and no focal masses were identified. 

RESULTS 

Dural biopsy showed proliferation of foamy histiocytes with 

absence of Langerhans granules leading to the diagnosis of 

disseminated juvenile xanthogranulomatosis confirmed by 

immunohistochemistry. 

CONCLUSION 

Juvenile xanthogranulomatosis can present clinically as 

pseudotumor cerebri, and can have diffuse leptomeningeal 

involvement of cranial and spinal compartments without 

focal solid lesions affecting the choroid plexus, dura or brain 

parenchyma. Differential diagnosis includes chronic infectious 

conditions and other granulomatous disorders. 

Diagnosis requires biopsy. 

217 

REFERENCES 

1. Favara BE, Feller AC, Pauli M, et al. Contemporary classification 

of histiocytic disorders. The WHO committee on histiocytic 

/ reticulum cell proliferations. Med Pediatr Oncol 

1997;29:157-166 

2. Lesniak MS, Viglione MP, Weingart J. Multicentric parenchymal 

xanthogranuloma in a child: case report and review of 

literature. Neurosurgery 2002; 51:1493-1498 

KEY WORDS: Xanthogranulomatosis, leptomeningeal, 

pseudotumor 


Air Emboli to the Brain: An Unusual Complication of 

Esophageal Cancer 

Hopkin, J. R. · Mamourian, A. J. · Williams, T. · Koff, M. 

Dartmouth-Hitchcock Medical Center 

Lebanon, NH 

PURPOSE 

Noniatrogenic air embolism is an uncommon cause for 

ischemic events of the brain. We present an unusual case of 

arterial air embolus caused by an esophageal to left atrial fistula 

as late complication of prior esophagogastrectomy and 

radiation therapy. 


A 58-year-old man with a remote history of esophagogastrectomy, 

and local radiation therapy was admitted with neurologic 

symptoms of numbness and weakness. Initial symptoms 

were suggestive of endocarditis. Blood cultures 

became positive for multiple bacteria and fungi, but TEE 

was negative for vegatations. Despite aggresive care, he had 

rapid neurologic decline and died. Autopsy showed multifocal 

infarcts within end organs, pericardial inflammation, and 

an esophageal to left atrial fistula. Several cavitary lesions 

consistent with parynchemal air collections were demonstrated 

in regions of infarct. 

Image 1,2: Large intracerebral air collection, and its atrial 

source. 

RESULTS 

Air embolism may cause infarcts and are similar to other 

sources of emboli causing ischemia in many vascular distributions. 

Early findings on CT resemble those from other 

causes of ischemia and include poor gray white differentiation 

and perhaps some mass effect. Initial MR findings are 

increased T2 signal on FLAIR imaging, and more specifically, 

restricted diffusion on diffusion-weighted imaging (1). 

While case reports exist of rapid gas production in certain 

types of abscess, the development of air in an abscess from 

Thursday

Thursday 

septic emboli occurs in a stepwise fashion after the initial 

embolic event (2). Findings of early ischemia and pockets of 

air suggest a rapid ischemic event, and can be a useful finding 

in differentiating between air embolus and abscess. 

CONCLUSION 

There must be a connection between air and the circulatory 

system and pressure gradient that favors movement of air 

into the circulation for air emboli to occur. The left atrium is 

the only large volume low pressure arterial site distal to the 

pulmonary “filter,” and esophageal pressures can be high 

enough to favor a gradient directed toward the left atrium 

(3). This case serves as a reminder that whenever air and 

infarcts are identified in the brain it should initiate a comprehensive 

search for the source and even improbable mechanisms, 

such as that described here, should be considered. 

REFERENCES 

1. Grossman, Yousem. Neuroradiology 

2. Lilay PC. Rapid gas forming brain abscess from Klebsiella 

pnumoniae. J Neurosurg 1999;91(6):1060 

3. Klein WA. Sphincter-like thoracoabdominal high-pressure 

zone after esophagogastrectomy. 

KEY WORDS: Emboli, air 


Rosai-Dorfman Disease Presenting as a Frontal Lobe 

Mass 

Kruger, A. Y. · Spearman, M. P. · Sanghvi, A. N. · 

Ragoowansi, A. 

Western Pennsylvania Hospital 


PURPOSE 

To describe the unusual presentation of isolated CNS 

involvement of Rosai-Dorfman disease as a frontal lobe 

mass. 


A 73-year-old male presented to the emergency department 

after 3 days of weakness, disorientation, and several 

episodes of falling. A noncontrast CT was performed initially 

on presentation to the emergency department. MR imaging 

of the brain then was performed for preoperative planning. 

RESULTS 

On the CT the mass centered in the inferior frontal lobe was 

of slight increased attenuation with surrounding edema. MR 

imaging of the brain demonstrated an irregular mass of slight 

increased signal on T1-weighted images and decreased signal 

on the T2-weighted images. Extensive surrounding 

edema extended throughout the frontal lobe. The irregular 

mass enhanced intensely with intravenous gadolinium contrast. 

There was no abnormal adjacent dural enhancement. 

218 

CONCLUSION 

Rosai-Dorfman disease was first described in 1969 as a rare 

idiopathic lymphohistiocytic proliferative disease characterized 

by sinus histiocytosis and massive lymphadenopathy. 

The disease most commonly affects children or young adults 

and has a slight male predominance. The most common presenting 

symptoms include massive painless cervical lymphadenopathy 

(95%), fever, and polyclonal hypergammaglobulinemia. 

Isolated extranodal disease is uncommon, but 

can occur in the skin, orbit, paranasal sinuses, and upper respiratory 

tract. Isolated intracranial disease is rare and has 

been described as abnormal dural enhancement or a duralbased 

enhancing mass mimicking a meningioma. We report 

a case of isoloted CNS involvement of Rosai-Dorfman disease 

presenting as a solitary frontal lobe mass without dural 

involvement. One of the histopathologic hallmarks of Rosai- 

Dorfman disease is the finding of emperiolesis (lymphocytophagocytosis), 

which describes the presence of lymphocytes 

within the histiocytes. On immunohistochemical 

examination, the histiocytes are S-100 protein positive 

which distinguishes it from granulomatous diseases such as 

sarcoid and Wegener’s granulomatosis. The histiocytes are 

CD1a negative which distinguishes it from Langerhans histiocytosis. 

REFERENCES 

1. Konishi E, Ibayashi N, Yamamoto S, Scheithauer BW. Isolated 

intracranial Rosai-Dorfman disease (sinus histiocytosis with 

massive lymphadenopathy). AJNR Am J Neuroradiol 

2003;24:515-518 

2. Wang E, Anzai Y, Paulino A, Wong J. Rosai-Dorfman disease 

presenting with isolated bilateral orbital masses: Report of 

two cases. AJNR Am J Neuroradiol 2001;22:1386-1388 

KEY WORDS: Rosai, Dorfman, histiocytosis 

Paper 400A Starting at 11:15 AM, Ending at 11:20 AM 

Superiority of Dynamic CT Myelography in Localizing 

an Occult Cerebrospinal Fluid Leak 

Eckel, L. J. · Campeau, N. G. 

Mayo Clinic 

Rochester, MN 

PURPOSE 

To illustrate the superiority of dynamic CT myelography in 

localizing an occult cerebrospinal fluid leak. 


A 45-year-old female with a long standing history of low

pressure headaches had undergone an extensive prior 

workup including two conventional CT myelograms. 

Extradural contrast was identified on these studies; however, 

the precise location of the leak could not be determined. 

Unsuccessful treatments including multiple blood patches 

were attempted. Open surgery including C1-2 laminectomy 

with dural repair was attempted also, but unsuccessful with 

persistence of headaches and presumed CSF leak. At this 

point, the patient underwent additional imaging with a 

dynamic CT myelogram. 

RESULTS 

Dynamic CT myelographic images of the entire spine were 

obtained with patient in the right and left lateral decubitus, 

prone, and supine positions. CT parameters for the study 

included: slice thickness 5 mm, FOV 25 cm, 140 kVP, 260 

mA. The entire spine was imaged in 36 seconds for each 

position. Initial imaging began coincidentally with intrathecal 

contrast administration (mixture of 15 cc Omnipaque 180 

and 10 cc nonbacteriostatic saline). Extradural contrast 

already was noted in the epidural space at the level of the 

right pedicle of T4 on the second imaging pass within 1 

minute of contrast administration. Later images demonstrated 

continued accumulation of contrast material within the 

epidural space, centered about the T4 level, confirming this 

as the definitive location of the leak. The patient then 

received definitive operative repair of the leak, and the 

patient’s symptoms promptly resolved. 

CONCLUSION 

Dynamic CT myelography is a highly useful tool in the 

detection of CSF leak in those patients where prior detection 

or treatment of leak has been unsuccessful. It provides both 

high spatial and temporal resolution required for identification 

of CSF leaks refractory to detection using conventional 

myelographic techniques. 

KEY WORDS: CSF leak, myelogram, dynamic 

Paper 400B Starting at 11:20 AM, Ending at 11:25 AM 

Suprasellar Hemangioblastoma Producing 

Hydrocephalus and Optic Tract Edema in a Patient with 

Von Hippel-Lindau Disease 

Kale, H. A. · Sklar, E. · Sternau, L. 

University of Miami/Jackson Memorial Hospital 

Miami, FL 

PURPOSE 

Pituitary infundibular hemangioblastomas are extremely 

rare. We present a case of histopathologically proven hemangioblastoma 

involving the pituitary stalk producing obstructive 

hydrocephalus and optic tract edema. 


A 31-year-old woman with past history of bilateral retinal 

hemangioblastomas was diagnosed with Von Hippel-Lindau 

disease (VHL) disease and treated with radiation. She also 

had a cerebellar hemangioblastoma that was resected. She 

now has headaches and visual loss. 

RESULTS 

Contrast-enhanced MR imaging of the brain and orbits per- 

219 

formed showed a lesion of the pituitary stalk. The lesion had 

high signal on the T2-weighted imaging and an isointense 

signal on the T1-weighted imaging with homogenous 

enhancement (Fig 1). An interesting finding demonstrated 

best on the FLAIR imaging was bilateral optic tract edema 

(Fig. 2). A pituitary stalk biopsy performed showed a hemangioblastoma. 

Fig 1: Enhanced sagittal T1-weighted imaging shows a 

markedly thickened homogenously enhancing pituitary 

stalk. 

Fig 2: FLAIR axial image demonstrating optic tract edema. 

CONCLUSION 

VHL is a multisystem neoplastic syndrome associated with 

mutation of the suppressor gene on Chromosome 3. Affected 

Thursday

Thursday 

individuals may develop CNS hemangioblastomas in cerebellum, 

spinal cord, and medulla in order of frequency. 

These tumors are often cystic with a solid mural component. 

Unusual sites of hemangioblastomas in VHL are anterior 

lobe of the pituitary, optic nerve, corpus callosum, wall of 

the third ventricle, temporal horn of the lateral ventricles, 

frontal/temporal lobe, and meninges. The best imaging technique 

available for hemangioblastomas is contrast-enhanced 

MR imaging. Angiography commonly is performed prior to 

surgery to demonstrate the feeding vessels. Other VHL manifestations 

are pheochromocytoma, cysts, microcystic adenomas, 

and neuroendocrine tumors of the pancreas, kidneys, 

and liver, epididymal cystadenoma and endolymphatic sac 

tumors. Suprasellar lesion differentials are granulomatous 

diseases like tuberculosis, sarcoidosis, histiocytosis and neoplastic 

processes like astrocytoma, lymphoma, germinoma, 

and craniopharyngioma. Differential of a hemangioblastoma 

should be considered in case of a pituitary stalk lesion in a 

known VHL patient. 

KEY WORDS: Von Hippel-Lindau disease, hemangioblastoma, 

pituitary stalk 


10:15 AM - 11:45 AM 

Room 103 

(65) ELC Workshop E: Advanced 

Website Production 


12:00 PM - 12:55 PM 

Room 201D 

⎯ Dale A. Charletta 

(66) American Society of Spine 

Radiology (ASSR) Annual Business 

Meeting (Members Only) 

220 


12:00 PM - 12:55 PM 

Room 201F 

(67) American Society of Pediatric 

Neuroradiology (ASPNR) Annual 


Thursday Afternoon 

1:00 PM - 2:30 PM 

Room 105/106 

(68) Review of Pediatric Brain 

Imaging (ASPNR) 

(402) The Neonate and Young Infant 

⎯ P. Ellen Grant, MD 

(403) The Older Child and Adolescent 

⎯ Susan I. Blaser, MD 

Moderator: Patrick D. Barnes, MD 

The Neonate and Young Infant 

P. Ellen Grant, MD 



able to: 

1) Identify patterns of brain injury due to compromised 

blood flow and oxygenation, and due to ischemic stroke in 

newborns and infants. 

2) Be aware of hints that help distinguish inborn errors of 

metabolism from other more common causes of brain injury. 

3) Identify common infections that present in infancy. 

4) Be familiar with congenital malformations that commonly 

present in infancy. 


In the newborn and young infant the brain is undergoing 

rapid changes and its imaging appearance is unlike the older 

child and adult. As a result the appearance of the normal 

brain changes with time. In addition the types of abnormalities 

that can be seen in the immature brain are often uncom-

mon in older children and adults. These factors make neonatal 

and young infant neuroradiology particularly challenging. 

In this session we will use a case-based approach and discussion 

of the differential diagnosis and optimal imaging 

strategy to cover a number of abnormalities that commonly 

are seen in the neonatal and young infant age group. These 

cases will illustrate how to identify neonatal and young 

infant brain injury using a pattern-based approach. Hints that 

help differentiate inborn errors of metabolism from other 

more common causes of brain injury will be discussed. 

Infections that commonly present in infancy will be presented. 

Finally malformations that are often diagnosed in this 

age group will be reviewed. 

The Older Child and Adolescent 



1:00 PM - 2:30 PM 

Theatre 

(69) Advanced Imaging in the Spine 

(ASSR) 

(404) Parallel Imaging in the Spine: Theory and 

Reality 

⎯ Christiane K. Kuhl, MD 

(405) Functional MR Imaging in the Cord 

⎯ Spyros Kollias, MD 

(406) Imaging of Cell-Based Therapies for Spinal 

Cord Injury 

⎯ Edward D. Wirth, III, MD, PhD 

Moderator: Christiane K. Kuhl, MD 

Parallel Imaging in the Spine: Theory and Reality 

Christiane K. Kuhl, MD 



able to: 

1) Understand the technical groundwork of parallel imaging 

with parallel imaging like "Sensitivity Encoding" 

("SENSE"). 

2) Understand how parallel imaging can be used in clinical 

221 

neuroradiology. 

3) Understand the specific difficulties of parallel imaging in 

spine applications. 

4) Learn the SENSE specific advantages for high-field MR 

imaging. 


For optimum detection and classification of disease states, 

virtually all clinical MR imaging applications in the field of 

neuroscience require both short acquisition times and a high 

spatial resolution. Up to now, increasing gradient strength 

has been the only strategy to meet the increasing demands of 

advanced diagnostic imaging applications. Yet, this strategy 

is limited by physical, economical, and medical considerations: 

Increasing gradient strength is technically difficult, it 

is associated with significant hardware costs, and goes along 

with the risk to induce unwanted side effects such as peripheral 

neuro-stimulation. With the advent of parallel imaging 

techniques like SENSE, this dilemma can be overcome. In 

parallel imaging, phase encoding steps are replaced by 

exploiting the spatial information that is inherent to the spatially 

variable sensitivity of an array of surface coils. Several 

different techniques for parallel imaging have been proposed 

by different MR system vendors (ASSET, IPAT, SENSE); 

they currently allow a reduction of phase encoding steps by 

a reduction factor of 2 to 6 (and recently even higher). The 

SENSE-mediated reduction of acquisition time can be used 

for an improved temporal or spatial resolution of any given 

pulse sequence, without change of image contrast. The prerequisite 

for using SENSE is the availability of multielement 

coils that are placed exactly in parallel to each other, which 

is difficult for spine applications and explains why the use of 

parallel imaging for spine applications is still lagging behind 

that of virtually all other applications in neuroimaging. New 

solutions are, however, to use the existing multi -element 

coils and use the profiles of two adjacent elements for generating 

the SENSE image. In addition to the mere increase of 

image acquisition speed, the reduction of phase-encoding 

steps brings about two further advantages that are particularly, 

but not only, important for high-field MR imaging: First, 

in single-shot EPI applications that usually are used for diffusion 

imaging, diffusion tensor imaging or for functional 

BOLD-contrast MR studies, SENSE helps to shorten the 

echo-train length in proportion to the reduction factor. The 

considerably shorter echo train reduces the accumulation of 

phase errors during the EPI readout and, accordingly, 

reduces susceptibility effects like image distortions and blurring. 

In addition, the shorter echo train translates into a significantly 

higher SNR compared to sequentially phase 

encoding. Second, parallel imaging helps reduce RF deposition 

(regular phase encoding requires an RF pulse for every 

step); this proves extremely helpful for high-field imaging, 

where, due to the higher SAR (specific absorption rate), 

most pulse sequences need to be “slowed down” to avoid 

excessive heating of the patient. Only with parallel imaging 

like SENSE, the actual high-field SNR advantage can be 

fully exploited. Using SENSE is associated with a 30% SNR 

penalty; at 1.5 T, its use is therefore confined to applications 

or pulse sequences with high inherent SNR, like MRA studies. 

For high-field MR imaging, the SNR loss due to SENSE 

is not important; in fact, higher reduction factors become 

clinically feasible, thus contributing to an even higher acquisition 

speed. For both contrast-enhanced and inflow MR 

angiography, SENSE helps improve spatial resolution and 

anatomical coverage at a given acquisition time. 

Thursday

Thursday 

Functional MR Imaging in the Cord 

Spyros Kollias, MD 



able to: 

1) Familiarize with the methodological issues related to 

obtaining functional MR imaging (fMRI) data from the 

human spinal cord. 

2) Critically discuss the results obtained so far from the 

application of spinal fMRI in the investigation of normal 

spinal cord function and its alteration in patients with spinal 

myelopathy. 

3) Evaluate future advances of spinal cord fMRI, its clinical 

feasibility and the applicability of the method for assessing 

residual and regained functions of the injured spinal cord. 


The gray matter of the spinal cord contains a high density of 

specialized neuron cell bodies and has a high capillary density. 

Motor tasks and sensory stimuli activate specific neuronal 

populations in the spinal cord and elicit local hemodynamic 

effects that can be detected with functional MR imaging 

methodologies. Challenges associated with the application 

of fMRI methodologies in the spinal cord arise from its 

fine structure and elasticity, the requirement for high inplane 

resolution and the presence of CSF, respiratory and 

swallowing motion artifacts. Blood oxygen level dependent 

(BOLD) and signal enhancement by extravascular water protons 

(SEEP) methods used for spinal fMRI and imaging protocols 

will be addressed briefly (1). Studies in healthy subjects 

using several types of stimulation including thermal, 

electrical, and motor tasks (2) and animal studies (3) confirming 

the presence of neuronal activation in sites in which 

functional MR imaging detected activity will be reviewed. 

Preliminary results from the application of the technique to 

assess function in spinal cord injury patients will be discussed 

(4). Determining the reproducibility of spinal fMRI is 

essential for its clinical applications and the results of a 

recent study assessing the reliability of the method for 

detecting and localizing neuronal function in the spinal cord 

will be presented. Distinct patterns of activity in the cervical 

and lumbar spinal cord during vibratory stimulation of different 

dermatomes will be demonstrated. Emphasis will be 

on the potential of the method for assessing residual and 

regained functions of the injured spinal cord with regard to 

the integration of reflex actions and locomotor pattern generating 

circuits and provide objective information additional 

to the one obtained by standard clinical neurologic scores. 

REFERENCES 

1. Stroman PW, Krause V, Malisza KL, Frankenstein UN, Tomanek 

B. Extravascular proton-density changes as a non-BOLD 

component of contrast in fMRI of the human spinal cord. 

Magn Reson Med 2002b;48:122-127 

2. Stroman PW, Ryner LN. Functional MRI of motor and sensory 

activation in the human spinal cord. Magn Reson Imag 

2001;19:27-32 

3. Lawrence J, Stroman PW, Bascaramurty S, Jordan LM, Malisza 

KL. Correlation of functional activation in the rat spinal 

cord with neuronal activation detected by immunohistochemistry. 

NeuroImage 2004;22:1802-1807 

222 

4. Stroman PW, Tomanek B, Krause V, Frankenstein UN, Malisza 

KL. Mapping of neuronal function in the healthy and injured 

human spinal cord with Spinal fMRI. NeuroImage 

2002;17:1854-1860 

Imaging of Cell-Based Therapies for Spinal Cord Injury 

Edward D. Wirth, III, MD, PhD 

Dr. Edward Wirth completed the MD/PhD program at the 

University of Florida in 1994. He elected to remain at the 

University of Florida to conduct postdoctoral research, and 

subsequently joined the faculty there in 1996. From 1997 to 

2002, Dr. Wirth led the University of Florida team that performed 

the first human embryonic spinal cord transplant in 

the United States. This pilot study demonstrated the feasibility 

and safety of implanting embryonic spinal cord cells into 

patients with posttraumatic syringomyelia. From 2002 to 

2004, Dr. Wirth held academic appointments at Rush- 

Presbyterian St. Luke’s Medical Center and at the University 

of Chicago. In 2004, he joined Geron Corporation as 

Associate Medical Director. 



able to: 

1) Interpret MR images of cellular transplants in the injured 

human spinal cord. 

2) Describe emerging MR techniques for assessing the survival 

and migration of implanted cells. 


Numerous studies in animal models have demonstrated that 

cell-based therapies can facilitate partial return of neurologic 

function following spinal cord injury (SCI). The overall 

goal of this approach is to provide and/or stimulate a permanent 

repair of the injured spinal cord tissue via a single 

implantation of cells into the injury site. Several clinical trials 

of various cell types are either in progress or planned to 

begin in the near future. A key challenge for these trials is the 

need to correlate changes in a patient’s functional outcome 

with survival, migration, and anatomical impact of the 

implanted cells. This presentation will address the feasibility 

and technical challenges of imaging these transplants in 

the human spinal cord. Emphasis will be placed on the ability 

of current and emerging MR imaging techniques to delineate 

implanted cells from host tissue and to assess their 

effects on host white matter. Since the volume of implanted 

cells is typically very small, strategies to achieve the necessary 

spatial resolution through improved signal-to-noise also 


REFERENCES 

1. Wirth ED, III, Reier PJ, Fessler RG, et al. Feasibility and safety 

of neural tissue transplantation in patients with 

syringomyelia. J Neurotrauma 2001;18:911-929 

2. Hoehn M, Kustermann E, Blunk J, et al. Monitoring of 

implanted stem cell migration in vivo: a highly resolved in 

vivo magnetic resonance imaging investigation of experimental 

stroke in rat. Proc Natl Acad Sci USA 2002;99:16267- 

16272 

3. Lakatos A, Franklin RJ. Transplant mediated repair of the 

central nervous system: an imminent solution? Curr Opin 

Neurol 2002;15:701-705

4. Bulte JW, Duncan ID, Frank JA. In vivo magnetic resonance 

tracking of magnetically labeled cells after transplantation. J 

Cereb Blood Flow Metab 2002;22:899-907 


1:00 PM - 2:30 PM 

Room 107 

(70) Emerging Functional Modalities 

(ASFNR) 

(407) Advances in Molecular Imaging 

⎯ A. Gregory Sorensen, MD 

(408) Clinical Magnetic Source Imaging 

— Timothy P.L. Roberts, PhD 

(409) Functional MR Imaging Brain Tumors: 

Advantages and Challenges 

— Jay J. Pillai, MD 

Moderator: A. Gregory Sorensen, MD 

Advances in Molecular Imaging 


Clinical Magnetic Source Imaging 

Timothy P.L. Roberts, MD 



able to: 

1) Understand the neural basis of magnetoencephalography. 

2) Understand the fusion of MEG and MR imaging into 

magnetic source images (MSI). 

3) Appreciate the value of MSI for functional mapping of 

eloquent cortex. 

4) Appreciate the value of the high temporal resolution 

inherent to MEG. 


Magnetoencephalography (MEG) is an emerging methodology 

for detection of the extracranial magnetic fields associated 

with the electrical current of neuronal activity. 

Mathematical modeling of the source of such activity, combined 

with anatomical registration to MR imaging allows the 

formation of magnetic source images (MSI or functional 

maps). Since MEG has submillisecond temporal resolution, 

223 

these maps have the potential to display the spatiotemporal 

dynamics of neuronal activity and signal propagation, as 

well as revealing temporal signatures in evoked responses. 

To date, the principle clinical success of magnetic source 

imaging has been the identification of somatosensory cortex 

in patients scheduled for neurosurgical resection of mass 

lesions, such as tumors, primarily in fronto-parietal locations. 

Preservation of functional tissues of the motor and 

somatosensory regions is a prime consideration in planning 

the treatment strategy and surgical approach. Prior knowledge 

of the functional organization, with respect to the 

lesion, is a vital aid to the neurosurgeon, radiologist, and 

patient in evaluating the risks versus likely benefits of invasive 

surgery and focused irradiation (gamma knife). A number 

of studies have documented the ability of MSI (with 

somatosensory stimulation) to define the central sulcus even 

in situations where anatomical methods fail. The essential 

validation of source localizations by comparison to intraoperative 

cortical stimulation mapping and/or SSEP recording 

has given further support to the increasingly routine clinical 

adoption of this modality in patients in whom there is suspicion 

of sensorimotor cortex involvement. A number of centers 

routinely incorporate MSI data into the neuronavigational 

systems used to guide operative procedures, such that 

MSI is used not only for presurgical planning, but also intraoperative 

guidance. This presentation will discuss the basis 

of magnetoencephalography and application to functional 

mapping of eloquent cortex prior to surgery in patients with 

brain tumors. Differences to fMTI will be discussed, as well 

as application to recording of spontaneous activity (e.g., 

epileptogenic discharges). Finally, the use of high temporal 

resolution to reveal temporal processing will be discussed, 

along with the outlook for new clinical indications. 

Functional MR Imaging Brain Tumors: Advantages and 

Challenges 




able to: 

1) Describe the advantages of use of BOLD fMRI in brain 

tumor presurgical planning. 

2) Recognize the limitations and challenges of use of BOLD 

fMRI in such presurgical planning. 

3) Identify ways of overcoming the various limitations of 

fMRI in such clinical scenarios. 

4) Design clinically useful fMRI paradigms for sensorimotor 

and language mapping. 


The lecture will begin with a brief introduction to the underlying 

principle of BOLD functional MR imaging (fMRI), 

followed by a brief review of various studies which have 

compared the presurgical lateralizing and localizing ability 

of fMRI to that of the established gold standards (i.e., intraoperative 

cortical stimulation and Wada testing) for technique 

validation. Advantages of BOLD fMRI over these 

other techniques will be discussed, as they relate to presurgical 

mapping, as well as to the study of CNS plasticity. In 

addition, various fMRI paradigms will be discussed for clinical 

presurgical motor and language mapping, and examples 

Thursday

Thursday 

of each will be shown. A thorough discussion of the various 

limitations and challenges of this technique thenwill ensue, 

with specific mention of problems such as motion artifacts, 

susceptibility artifacts, statistical thresholding issues, 

patient-specific task performance limitations, and alterations 

in the intensity of the BOLD effect reflecting tumor-related 

neurovascular uncoupling (secondary to impaired vascular 

autoregulation/vasoactivity or local hemodynamic alterations 

related to possible factors suxh as tumor angiogenesis, 

tumor infiltration, or extrinsic vascular compression) or 

adjacent large draining vein effects. Approaches for dealing 

with many of these challenges will be discussed, and some 

examples will be shown of how these various problems can 

be minimized or actually resolved in routine clinical practice. 

Thus, despite potential limitations, BOLD fMRI 

remains as a powerful noninvasive diagnostic method for the 

delineation of eloquent cortex in the vicinity of brain tumors. 


1:00 PM - 2:00 PM 

Room 103 


(NLM) Hands-on Workshop: When 

PubMed ® Is not the Answer 


3:00 PM - 4:30 PM 

Room 105/106 

(72a) ADULT BRAIN: Innovative 

Techniques and Miscellaneous 



(72b) SPINE: Interventional and Innovative 

Techniques 

(72c) PEDIATRICS: Vascular, Trauma and Tumors 

(72d) PEDIATRICS: Vascular and Miscellaneous 

Moderators: Walter Kucharczyk, MD 


— Maryanne Blake 

224 


MR Angiography at 3.0 T versus Digital Subtraction 

Angiography in the Follow Up of Intracranial 

Aneurysms Treated with Detachable Coils 

Majoie, C. B. 1 · Sprengers, M. E. 1 · van Rooij, W. J. 1 · Lavini, 

C. 1 · Sluzewski, M. 2 · van Rijn, J. C. 1 · den Heeten, G. J. 1 

1 Academic Medical Center, Amsterdam, THE 

NETHERLANDS, 2 St. Elisabeth Ziekenhuis, Tilburg, 

THE NETHERLANDS 

PURPOSE 

To assess the feasibility and usefulness of 3 dimensional 

time-of-flight (3D TOF) MRA performed at 3.0 T compared 

to digital subtraction angiography (DSA) for the follow up of 

coiled intracranial aneurysms. 


In a prospective study, 20 consecutive patients with 21 

coiled intracranial aneurysms underwent DSA, unenhanced 

and enhanced multiple overlapping thin slab acquisition 

(MOTSA) 3D TOF MRA at 3.0 T on the same day at a mean 

follow up of 6 months (range 4-14 months) after coil placement. 

MR angiography was evaluated for presence of artifacts, 

presence and size of aneurysm remnants, and recurrences, 

patency of parent and branch vessels, and added 

value of contrast enhancement. MR angiographic findings 

were compared to DSA. 

RESULTS 

Interobserver agreement of MRA was good as was agreement 

between MRA and DSA. All three recurrences that 

needed additional treatment were detected on MRA. Minor 

disagreement occurred in 4 cases: 3 coiled aneurysms were 

scored on MRA as having a small remnant, while on DSA 

these aneurysms were occluded and the other aneurysm was 

scored on MRA as having a small remnant, while on DSA 

this was a small recurrence. The use of contrast material had 

no additional value. Coil-related MR imaging artifacts were 

minimal and did not interfere with evaluation of the occlusion 

status of the aneurysm. 

CONCLUSION 

High-spatial resolution 3D TOF MRA at 3.0 T is feasible and 

useful in the follow up of patients with intracranial 

aneurysms treated with coil placement 

KEY WORDS: MR angiography, aneurysm, coil 


Hyperacute, Acute, and Subacute Intracranial 

Hemorrhage: Detection and Quantitation Using CT and 

Multimodal MR Imaging 

Omura, M. C. · Villablanca, J. 

University of California Los Angeles, Center for the Health 

Sciences 


PURPOSE 

CT is considered the gold standard for detecting and diagnosing 

hyperacute intracerebral hemorrhage (ICH). Recent

studies suggest that MR imaging also may be useful in the 

hyperacute stage. The purpose of this study was to assess the 

accuracy of MR imaging in the detection and quantitation of 

hyperacute, acute, and subacute ICH. 


Thirty-two patients, male/female: 18/14, age 27 to 95 years, 

diagnosed with either primary ICH (n = 21) or hemorrhagic 

transformation of an infarct (n = 11) between 1998 and 2003, 

who had MR imaging performed within 24 hours of CT 

(mean time between CT and MR imaging = 4 hours 53 minutes), 

were studied retrospectively. CT was used as the reference 

standard. Lesion conspicuity on MR imaging was 

rated “greater than CT,” “equal to CT,” “less than CT,” or 

“not detectable.” Lesion volume for CT and MR imaging 

was calculated by multiplying lesion length, width, and 

height. Fisher’s exact test and Student’s t-test were used to 

evaluate categorical and continuous data, respectively. A pvalue 

of 0.05 was considered statistically significant. 

RESULTS 

Of the 32 subjects, 16 underwent MR imaging within 7 

hours of symptom onset (hyperacute), 10 within 24 hours 

(acute), and 6 within 7 days (subacute). MR sequences performed 

included T2*GRE (n = 32), T2-weighted (n = 30), 

FLAIR (n = 23), T1-weighted (n = 13), DWI (n = 26), and 

ADC (n = 8). All 32 lesions were detected on GRE. GRE 

was more sensitive than ADC (p < 0.0001), T1 (p = 0.0007), 

T2 (p = 0.0016), and FLAIR (p = 0.0129). Lesion conspicuity 

on MR imaging did not vary significantly between hyperacute 

and acute/subacute lesions. There were 5 cases in 

which lesions were more conspicuous on MR imaging compared 

to CT, 2 of which were petechial in nature. These 5 

lesions were significantly smaller than the rest, with a mean 

area of 1.080 cm^2 vs 8.850 cm^2 (p = 0.0383). CT and 

GRE-determined lesion volumes were not significantly different 

in hyperacute period, but GRE overestimated the size 

of acute and subacute lesions, with mean volume on GRE = 

12.91 cm^3 vs CT = 10.49 cm^3 (p = 0.0309). In all time 

periods, T2/FLAIR, DWI and T1 underestimated lesion volume 

compared to CT (p = 0.0306, p = 0.0238, p = 0.0543, 

respectively). 

CONCLUSION 

MR imaging is as effective as CT at detecting ICH, regardless 

of lesion age. In addition, MR imaging may be superior 

to CT in the detection of small parenchymal hemorrhages, 

including petechial bleeding. GRE is more sensitive than 

other MR sequences. GRE lesion volumes correlated most 

closely with CT volumes in the hyperacute period, but GRE 

sequences significantly overestimate lesion volumes in the 

acute and subacute periods, while all other sequences underestimated 

lesion volume at all time periods vs CT. The 

results of this study suggest that MR imaging is a valuable 

tool for the detection of hyperacute, acute and subacute ICH, 

including petechial blood, but is less accurate for the quantitation 

of hemorrhage volume. 

KEY WORDS: MR imaging vs CT, intracranial hemorrhage, 

quantitation 

225 


Evaluation of the Cervicocerebral Vasculature by High 

Resolution 64-Slice Spiral CT Angiography 

Klingebiel, R. · Kentenich, M. · Breitwieser, C. · Mühler, M. 

· Bohner, G. 

Charité 


PURPOSE 

Preliminary evaluation of recently introduced 64-multislice 

spiral CT in performing cervicocerebral CT angiography (cc 

64-MS CTA), focussing on detail resolution, radiation exposure, 

and low-dose protocols. 


Thirty patients (13 male, 17 female; age range: 32 to 90 

years), transferred by the departments of neurology, vascular 

surgery, and radiation therapy for noninvasive assessment of 

the cervicocerebral vasculature underwent 64-MS CTA (32 

mm detector width, 0.5/0.4 mm collimation/increment, 120 

kv, 150 mAs, pitch 0.75). In a subset of patients (n = 5), suspected 

of suffering from major vessel occlusive disease, a 

low-dose protocol (50 mAs) was used. Image quality and 

detail resolution of x2 and x3 segments were assessed by two 

radiologists (5-point scale). Effective dose (ED) values were 

calculated and compared to 4-slice CTA. 

RESULTS 

Image quality and detail resolution of standard 64-MS CTA 

proved superior (p < 0.05) compared to the low-dose protocol. 

Yet, major vessel assessment was possible by low-dose 

64-MS CTA. Effective dose values amounted to 2.2/0.7 mSv 

(150/50 mAs). Corresponding values for 4-slice CTA were 

4.3/1.4 mSv. 

CONCLUSION 

Sixty-four-slice spiral CT angiography provides high-detail 

resolution and simultanouesly decreases radiation exposure 

as compared to 4-slice CTA, probably due to decreased overbeaming. 

Low-dose protocols may be beneficial for major 

vessel assessment such as sinus vein thrombosis, especially 

in younger patients. 

KEY WORDS: Multislice, CT angiography, cerebrovascular 


New 3D Angiography System with Flat Panel Detector of 

Direct Conversion Type for Evaluation of Intracranial 

Vessels 

Kakeda, S. 1 · Ohnari, N. 1 · Hatakeyama, Y. 1 · Moriya, J. 1 · 

Oda, N. 1 · Korogi, Y. 1 · Nishino, K. 2 · Miyamoto, W. 2 

1 University of Occupational and Environmental Health 

School of Medicine, Kitakyushu-shi, JAPAN, 2 Medical 

Systems Division, Shimadzu Corporation, Sohraku-gun, 

JAPAN 

PURPOSE 

Although 3D angiography system using conventional image 

intensifier (II) has been reported helpful for determining 3D 

vessel structures, there are some problems related to the 

Thursday

Thursday 

characteristics of II. Recently, we have developed a new 3D 

angiography system using the flat panel detector (FPD) of 

direct conversion type. Compared with II TV system, our 

system has several theoretical advantages such as higher spatial 

resolution, wide dynamic range, square FOV, and no 

image distortion. The purpose of this study is to compare the 

image quality of 2D DSA between the FPD system and II TV 

system, and to assess 3D DSA with the FPD system in the 

depiction of intracranial vessels. 


Sixteen consecutive patients (9 men, 7 women; age range, 

18-68 years; mean age, 53 years) who underwent intracranial 

angiography were examined prospectively. All patients 

underwent 2D DSA with both II TV system and FPD system 

in at least one projection. Three-dimensional DSA images 

were created from the rotational DSA data with the FPD system. 

Two blinded radiologists independently evaluated 2D 

DSA with FPD system and II TV system using a 5-point 

scale (excellent ~ not visible) in the depiction of intracranial 

vessels. MIP and VR images of 3D DSA with FPD system 

were evaluated also using the same scale. The following vessel 

segments were assessed: A2-3 segment of ACA, perforating 

vessels, peripheral small vessels which defined as cortical 

segment of ACA and MCA in less than 2 cm from the 

brain surface. 

RESULTS 

For 2D DSA, the FPD system was significantly superior to 

II TV system regarding the visibility of peripheral and perforating 

vessels (p < .05), although differences were not significant 

for the main arterial segments such as A2-3 segment 

between both systems. The peripheral and perforating vessels 

were visualized sufficiently on MIP images of 3D DSA 

in all 16 cases. 

CONCLUSION 

Our FPD system was superior to the II TV system in the vis- 

226 

ibility of small intracranial vessels, and could create the high 

quality 3D DSA images on which high spatial resolution 

allows precise visualization of small vessels such as perforating 

vessels. 

KEY WORDS: 3D angiography, flat panel detector, intracranial 

vessel 


Multidetector Helical CT Angiography versus Digital 

Subtraction Angiography in the Evaluation of Neck 

Remnants following Neurosurgical Clipping of 

Intracranial Aneurysms Using Titanium Clips 

Dhaliwal, S. G. · Gifford, W. · Morales, T. · Frazee, J. · 

Duckweiler, G. · Jahan, R. · Vinuela, F. · Sayre, J. · 

Villablanca, J. P. 

University of California Los Angeles Medical Center 


PURPOSE 

Patients who have undergone neurosurgical aneurysm clipping 

frequently require digital subtraction angiography 

(DSA) to exclude the presence of a residual aneurysm neck. 

CT angiography (CTA), historically, has not been a viable 

alternative to DSA postclipping due to beam-hardening artifacts 

arising from metallic clips. With the advent of titanium 

clips, CT artifacts have been reduced greatly, opening the 

possibility of noninvasive evaluation of these patients. In 

this study, we consider whether CTA is an acceptable alternative 

to DSA in patients with intracranial aneurysms treated 

with titanium clips. 


Patients who have undergone: 1) neurosurgical clipping of 

an intracranial aneurysm with titanium clips, 2) a postoperative 

DSA and 3) a postoperative CTA were included in the 

study. CT angiography was performed using a timing injection, 

pitch of 1.5, 1 mm slice collimation, 0.5 mm reconstruction 

interval, FOV 18 cm, matrix 512 X 512, soft tissue 

kernel, using Omnipaque 350 injected intravenously at 3 

cc/sec. Image analysis was performed using 2D gray-scale 

MPR and 3D volume rendered images. DS angiography 

studies used 1K bi-plane angiography. After reporting of 

CTA and DSA studies, the results of the two procedures were 

compared taking the DSA report results as the gold standard. 

All studies were interpreted prospectively in a blinded manner. 

RESULTS 

Nineteen patients with a total of 22 clipped aneurysms, who 

had undergone postoperative DSA and CTA, were enrolled. 

According to DSA, 5/22 (23%) aneurysms showed a residual 

neck, 1/22 (5%) was indeterminate, and 16/22 (72%) 

aneurysms showed no residual neck. The CTA results 

matched the DSA results in 15/22 (68%) cases. In 12/15 

(80%) matches CTA matched DSA for no residual neck. In 

2/15 (13%) matches CTA and DSA both detected residual 

necks, while in 1/15 (7%) CTA matched DSA for indeterminate. 

In 4/22 (18%) cases, the CTA was indeterminate. In this 

group, DSA showed no residual aneurysm in 3/4, and was 

indeterminate in 1/4. In 3/22 cases (14%), the CTA reported 

a false negative for residual neck. There was 1 false positive

CTA (5%). The overall sensitivity of CTA for residual neck 

or concordant indeterminate finding was 50%, with a specificity 

of 92%. 

CONCLUSION 

CT angiography matched DSA findings in over 2/3 of the 

cases studied. CT angiography accurately detected the presence 

of a residual neck in 40% of the patients with residual 

found on DSA. CT angiography accurately detected the 

absence of residual in 75% of patients in whom DSA found 

no residual. CT angiography had an intermediate sensitivity 

and a high specificity for the detection of residual neck or 

concordant indeterminate findings as compared to DSA. In 

the cases for which the CTA results were indeterminate, we 

found that beam-hardening artifact which obscured the area 

of interest increased with the number of clips used. In these 

patients, a supplemental DSA should be performed in order 

to determine the presence of a residual aneurysm neck. A 

larger patient population is being accrued to more accurately 

characterize the role of CTA in this patient population. 

KEY WORDS: Aneurysm, CT angiography, clip 


Cerebral Arteriovenous Malformation: Hemodynamic 

Characterization by Ultrafast Contrast-Enhanced MR 

Angiography and Nonenhanced Dynamic Active Tagging 

MR Angiography 

Essig, M. 1 · Bock, M. 1 · Debus, J. 2 

1 German Cancer Research Center, Heidelberg, GERMANY, 

2 University of Heidelberg, Heidelberg, GERMANY 

PURPOSE 

To assess the hemodynamic characteristics of patients with 

angiographically proven cerebral arteriovenous malformations 

(AVMs), we applied both an ultrafast, time-resolved 

contrast-enhanced MR DSA technique (cMRA) as well as a 

nonenhanced dynamic tagging MR DSA (dMRA). 


In an ongoing prospective protocol, so far, we have examined 

20 patients with the cMRA and more than 150 patients 

with dMRA. The cMRA is based on an ultrafast 3D FLASH 

acquisition after bolus injection of 10 cc of MultiHance® 

(Bracco Diagnostics, Princeton, NJ) using parallel imaging 

techniques (IPAT) at a TR of < 2 ms and a TE of < 1 ms. The 

cMRA is performed as a multiphasic acquisition with a temporal 

resolution of about 250 ms. The dMRA sequence is 

based on the spin labeling sequence (STAR) in which a bolus 

of blood upstream from the AVM is tagged by an inversion 

pulse and imaged downstream at different time intervals 

after tagging using also a 2D FLASH acquisition. This 

results in multiple angiograms between 100 ms and 1200 ms 

after tagging. For both techniques morphologic parameters 

(e.g., size, location, number and origin of vessel components), 

and hemodynamic parameters were assessed and 

compared with the clinical presentation. For hemodynamic 

information the time between the arterial feeders and the 

draining veins was estimated as the AVM shunt time. 

RESULTS 

A hemodynamic assessment as well as a detailed analysis of 

227 

the angioarchitecture is possible with both techniques (Fig 

1). Feeding arteries, AVM nidus, and draining veins were 

detected easily on both techniques. The time resolution with 

the dMRA is better; however the examination time is 10 

times longer. Artifacts are more pronounced on the dMRA 

while the cMRA proved to be a very robust technique. 

Smaller AVMs generally showed shorter shunt times; however, 

a short shunt-time was associated with a higher risk of 

bleeding, independent of the AVM size. The hemodynamic 

parameters changed significantly after therapeutic interventions 

(e.g., radiosurgical treatment). 

CONCLUSION 

Both cMRA using only 10 cc of MultiHance® and dMRA 

proved to be very helpful diagnostic tools in the assessment 

of the angioarchitecture and the hemodynamic changes in 

patients with cerebral AVMs. The time-resolved acquisition 

allowed interesting insights into the hemodynamics of such 

malformations and correlated well with the clinical presentation. 

A shorter shunt-time was associated with a higher risk 

of intracerebral bleeding independent of the malformation 

size. The cMRA is more robust and significantly faster, 

while the dMRA still provides the better time resolution. 

KEY WORDS: Arteriovenous malformations, MR digital subtraction 

angiography, contrast media 


Imaging Capillary Density Using Susceptibility- 

Weighted Imaging 

Haacke, E. M. 1 · Khan, A. 2 · Ayaz, M. 2 · Kirsch, W. 3 

1 The MRI Institute for Biomedical Research, Detroit, MI, 

2 Wayne State University, Detroit, MI, 3 Loma Linda 

University, Loma Linda, CA 

PURPOSE 

To image capillary density in the brainstem using MR imaging. 

In this paper we use susceptibility-weighted imaging to 

correlate filtered phase images with anatomical images of 

the brain using India ink staining. 

Thursday

Thursday 


Susceptibility-weighted images are collected at 1.5 T using a 

fully flow compensated, 3D, gradient-echo sequence. The 

normal acquisition parameters are used: in-plane resolution 

0.5 mm x 1.0 mm; TH = 2 mm, FOV = 256 mm x 256 mm; 

Nx = 512; Ny = 256; Nz = 48; TE = 40 ms; TR = 57 ms; FA 

= 20 degrees. The phase images are high-pass filtered. The 

area around the substantia nigra and red nucleus was evaluated 

over 4 to 5 slices. We then compared these results to the 

recent brainstem cadaver analysis by Duvernoy (2). In this 

book, he presents results from an India ink stained brain 

using 2 mm to 3 mm thick slices. 

RESULTS 

We found that there is a one-to-one match of the filtered hase 

susceptibility-weighted images with the India ink stained 

brain slices. There are usually four 2 mm slices that are needed 

to visualize the entire red nucleus. The subtantia nigra, 

both pars compacta and pars reticulata, are seen clearly 

along with the interdigitation at their boundary. The crus 

cerebri is seen, as is the medial geniculate. The depth of signal 

change seen on the MR images also matches well with 

that seen in the India ink stained results from Duvernoy (2). 

CONCLUSION 

There is an excellent match between the structures here and 

the India ink stained images. It is interesting that the changes 

in the phase also appear to correlate with what has been 

thought to be an increase in iron content. This raises the 

question of iron origin: is it heme iron, or is there a nonheme 

iron contribution? To address this question, it will be necessary 

to perform experiments to distinguish between heme 

and nonheme iron. In summary, there is a strong correlation 

between the contrast in the susceptibility-weighted phase 

images and the anatomical structures revealed in India ink 

stained cadaver brains. 

REFERENCES 

1. Reichenbach JR, et al. Small vessels in the human brain. 

Radiology 1997;204:272-277 

2. Duvernoy HM. Human brainstem vessels. Springer-Verlag, 

Berlin, Germany 1999;206-209 

KEY WORDS: Susceptibility-weighted imaging, capillary 

density, neuroanatomy 

228 


Idiopathic Intracranial Hypertension: The Validity of 

Cross-Sectional Neuroimaging Signs 

Agid, R. · Farb, R. I. · Willinsky, R. A. · Mikulis, D. J. · 

Tomlinson, G. 

Toronto Western Hospital 


PURPOSE 

To evaluate the utility of previously reported neuroimaging 

signs in establishing or excluding the diagnosis of idiopathic 

intracranial hypertension (IIH). 


Thirty patients with confirmed IIH and 56 controls were 

evaluated prospectively with brain MR imaging. All exams 

included standard T1 and T2 as well as enhanced 3D fast GE 

high-resolution volume acquisition. All examinations were 

evaluated in a blinded fashion by three neuroradiologists for 

the presence or absence of the “traditional” signs of IIH: 

empty sella turcica, deformation of the pituitary, slit-like 

ventricles, “tight” subarachnoid spaces, flattening of the posterior 

globe, protrusion of the optic nerve, distension of the 

optic nerve sheath, and vertical tortuosity of the optic nerve. 

RESULTS 

Optic nerve protrusion and enhancement, slit-like ventricles 

and tight CSF spaces were not significantly associated with 

IIH (p > 0.05). Posterior globe flattening, optic nerve sheath 

distension, optic nerve tortuosity, pituitary deformity and 

empty sella turcica were associated significantly with IIH (p 

< 0.05). However, these are not helpful in a clinical setting 

with the exception of posterior globe flattening. This is the 

only sign that, if present, strongly suggests the diagnosis of 

IIH (specificity 99.1, sensitivity 43.5, positive likelihood 

ratio 49.7). 

CONCLUSION 

The majority of the reported signs for IIH on planar imaging 

are not helpful in establishing or excluding the diagnosis of 

IIH and are of no value in the clinical setting. Flattening of 

the posterior aspect of the globe is the only sign that if present 

is suggestive of the diagnosis of IIH. 

KEY WORDS: Pseudotumor, MR imaging, cross-sectional 

imaging 


MR Imaging, MR Angiography, and MR Perfusion in 

Patients with Posterior Reversible Encephalopathy 

Syndrome 

Bartynski, W. S. · Boardman, J. F. 

Presbyterian University Hospital 


PURPOSE 

The purpose of this study is to correlate MR imaging, MR 

angiography (MRA), and MR perfusion (MRP) obtained 

concordantly in patients with posterior reversible 

encephalopathy syndrome (PRES).


Eleven patients with brain imaging findings consistent with 

PRES were identified who had routine MR imaging, MRA, 

and MRP during the same imaging session available for correlation. 

Three patients had eclampsia, 2 had immune suppressive 

therapy, 1 had lupus and 5 had multiorgan failure. 

The routine imaging studies, MRA and MRP findings were 

compared and correlated. 

RESULTS 

Brain imaging studies demonstrated the typical findings of 

PRES consistent with vasogenic edema including lesions in 

the occipital poles, parietal region, frontal lobes, inferior 

temporal-occipital junction, and cerebellar hemispheres. 

Vascular irregularity consistent with spasm was identified in 

10/11 patients (4 severe, 6 mild-moderate). In these 10 

patients MRP demonstrated patchy regions of decreased and 

delayed perfusion most commonly correlating with the areas 

of vasogenic edema. Regions of reduced and delayed MR 

perfusion also were found in areas of otherwise normalappearing 

brain, typically in the watershed regions. In a single 

patient with eclampsia, the MRA appeared normal with 

no suggestion of spasm. MR perfusion in this patient demonstrated 

regions of reduced perfusion consistent with the areas 

of vasogenic edema, but delayed perfusion was not present. 

CONCLUSION 

In regions of vasogenic edema identified by MR imaging in 

patients with the PRES syndrome, MRP most consistently 

demonstrates areas of hypoperfusion and delayed perfusion. 

MR angiography coincidentally obtained demonstrates vessel 

irregularity confirming or suggesting vasospasm in these 

patients. These findings suggest that the lesions identified in 

PRES syndrome are related most likely to local or global 

brain hypoperfusion. 

KEY WORDS: PRES, MR imaging, MR perfusion and 

angiography 


Intracerebral Hemorrhage after Rapid Decompression 

of Chronic Subdural Hematoma: An Analysis of 

Contributing Factors, Interpolation of Existing Theories, 

and Hypothesis of Pathophysiologic Mechanism 

Mirsky, D. M. 1 · Bello, J. A. 1,2 

1Albert Einstein College of Medicine, Bronx, NY, 

2Montefiore Medical Center, Bronx, NY 

PURPOSE 

Intracerebral hemorrhage represents the most clinically devastating 

postoperative occurrence following surgical 

drainage of chronic subdural hematoma (CSH). One third of 

patients with this complication die, and another third are 

severely disabled. Historically considered to be rare, the true 

incidence of an intracerebral hematoma post-CSH evacuation 

ranges from 0.7% to 4%. The etiology of this potentially 

fatal complication has been the subject of debate in the literature. 

Entering the debate over the underlying mechanism 

for this complication, we first reviewed existing theories. By 

integrating well defined pressure relationships between the 

CSF and venous systems, we then synthesized a plausible 

mechanism focusing on the vulnerable venous circulation. 

229 


An extensive literature review of underlying principles and 

postulated mechanisms for parenchymal hemorrhage following 

treatment of CSH was performed. In particular, we 

examined the following theories: a) perfusion breakthrough 

bleeding, b) the effects of disrupted cerebral vascular 

autoregulation, c) vascular damage secondary to mechanical 

shift of cranial contents, and d) the correlation between 

intracranial and venous pressures. These theories underlie 

several well established clinical conditions, including reperfusion 

bleeding following treatment of arteriovenous malformations, 

hemorrhage in the setting of venous thrombosis, 

and various types of hemorrhage related to shunting hydrocephalus. 

We postulate an etiology for parenchymal cerebral 

hemorrhage following CSH drainage by integrating the 

above theories into the pathophysiology represented in case 

material. 

RESULTS 

We observed parenchymal hemorrhage develop immediately 

postdrainage of large bilateral CSH. These subcortical hemorrhages 

occurred bilaterally, remote from the drainage sites. 

This took place within the clinical setting of a patient previously 

shunted for communicating hydrocephalus. Cerebral 

spinal fluid diversion in this patient potentially reduced his 

subcortical venous pressures. When he became symptomatic 

from bilateral large CSH, his increased intracranial pressure 

necessarily would have reduced the blood flow in these subcortical 

veins. With an increase in parenchymal blood flow 

following CSH drainage, we postulate that these veins are 

vulnerable to rupture by a variety of mechanisms. With 

altered pressure dynamics at play, the resultant venous vulnerability 

reflects disrupted autoregulation, and leads to a 

form of perfusion breakthrough bleeding. The integration of 

these principles forms the basis for our hypothesis. 

CONCLUSION 

Our hypothesis for the underlying pathophysiologic mechanism 

for intracerebral hemorrhage following decompression 

of CSH is based on contributing factors including ICP, 

altered venous pressure dynamics, and underlying venous 

CSF pressure equilibrium. It is the unique pressure environment 

established by CSH that makes venous vulnerability 

the plausible mechanism. 

KEY WORDS: Chronic subdural hematoma, intracerebral 

hemorrhage 

Thursday

Thursday 


3:00 PM - 4:30 PM 

Theatre 

(72b) SPINE: Interventional and 

Innovative Techniques 



(72a) ADULT BRAIN: Innovative Techniques and 

Miscellaneous 



Moderators: Brian C. Bowen, MD, PhD 

Adam E. Flanders, MD 


MR Neurography in Extraspinal Sciatica 

Lewis, A. M. · Layzer, R. · Engstrom, J. W. · Barbaro, N. M. 

· Chin, C. T. 



PURPOSE 

Patients with leg pain resembling lumbosacral radicular sciatica 

sometimes have a normal lumbar MR imaging. The 

term “piriformis syndrome” has been used in such cases, but 

evidence of sciatic nerve entrapment by the piriformis muscle 

is controversial. MR neurography (MRN) has emerged 

recently as a powerful tool for evaluating peripheral nerve 

pathology. To evaluate the sciatic nerve on MRN in patients 

presenting with extraspinal sciatica. 


Patients with unexplained sciatic distribution pain without evidence 

of nerve root compression on routine lumbar MR imaging 

were referred by neurologists and neurosurgeons for MRN 

(August 2003-present). Coronal and axial inversion recovery 

sequences were obtained of the lumbosacral plexus and sciatic 

nerves (Philips 1.5 T magnet TR 2530 TE 20 TI 160 3/3.5 mm 

FOV 34 NEX 3). Clinical history, examination, electrodiagnostic 

tests, and operative findings were obtained. 

RESULTS 

Nine patients (7 female; 2 male; ages 35-67 years) underwent 

MRN for extraspinal sciatica. In eight patients, MRN 

demonstrated abnormal increased signal in the ipsilateral 

sciatic nerve at the level of the sciatic notch and piriformis 

muscle. In one of these patients, abnormal signal was 

demonstrated in both sciatic nerves. Four of these patients 

also had an abnormal ipsilateral piriformis muscle (two with 

denervation atrophy; two with relative hypertrophy). One 

patient had a normal MRN. Typical clinical symptoms con- 

230 

sisted of buttock pain radiating down the posterior thigh into 

the leg, aggravated by sitting, bending at the waist, or exercise, 

which sometimes could be reproduced by pressing on 

the sciatic notch or by straight leg raising. Neurologic 

deficits were mild or absent. Three of four patients who 

underwent electrodiagnostic testing had findings consistent 

with mild sciatic neuropathy; the fourth patient had a normal 

EMG. To date, two patients (including the patient with a normal 

MRN) have undergone surgical exploration and decompression 

of the sciatic nerve. Both patients had intraoperative 

evidence of entrapment by a fibrous portion of the piriformis 

muscle. Postoperatively, both patients reported complete resolution 

of their symptoms. Five patients (2 female; 3 male; 

ages 54-80 years) without sciatica who underwent MRN did 

not demonstrate abnormal signal within the sciatic nerve. 

CONCLUSION 

In patients presenting with unexplained extraspinal sciatica, 

MRN often demonstrates focal increased signal within the 

sciatic nerve at the level of the sciatic notch and piriformis 

muscle. It remains to be determined whether nerve entrapment 

by the piriformis muscle is responsible for these MRN 

findings. 

KEY WORDS: MR neurography, piriformis syndrome, sciatica 


Time-Resolved Spinal MR Angiography with TRICKS 

and Parallel Imaging: A New Technique and Early 

Clinical Experience with Spinal Dural Fistulae 

Walker, M. T. · Cashen, T. · Khawar, S. · Shaibani, A. · Carr, 

J. · Hopkins, J. · Shin, W. · Batjer, H. · Carroll, T. 

Northwestern Memorial Hospital 

Chicago, IL 

PURPOSE 

The gold standard examination to identify and characterize a 

spinal dural arteriovenous fistula (DAVF) is digital subtraction 

angiography (DSA). The spatial and temporal resolution 

of DSA are unrivaled but there are downsides including 

time, expense, morbidity and mortality related to this inva-

sive procedure. High-resolution contrast-enhanced spinal 

MR angiography (CE MRA) can anatomically depict these 

lesions but cannot display the temporal information that is 

critical to their recognition. With advances in MR technology, 

time-resolved MRA has become possible and may, with 

further work, approach the diagnostic power of DSA. The 

purpose of this study is to describe a new technique for timeresolved 

MRA which includes echo-sharing, a short TR and 

parallel imaging and to report our early clinical experience. 


Patients with suspected/known spinal dural fistula were 

imaged using a 1.5 T whole body MR scanner (Avanto, 

Siemens Medical Solutions, Erlangen, Germany) with an 

abdominal array. A 3D multiphase time-resolved echoshared 

angiographic technique (TREAT) pulse sequence was 

combined with parallel imaging (GRAPPA). The 3D TREAT 

sequence is a segmented k-space acquisition, which uses the 

TRICKS variable rate k-space acquisition. Coronal, coronal 

oblique, and sagittal acquisitions were employed depending 

upon the level imaged. Gadolinium-based contrast material 

was administered as a single dose in an antecubital vein at an 

injection rate of 4.0 ml/s. 

RESULTS 

Screening of the entire neuroaxis in one patient with suspected 

DAVF localized the fistula to the pelvis with bilateral 

lateral sacral arterial supply. This information was used to 

focus the spinal DSA and subsequent intervention. After 

treatment, time-resolved MRA showed no residual fistula, 

which correlated with the posttreatment DSA. A second 

patient underwent embolization of a lumber DAVF followed 

immediately by a time-resolved MRA for baseline follow up. 

During that exam, a second dAVF was identified remote 

from the primary DAVF. The second fistula was subsequently 

treated. Three additional patients underwent spinal MRA 

for unexplained SAH all of which were negative. 

Coronal oblique early arterial phase showing midline early 

draining vein (temporal resolution = 6 secs). 

CONCLUSION 

Time-resolved MRA with parallel imaging is viable in the 

231 

spine and can identify pathologic flow states such as DAVFs. 

Potential applications include diagnosis and localization of 

fistulous sites and the effectiveness of endovascular or surgical 

therapy. If proven sensitive and accurate, this tool will 

decrease the morbidity and mortality related to conventional 

spinal DSA and reduce cost, which can be significant. 

KEY WORDS: Time-resolved MR angiography, parallel 

imaging, spinal dural fistula 


CT-Guided High Cervical Root Block for Cervicogenic 

Headache 

Marin, H. · Jain, R. · Pace, M. · Patel, S. C. · Mitsias, P. D. 

Henry Ford Hospital 

Detroit, MI 

PURPOSE 

Cervicogenic headache (CHA) is a common type of 

headache generated by abnormalities or dysfunction of cervical 

structures, and is often severe and disabling. Response 

to selective root or nerve blocks can establish the diagnosis 

and offer temporary relief. We performed this study in order 

to assess the value of CT-guided C2 and C3 root blocks for 

confirmation of diagnosis, pain relief, and preoperative evaluation 

of patients with presumed cervicogenic headache. 

The regional anatomy, indications, technique and potential 

complications will be illustrated with representative cases. 


We performed retrospective record review of all patients 

referred for C2 or C3 nerve blocks for the period 12/1999- 

10/2004. All patients had extensive neurologic and imaging 

evaluations, and failed to respond satisfactorily to medical 

treatment. C2 and C3 root blocks were performed uni or 

bilaterally, using a mixture of 1 mL of triamcinolone 40 

mg/mL and 1 mL of Bupivacaine 0.25%. Patients were contacted 

for assessment of pain relief during the first week and 

at 1 month after the block. 

RESULTS 

Twenty-six patients had 34 nerve blocks. The etiology of CHA 

was: posttraumatic (n = 8), degenerative spine disease (n = 4), 

migraine with cervicogenic trigger (n = 9) and idiopathic (n = 

7). Bilateral blocks were performed simultaneously in 7 

patients. One patient had only local anesthetic and 4 patients 

had steroid injection alone. Pain relief in the first week after the 

block was reported by 18 out of 26 patients (69%). Sustained 

relief at 1 month was reported by only 10 patients (38%). Of 

the latter, three benefited from either placement of an occipital 

stimulator (n = 2) or C2 ganglionectomy (n = 1). In the 8 

patients who did not respond to the blocks, the headache was 

diagnosed as not being of cervicogenic origin. 

CONCLUSION 

CT-guided percutaneous upper cervical root block is a simple 

and safe procedure. It primarily has diagnostic value for 

the confirmation of the cervicogenic origin of chronic 

headache. It also has therapeutic value for pain relief in carefully 

clinically selected patients. 

KEY WORDS: Cervical nerve block, cervicogenic headache 

Thursday

Thursday 


CT Fluoroscopically Guided Neurolytic Sympathetic 

Plexus Block: An Efficacious Pain Management 

Approach in Patients with Unresectable Cancer 

Shah, L. M. · Keyserling, H. · Pekala, J. · Huh, B. · Gray, L. 


Durham, NC 

PURPOSE 

Neurolytic sympathetic plexus block (NSPB) is an established, 

effective procedure for the palliation of visceral pain 

in patients with unresectable cancer. Conventional techniques 

using surface landmarks, fluoroscopy, and ultrasound 

are wrought with difficulty and complications. CT-guided 

NSPB provides more specific localization of the needle tip in 

relation to the celiac artery and avoids penetration or injection 

of the spinal cord, major vascular structures, or adjacent 

viscera. CT fluoroscopy offers the added benefit of more 

rapid and accurate needle guidance, as the operator is able to 

remain at the bedside and manipulate the needle with maximal 

patient comfort and less potential for movement 

between imaging. Accurate placement also may allow more 

effective distribution of the neurolytic agent. We present four 

patients who underwent CT fluoroscopically guided NSPB 

with excellent results. 


Four consecutive patients underwent NSPB utilizing CT fluoroscopic 

guidance over an 8-month period. The procedures 

were performed by a single neuroradiologist in conjunction 

with the anesthesiology service on a single-detector spiral 

CT scanner modified for CT fluoroscopy. Neurolytic sympathetic 

plexus block was performed with the patients in prone 

or prone-oblique position. After selected axial imaging 

through the upper abdomen for procedure planning, two 22 

gauge spinal needles were advanced via bilateral posterior 

oblique approaches into the retrocural space at the level of 

the sympathetic plexus. Once optimal needle placement was 

achieved, 10 cc of 100% ethanol and 10 cc of 0.5% bupivacaine 

were infused through each needle. 

RESULTS 

Successful needle placement was achieved in each case, and 

there were no immediate or delayed complications. Three of 

the four patients experienced total to near complete pain 

relief. Continued oral analgesia was required in the fourth 

patient, despite moderate improvement in pain symptoms. 

232 

Fig. 1. Static image obtained uncer CT fluoroscopic-guided 

neurolytic sympathetic plexus block in patient #1. 

CONCLUSION 

CT fluoroscopic NSPB has the advantages of CT guidance 

for accurate needle placement while avoiding vital structures 

and the speed of fluoroscopic guidance with maximal patient 

tolerance and comfort. 

KEY WORDS: Celiac plexus block, CT fluoroscopic guidance 


Spread of Contrast during Selective CT Fluoroscopic 

Epidural Injections 

Wagner, A. L. 1,2 

1Rockingham Memorial Hospital, Harrisonburg, VA, 

2University of Virginia School of Medicine, Charlottesville, 

VA 

PURPOSE 

To document the spread of contrast during low-volume 

epidural steroid injections (ESI) using CT fluoroscopic 

(CTF) guidance in order to access for the feasibility of selective 

ESI in the treatment of single-level radiculopathy. 


Twenty-two patients scheduled for CTF-guided ESI were 

injected with 1cc of Omnipaque 180 after standard slightly 

paracentral posterior needle placement and images were 

obtained of the proximal ipsilateral lateral recesses at 3 adjacent 

levels. The density of contrast adjacent to the nerve was 

classified as 0 (none) to 3 (high) and the presence or absence 

of contralateral spread was documented. 

RESULTS 

At least moderate contrast was present in the inferior lateral 

recess in 21 of 22 patients (95%) and in the superior lateral 

recess in 20 of 22 (91%). Contralateral spread was present 

only in three patients (14%), one of which the contrast was 

primarily contralateral. There were no intrathecal injections.

CONCLUSION 

Small volume epidural injections preferentially stay on the 

side of the injection, even with a paracentral approach. One 

cc is adequate to ensure spread into the adjacent lateral 

recesses in most cases, indicating that a selective ESI, using 

a more lateral approach, could be effective in treating a 

unilevel radiculopathy when the etiology is in the lateral 

recess or proximal neural foramen. 

KEY WORDS: CT fluoroscopy, epidural injections, spine 

intervention 


Vertebroplasty in the Treatment of Painful Spine: A 

Cooperative Study on 1084 Patients 

Muto, M. 1 · Anselmetti, G .2 · Bonaldi, G. 3 · Manfrè, L. 4 · 

Baruzzi, F. 5 · Vallone, S. 6 · Carpegiani, P. 7 

1 2 Cardarelli Hospital, Naples, ITALY, Institute Research 

Scientifics Center Candiolo, Turin, ITALY, 3Civile Hospital, 

Bergamo, ITALY, 4Cannizzaro Hospital, Catania, ITALY, 

5 6 Civile Hospital, Varese, ITALY, Policlinico, Modena, 

ITALY, 7S. Chiara Hospital, Pisa, ITALY 

PURPOSE 

The utility of vertebroplasty in treatment of painful spine has 

been accepted worldwide. In this cooperative study we want 

to show the success and failure rate related to the treated 

pathology (porotic, metastatic, or angioma) the incidence of 

new vertebral fractures adjacent to the previously treated 

one, and the incidence of symptomatic and asymptomatic 

side effects. 


One thousand and eighty-four patients were treated from April 

2001 through September 2004 for a total of 2342 vertebral 

bodies. Eight hundred and six patients (73%) were treated for 

porotic abnormality, 247 patients (24%) were neoplastic and 

31 patients (3%) presented compressive or painful angioma. 

Different type of cement have been used. All patients were 

treated after MR and CT examination or, in alternative to MR 

imaging, bone nuclear medicine scan, always associated to 

233 

clinical evaluation. The patients were treated under CT and C 

arm fluoroscopy control (2 centers) or under fluoroscopy in 

angio suite (5 centers). We never performed general anesthesia, 

but only local sedation or neurolepto. Blood tests were 

always performed before the treatment to exclude coagulopathy. 

The only real contraindications was the consideration of 

the presence of local or systemic infection. We evaluated the 

incidence of new fractures adjacent to the vertebral body previously 

treated within 3 months. 

RESULTS 

We had a success rate in porotic patients of 93%, in neoplastic 

patients of 73% and in patients with angioma of 100%. We had 

asymptomatic venous leak in 135 patients (15%) and CT 

proved asymptomatic pulmonary embolus in 8 patients (0.9%). 

No symptomatic venous and pulmonary embolus were identify. 

We had new fractures adjacent to a previously treated vertebra 

body in 14 patients (1.3%) while 6 patients showed 

radiculopathy (0.6%) due to posterior leakage in the lateral 

lumbar recess, treated and resolved with medical therapy in 3 

weeks. In 6 patients we also had psoas hematoma treated with 

medical therapy; no cord compression was identify. We had 

mild disk leakage in 90 patients (10%) and we found 45 cases 

of new adjacent vertebral fractures within 3 months after the 

previous treatment. There was an increased of the height of the 

vertebral body in 33% of the patients with porotic fractures 

(272 patients) with a mean increase height of 2 mm. 

CONCLUSION 

Vertebroplasty represent a safe and effective technique in the 

treatment of painful spine related to porotic, neoplastic 

changes or in case of angioma. The low incidence of major 

and minor side effects is related to high quality technology 

used by all centers. The incidence of symptomatic effects 

and of new fractures related to previously treated vertebrae 

or disk leakage are very low. 

KEY WORDS: Vertebroplasty, pain, spine 


Vertebroplasty and Kyphoplasty in the Treatment of 

Acute Posttraumatic Fractures 

Muto, M. 1 · Piovan, E. 2 · Guarnieri, G. 3 · Cirillo, L. 3 · 

Zeccolini, F. 1 · De Falco, R. 4 · Beltramello, A. 2 

1 2 Cardarelli Hospital, Naples, ITALY, Borgotrento Hospital, 

Verona, ITALY, 3Policlinico, Naples, ITALY, 4S. Maria delle 

Grazie Hospital, Naples, ITALY 

PURPOSE 

To show the difference between vertebroplasty and kyphoplasty 

in the treatment of acute vertebral trauma and to 

define which one of those two techniques can be consider as 

a correct way to treat this type of patients. 


We treated 35 patients affected by acute posttraumatic thoracic 

(7 patients) and lumbar (28 patients) vertebral fractures from 

January 2003 through February 2004. The patients presented 

a Magerl 1 type fracture in 25 cases and a Magerl type 3 fracture 

in 10 cases. Patients ranged from 25 up 78 years old with 

a mean age of 55 years (22 females and 13 males) and an 

exclusion criteria was a trauma older than 4 weeks. In 5 cases 

Thursday

Thursday 

we performed vertebroplasty while in the remaining 30 a 

kyphoplasty was obtained. In 23 cases the treatment was performed 

within a week after the trauma while in other cases it 

was permormed within 4 weeks. The trauma was related to 

car, work, or sports accident; the clinical follow up was performed 

up to 6 months after the treatment while right after the 

percutaneous therapy the distribution of the cement, cement 

leakage, height restoration, and clinical outcome were evaluated. 

A low density cement and a bipeducolar approach was 

used also in all cases with kyphoplasty or vertebroplasty to try 

to obtain the best filling of the vertebral body. 

RESULTS 

An incomplete and very irregular filling of the vertebral 

body through the lines of the fractures was obtained in the 5 

patients in which vertebroplasty was choosen as percutaneous 

technique. A better homogeneous and symmetrical 

filling was obtained in the 30 patients treated by kyphoplasty. 

There was no significant abnormal cement leakage and no 

new fractures were discovered within 6 months after the 

treatment; a height restoration was obtained in 13 cases 

(mean restoration of 1.5 mm) in the cases treated by kyphoplasty. 

At postprocedure clinical evaluation there was a pain 

reduction in 28 patients (25 patients treated by kyphoplasty 

and 3 patients treated by vertebroplasty) while at the 6month 

follow up all patients had completely recovery. 

CONCLUSION 

Kyphoplasty in this experience represents a safe and effective 

technique to obtain a correct cement distribution in 

patients with acute Magerl Type 1 and type 3 vertebral fracture. 

Vertebroplasty does not assure the same cement filling 

and in the same patients category and this could represent a 

bad result for the biomechanics of the spine, especially in 

younger patients. 

KEY WORDS: Vertebroplasty, kyphoplasty, vertebral fracture 


Vertebroplasty in the Inpatient Population 

Trout, A. T. 1 · Gray, L. A. 2 · Kallmes, D. F. 2 

1 Medical School, Mayo Clinic College of Medicine, 

Rochester, MN, 2 Mayo Clinic, Rochester, MN 

PURPOSE 

Percutaneous vertebroplasty is considered standard care in 

many geographic regions for treatment of vertebral compression 

fractures. Vertebroplasty is offered routinely to 

patients hospitalized for refractory pain, often without regard 

to the chronicity of the fracture, since it is assumed that the 

procedure will facilitate resolution of pain and a rapid hospital 

discharge. However, there exists no data regarding the 

ability of vertebroplasty to speed hospital discharge. We 

report our experience with inpatient vertebroplasty, with specific 

attention to rapidity of discharge and correlation to relevant 

clinical parameters. 


We performed a retrospective review of patients admitted 

between April 2000 and September 2004 with primary diagnoses 

of back pain or vertebral compression fracture who 

were treated subsequently with vertebroplasty. We cata- 

234 

logued total hospital stay and duration of hospitalization 

prior to and after vertebroplasty. We collected outcomes data 

in the form of verbal pain scales (0-10) for pain at rest and 

pain with activity, in-hospital medication use (graded 0-6), 

and posthospitalization medication use. Outcomes were 

assessed at baseline, 1 week, 1 month, 6 months, 1 year and 

2 years postvertebroplasty. Spearman’s Rho was used for 

correlations. Improvement in outcomes over time was analyzed 

with paired t-test comparison to the preceding timepoint 

or the Wilcoxon signed rank test. The Wilcoxon rank 

test and linear regression were used for analysis of the relationship 

between time periods, binary variables (sex, single/multiple 

levels treated, thoracic/lumbar levels) and age. 

RESULTS 

We identified 66 patients (mean age 77.6 years) with a median 

total hospital stay of 6.0 days (range 1-26 days). Median 

length of stay prior to and following vertebroplasty were 4.0 

(range 1-24 days) and 1.5 days (range 0-7 days), respectively. 

Ten of 66 patients were discharged the day of vertebroplasty. 

By days two and three, 33 (50%) and 48 (72.7%) of 66 

patients had been discharged. None of the durations cataloged 

were associated significantly with age, sex, number of levels 

treated, or thoracic vs lumbar treatment. Days between admission 

and vertebroplasty were correlated significantly with the 

change in medication strength between admission and discharge 

(p = 0.045) (i.e., patients who received vertebroplasty 

sooner had greater decreases in the strength of their medication 

by discharge). Days between vertebroplasty and discharge 

were correlated significantly with the medication 

strength on admission (p < .01)(i.e., patients on stronger medications 

at the time of admission required longer hospital stays 

postvertebroplasty). There was significant improvement in 

rest and activity pain by 1 week (p < .0001), with continuing 

improvement at 1 month (p = .02, p = .17, respectively) and 6 

months (p = .047, p = .032, respectively). Medication use significantly 

decreased between all follow-up time-points (p < 

.001 for all time-points). None of the catalogued time-periods 

in hospital were correlated significantly with outcomes. 

CONCLUSION 

This study confirms that vertebroplasty facilitates a rapid 

discharge in the majority of patients admitted to the hospital 

with refractory pain. Vertebroplasty administered earlier in 

hospitalization also leads to greater decreases in the strength 

of analgesics required. After discharge, patients demonstrate 

ongoing improvement over time with regard to both pain and 

medication requirements. 

KEY WORDS: Vertebroplasty, inpatient, outcomes 


Fatal Fat Embolism following Vertebroplasty: 

Identification of the High Risk Patient 

Syed, M. · Jan, S. · Patel, N. 

Mercy Medical Center 

Springfield, OH 

PURPOSE 

A case report of compression fractures at 3 levels treated 

using percutaneous polymethylmethacrylate vertebroplasty. 

To present an autopsy proven case showing a fatal compli-

cation of vertebroplasty although low volumes of cement 

were injected and no radiographic signs of cement leakage 

were recognized. Furthermore, it is important to evaluate 

patients and categorize them as “high risk” before attempting 

vertebroplasty. 


A retrospective review of 542 patients who had undergone 

vertebroplasty. Out of this sample size only one patient presented 

with a complication of death. In this particular 

patient, a transpedicular approach was used to inject codman 

cranioplastic cement mixed with barium sulfate and vancomycin. 

Patient was discharged 2 hours postprocedure. 

RESULTS 

Upon discharge, patient was noticed to be lethargic. On 

arrival to patients home, patient was found to be unresponsive. 

Emergency medical service then was called and patient 

was admitted to emergency room. After arrival to emergency 

room, patient pronounced dead. Autopsy was performed 

showing fat and bone marrow embolization in pulmonary 

arteriole. 

CONCLUSION 

Vertebroplasty has been proven to be safe with complication 

rates as low as 1%. Complications that can occur include 

pulmonary cement emboli, extravasation causing spinal cord 

or nerve root compression, infection, and hemorrhage. Most 

patients have extravasation but of no clinical significance. 

Although relatively safe, it is important to recognize the 

high-risk patient before attempting vertebroplasty. These 

patients include: 1) COPD (especially on home O2), 2) pulmonary 

arterial hypertension/cor pulmonale, and 3) any 

patient with a history of deep venous thrombosis or pulmonary 

embolism. 

KEY WORDS: Vertebroplasty, complication, fat embolism 


Kyphoplasty for the Treatment of Painful Osteoporotic 

Vertebral Compression Fractures: Assessment of the 

First 375 Cases 

Myers, M. E. · Madison, M. T. · Goddard, J. · Myers, T. V. 


St. Paul, MN 

PURPOSE 

Kyphoplasty traditionally has been performed by surgeons 

under general anesthesia in the operating room. We wish to 

report our experience as neuroradiologists performing 

kyphoplasty using conscious sedation in radiology suites. 


We retrospectively reviewed medical records and radiographic 

images from 375 consecutive kyphoplasty procedures 

performed between 2001 and 2004. Sedation records, 

fluoroscopy and procedure time, procedural complications 

and clinical outcomes were examined. 

RESULTS 

The mean age of patients was 75.5 years with mean follow 

up of 15 months. There were no clinical complications. 

235 

Asymptomatic cement leakage occurred at 4 levels. Mean 

procedure time for one level case was 19 minutes with mean 

recorded fluoroscopy time of 6.8 minutes. IV sedation average 

dose was 1.5 mg versed and 75 mcg Fentanyl. 

Significant improvement in SF36 and VAS scores were present 

in 84% of patients. 

CONCLUSION 

Kyphoplasty can be performed safely with good patient 

comfort levels using conscious sedation. Short procedure 

and fluroscopy times can be achieved in dedicated single and 

biplane radiology suites. Clinical pain reduction is similar to 

other reported kyphoplasty and vertebroplasty studies. 

KEY WORDS: Kyphoplasty 


New Vertebroplastic Procedures: Besides the Vertebral 

Body 

Manfrè, L. 1 · Bonetti, M. 2 · Cristaudo, C. 1 

1 Azienda Ospedaliera Cannizzaro, Catania, ITALY, 2 Clinico 

Città di Brescia, Brescia, ITALY 

PURPOSE 

We investigated the use of vertebroplastic procedures in the 

treatment of sacral neoplastic and osteoporotic fracture, in 

cases of neoplastic involvement of the vertebral pedicles, 

and in patients affected by severe spinal cord compression 

and paresis. 


From October 2001 to September 2004, 204 vertebroplastic 

procedures were performed in our department. In 5 patients, 

neoplastic involvement of the pedicle occurred, with 

destruction of the lateral wall of the spinal canal. Nine 

patients had sacral neoplastic involvement and one patient 

suffered from osteoporotic sacral fracture. Two patients had 

paraparesis related to severe collapse of the seventh thoracic 

vertebral body. Both pedicle and sacral involvement were 

cemented using a CT-guided and PMMA microinjection 

technique. The two patients affected by paraparesis underwent 

intraoperative vertebroplasty in the operating theater. 

RESULTS 

After vertebroplasty, all the patients with sacral and pedicle 

involvement had reconstruction of the pedicle and the sacral 

bone, mantaining regular neural foraminal shape. No foraminal 

PMMA leakage was observed. They became painless in 24 

hours. In patients who underwent a combination of surgery 

and vertebroplasty, reduction of severe kyphosis and restoration 

of the spinal curve was obtained, with pain resolution and 

reduction of spinal cord compression. 

CONCLUSION 

Using a CT-guided vertebroplastic technique and PMMA 

microinjection, vertebroplasty procedures can be performed 

in pedicle and sacral reconstruction. In combination with 

spinal surgery, vertebroplasty can be useful in restoring a 

regular shape in case of vertebral body collapse and spinal 

cord symptomatic compression. 

KEY WORDS: Vertebroplasty, sacrum, pedicle 

Thursday

Thursday 


Treatment of Compression Fractures with a Novel 

Bioceramic; 1-Year Follow Up on the First 20 Patients 

Maurer, P. 1 · Bae, H. 2 · Westerlund, E. 3 · Peppers, T. 3 · 

Linovitz, R. 3 

1 2 Pennsylvania Hospital, Philadelphia, PA, St. John’s Health 

3 Center, Santa Monica, CA, Centralized Orthopedic 

Research and Education, San Dieguito Orthopedic Medical 

Center, Encinitas, CA 

PURPOSE 

To assess the feasibility and clinical outcome of the use of 

Cortoss in treating osteoporotic vertebral compression fractures 

(VCF). 


Cortoss is a high-strength, bioactive, radiopaque self-setting 

composite engineered to mimic the characteristics of human 

structural bone. This study evaluated the safety and effectiveness 

of Cortoss. Twenty patients enrolled with persistently 

painful ( > 6 weeks) vertebral compression fractures 

(VCF): 1-2 levels (T6-L5), X-ray < 70% height loss, acute 

findings on MR imaging or bone scan, and maintained cortex 

on CT. Outcome measures of pain (VAS), function, and 

quality of life (SF-12, ODI) preoperatively, and 1-month, 3month, 

1-year follow up. CT scans postprocedure, and plain 

X-rays at all follow-up visits. 

RESULTS 

Eight males and 12 females were enrolled with a mean age of 

72. A total of 26 levels treated (average 1.9 cc/level). Mean 

VAS decreased from 74.1 (n = 20) preop to 32 (n = 14) at 1 

year. ODI was 52 preop and 23 at 1 year. Cortoss was 

observed to have a unique trabecular fill pattern on radiologic 

studies. There were no cardiac or pulmonary complications. 

CONCLUSION 

The results obtained in this pilot IDE study indicate that 

Cortoss is safe and effective in the treatment of osteoporotic 

VCF. These results are consistent with the results obtained in 

the prospective vertebroplasty study using Cortoss conducted 

in Europe as well as others found in the literature (see 

graph). These studies suggest that a relatively small volume 

of Cortoss appears to successfully reinforce the vertebrae 

and achieve symptomatic relief, as compared to the volumes 

reported in the literature for PMMA (1). This may be due to 

the interdigitated fill pattern observed with Cortoss which 

results in trabecular reinforcement. The positive results of 

this study support pursuit of a larger, prospective randomized 

controlled investigation (currently underway). 

236 

REFERENCES 

1. Liebschner, et al. Effects of bone cement volume and distribution 

on vertebral stiffness after vertebroplasty. Spine 

2001;(26):14 

KEY WORDS: Osteoporosis, vertebroplasty, bioceramic 


3:00 PM - 4:33 PM 

Room 107 

(72c) PEDIATRICS: Vascular, Trauma 

and Tumors 




Miscellaneous 


Techniques 


Moderators: Tina Young Poussaint, MD 

Patrick D. Barnes, MD 


Functional MR Imaging of the Primary Visual Cortex in 

the Newborn 

Erberich, S. G. 1 · Panigrahy, A. 1 · Chen, V. J. 1 · Seri, I. 2 · 

Nelson, M. D. 1 · Gilles, F. 1 

1Children’s Hospital Los Angeles, University of Southern 

California, Los Angeles, CA, 2Division of Neonatal 

Medicine, University of Southern California, Los Angeles, 

CA 

PURPOSE 

Functional MR imaging (fMRI) is a well established technique 

in the adult to measure brain activation from passive 

or active tasks performed during imaging, including the primary 

visual cortex (V1). Studies have shown that the visual 

cortex develops early and myelination of optic radiation is 

present in the third trimester of gestation in the fetus. 

However imaging of early functional responses after birth 

using fMRI remains to be a challenge. The limited studies on 

visual fMRI in infants showed controversial results which 

might be caused by poor signal-to-noise. Here we present an 

fMRI protocol to test the hypothesis that fMRI is feasible 

and reliable to image activation in the primary visual cortex 

in the newborn using an MR-compatible incubator with newborn 

head-coil. 


Fifty-two preterm and term newborn patients of the NICU in

need for routine MR imaging were enrolled in this study (GA 

32-41 weeks, mean = 37 weeks). All parents gave written consent 

for this IRB-approved study. Patients were sedated before 

imaging using chloral hydrate (25-75 mg/kg). Newborns were 

prepared and placed in an incubator with the newborn headcoil. 

During imaging vital signs, movement, and sleep state 

were constantly monitored. Imaging was performed with a 1.5 

T MR system (CV/i, 9.1 software GE/MS, Milwaukee, WI). 

Functional acquisition: blood oxygen level dependent 

(BOLD) single-shot gradient-echo echo-planar imaging (GR 

EPI) sequence (TR3000, TE60, FOV180, FA90, 64 x 64 

matrix, 3 x 3 x 3 voxel resolution) and T2-weighted FSE 

images used to overlay. Functional imaging consisted of 3 

alternating epochs of control (darkness) and activation (flicker 

light at 1Hz). Statistical parametric mapping software 

(SPM99) was used for spatial preprocessing and t-test statistics. 

Areas of activation/deactivation were identified (p < 0.01) 

and mapped onto the anatomical T2 correlate. 

RESULTS 

The fMRI was performed successfully in 45 newborns without 

adverse effects. Findings showed BOLD responses in 

about 38 (~85%) of the cases; 7 cases did not respond to the 

visual task. The responding cohort shows reliable and well 

defined activation of the calcarine cortex and V1 for the 1Hz 

task, comparable to the adult brain. In addition positive and 

negative BOLD reposes significantly correlated with the 

task paradigm were identified as the prefrontal, frontal cortex, 

and in sensory-motor areas of the pre and postcentral 

gyri. Partial absence of activation in one hemisphere could 

be linked to brain pathology (e.g.. hydrocephalus). Eighteen 

newborns startled at the first onset of stimulation, but overall 

head motion was insignificant, < 4 mm (translation) and 

< 3° (rotation), and was corrected by postprocessing. 

CONCLUSION 

We found that 1Hz visual flicker light fMRI is a feasible 

technique to study brain activation in the primary visual cortex 

of the newborn. Based on this data we are investigating 

if increased frequency of 4 Hz strobe flicker light will proportionately 

increase BOLD response compared to the 1 Hz 

data, as it is known for the adult brain. The current 1 Hz data 

demonstrates 1) reliability of this technique and 2) provides 

normal baseline information of visual fMRI in the perinatal 

population. This is relevant for the potential clinical application 

to diagnosis perinatal white matter injury. 

KEY WORDS: Visual fMRI, newborn, infant 


Development of Visual Spatial Integration in Children: A 

Combined Functional MR Imaging/Magnetization 

Transfer Imaging Study 

Maeder, P. P. · Fornari, E. · Knyazeva, M. G. · Martuzzi, R. · 

Meuli, R. A. 

Centre Hospitalier Universitaire Vaudois 

Lausanne, SWITZERLAND 

PURPOSE 

In humans, visual spatial integration belongs to the higherorder 

functions, which develop slowly (Atkinson, 2000). 

Spatial integration across visual fields is based on interhemi- 

237 

spheric interaction via the corpus callosum (CC), which 

gradually matures during the first two decades of life. This 

combined functional MR imaging (fMRI) and magnetization 

transfer imaging (MTI) study was aimed at comparison 

between children and adults in terms of local neural circuits 

involved in the interhemispheric visual integration. 


In the fMRI experiments, subjects (14 children, aged 7-13 

years and 14 adults) viewed bilateral iso-oriented (IG) and 

orthogonally-oriented (OG) moving gratings alternated with 

background. The gratings were centered on a fixation point, 

had a spatial frequency of 0.5 cpd, and drifted with a temporal 

frequency of 2 Hz. Gaze fixation was monitored with an 

eye tracking system. In children, MTI was used to assess the 

brain maturation with a general index of myelination (van 

Buchem et al. 2001). We restricted our analysis to the magnetic 

transfer ratio (MTR) changes in the ROI defined as the 

CC splenium. Further, we performed an AnCova on the 

group data with individual IG activation as dependent variable, 

splenium mean MTR as covariate, and total brain mean 

MTR as nuisance variable. 

RESULTS 

Our previous experiments showed that, in adults, bilateral IG 

stimulus fusible into a single image-induced interhemispheric 

synchronization between occipital EEGs signifying interhemispheric 

integration (Knyazeva et al. 2002-2004). The 

same stimulus, compared to OG, also induced BOLD 

increase within VP/V4 areas (fusiform and lingual gyri). In 

children, the BOLD differential activation for this contrast 

was significant in the lingual gyrus although less extended. 

Assuming that the difference between children and adults 

could be due to the late myelination of the interhemispheric 

fibers, we correlated the MTI and fMRI measures in children. 

The AnCova results revealed areas of visual activation 

associated with the splenium myelination. Specifically, lingual 

gyrus activation was proportional to splenium myelination 

across subjects (P (corrected) < 0.05, Fig. 1). 

Furthermore, using MTR mean values to classify our children 

population in two groups, for the group with more 

advanced myelination, we have shown an increase in VP 

activation for the IG > OG contrast, which resulted in activation 

patterns intermediate between low myelinated children 

and adults (Fig. 2). 

CONCLUSION 

Our data suggest that interhemispheric integration is mediated 

by cortico-cortical connections. Their myelination is a 

significant factor behind the maturation of visual spatial integration. 

KEY WORDS: Development, vision, functional MR imaging 

Thursday

Thursday 

238 

Paper 436 Starting at 3:16 PM, Ending at 3:24 PM mended. Also given changes in ADC that occur with b, the b 

Optimizing Diffusion-Weighted Imaging in Neonatal 

Vascular Territory Injuries 

value of the acquired data should be reported when providing 

quantitative ADC data. 

Pectasides, M. · Pienaar, R. · Matsuda, K. · Lopez, C. J. · 

Krishnamoorthy, K. S. · Grant, P. E. 


Boston, MA 

PURPOSE 

To determine the optimal b value for neonatal focal ischemic 

brain injury. 


Eight neonates with focal vascular territory lesions on diffusion-weighted 

imaging (DWI) were included. Diffusionweighted 

imaging included multiple b values (TR = 7500 

ms, TE = min, b = 0 and at least 2 of 500, 700, 1000, 1500 

or 2000 s/mm 2 ). Raw data were eddy-current corrected and 

processed to create DWI and ADC maps. Intrasubject data 

sets at different b values were coregistered using FLIRT. 

Lesions were manually outlined on DWI images and average 

DWI signal intensity (S) determined. ∆S = difference in signal 

between DWI lesion and normal tissue. Signal-to-noise 

ratio (SNR) = S/σ, where σ = noise, calculated from an ROI 

placed in air. Contrast-to-noise ratio (CNR) = ∆S/σ. The 

lesion was outlined also on ADC maps to determine mean 

ADC and to compare the size of the ROIs to those of the corresponding 

DWI map. 

RESULTS 

Diffusion-weighted imaging maps: Noise remained approximately 

the same at all b values. Signal-to-noise ratios 

dropped with increasing b values. ∆S and CNR increased 

when b value increased from 500 or 700 to 1000 and then 

dropped with further increase of the b value to 1500 and 

2000 (Fig. 1). The size of the lesion on the DWI maps 

decreased with increasing b values. ADC maps: The size of 

the lesion on the ADC maps was the same at all b values and 

smaller than the lesion outlined on the corresponding DWI 

maps. This difference between ADC and DWI lesion volume 

was more pronounced at the b = 700 and almost nonexistent 

at 1500 (Fig. 2). The observed ADC values were lower at 

higher b values for both normal tissue and lesions. 

Discussion: At a b = 1000 s/mm 2 , CNR is highest and SNR, 

although lower than at 700, is still at acceptable levels. T2 

shine-through effects are lower at a b = 1000 s/mm 2 than at 

700 s/mm 2 and therefore the lesion size on the DWI map is 

closer to the ADC size. Although DWI lesion size was closer 

to ADC size at a b = 1500 s/mm 2 , the lower SNR and CNR 

made this b value less favorable. 

CONCLUSION 

In conclusion if a single b value is used to image focal 

ischemic injuries in neonates, a b = 1000 s/mm 2 is recom- 

KEY WORDS: Diffusion, ischemia, b value 


Prominent Lenticulostriate Vasculature following 

Hypoxic-Ischemic Injury 

Epelman, M. · Daneman, A. · Aziz, A. · Halliday, W. · 

Moodie, R. · James, A. · Blaser, S. I. 



PURPOSE 

To evaluate increased flow to the brain in neonates with HIE 

evidenced by hyperemia on color Doppler sonography 

(CDS) and by prominent lenticulostriate vessels (LSV) on 

MR imaging. 


Informed consent and ethics approval were obtained. All 

newborns who underwent MR imaging had subsequent head 

US within 2 hours. Studies of 76 neonates performed 

prospectively during a 29-month period were reviewed retrospectively 

to determine frequency of hyperemia on CDS 

and prominent LSV on MR imaging. All sequences obtained 

including MRA, MRS and diffusion-weighted imaging with 

ADC maps were reviewed. 

RESULTS 

Of the 76 neonates MR imaging was considered positive for 

HIE in 52 of whom only 35 were suspected of having HIE 

clinically. Hyperemia was noted in 17/52. Prominent LSV on 

SET2 were appreciated on 23/52. MR angiography was 

available in 2 patients confirming these findings.

CONCLUSION 

Increased flow accompanies the changes that occur in HIE. 

Only a few case reports are described in the literature about 

the use of CDS for hyperemia assessment in HIE. Prominent 

LSV on SET2 in association with HIE has not been 

described to the best of our knowledge. It is suggested that 

these findings are due to increase in vessels diameter owing 

to luxury perfusion following ischemic injury. 

KEY WORDS: Lenticulostriate vessels, hypoxic-ischemic 

encephalopathy, ultrasound 

239 


Xenon CT, MR Imaging, and CT Cerebral Perfusion in 

Children at Risk for Stroke 

Poskitt, K. J. · Heran, M. K. 

British Columbia’s Children’s Hospital 

Vancouver, BC, CANADA 

PURPOSE 

To determine the efficacy of Xenon CT and MR or CT perfusion 

in children who are at risk for stroke. 


Excluding moyamoya disease, we have identified 28 patients 

at risk for cerebral ischemia from diseases known to cause 

stroke. These patients have been followed clinically for an 

average of 3.2 years and have undergone serial investigations 

including MR imaging, MRA, MR perfusion, CT, 

CTA, CT perfusion, Xenon CT (XeCT) and angiography. 

Xenon CT studies were performed at baseline and following 

20 mg/kg acetazolamide. Absolute regional CBF and vascular 

reserve were calculated from XeCT data and compared to 

clinical outcome and perfusion values of relative CBV, CBF, 

MTT, and TTP. 

RESULTS 

Twelve of 28 children were at risk for stroke from inherited 

conditions, 7 suffered from vasculitis, 7 from trauma and 2 

cases are unclassified. Inherited conditions ranged from 

Progeria and Homocysteinuria to PHACES and Alagille’s 

syndrome. Chicken pox, Takayasu’s and lupus caused vasculitis 

while trauma was iatrogenic, inflicted, or accidental. 

In 28 patients, 62 MR images, 70 CTs, 53 XeCTs, 75 CTAs 

or MRAs, 16 perfusion studies and 39 angiograms were performed 

over 90 person years of ischemic risk. Serial XeCTs 

identified 10 periods of ischemic and 21 periods of nonischemic 

risk. Eight of 10 patients identified with ischemic 

risk suffered a stroke within 9 months of the XeCT study. 

Only 1 patient of 21 judged to not be at risk has suffered 

from a stroke. This occurred in an active, undiagnosed, and 

untreated case of Takayasu’s arteritis. Twenty of 21 cases 

have not suffered from stroke with an average follow up of 

3.2 years or more than 65 person years. When estimating 

stroke risk in this series XeCT had a 95% negative predictive 

value and 89% positive predictive value. Sixteen MR and/or 

CT perfusion studies were performed. The results were 

either incorrect or misleading in predicting stroke risk in 

9/16 (56%) cases when compared to outcome and XeCT 

data. In patients suffering from chronic vascular changes, 

perfusion studies and XeCT disagreed in 9/12 (75%) cases. 

In 3/4 cases there was agreement between perfusion and 

XeCT in acute vascular events, although 2/4 cases were classified 

incorrectly by perfusion. Overall perfusion studies had 

a 50% negative predictive value and a 38% positive predictive 

value. In this series there are 5 additional XeCT cases of 

reduced rCBF greater than 50% that did not have perfusion 

studies. These would have been expected to show prolonged 

TTP and MTT, but XeCT-documented CBF above 30 ml/100 

gm tissue/minute and no patient has suffered from a stroke. 

CONCLUSION 

In our series of children at risk for stroke from known causes 

that were inherited, traumatic or secondary to a vasculitis, 

Xenon CT accurately estimated the potential risk of each 

Thursday

Thursday 

child. Perfusion studies with either MR or CT were unable to 

accurately stratify the patient’s risk, particularly if the condition 

was chronic. 

KEY WORDS: Stroke, perfusion, pediatric 


Clinical Impact of Repeat Head CT in Pediatric Patients 

Admitted following Blunt Head Trauma 

Hollingworth, W. · Linnau, K. F. · Vavilala, M. S. · Tirschwell, 

D. L. · Johnston, B. D. · Jarvik, J. G. · Wang, M. C. 


Seattle, WA 

PURPOSE 

Serial head CT is used frequently to monitor head trauma 

patients although little information exists to help guide 

physicians in predicting which patients may benefit from 

repeat scans. We evaluated the proportion of repeat head CTs 

following blunt trauma that lead to neurosurgical interventions 

in pediatric patients. 


A retrospective study of all pediatric patients (age < 15 

years) admitted to the emergency department (ED) at 

Harborview Medical Center, Seattle from 1994-2003 who 

had head CT following blunt trauma. Data were gathered 

from the trauma registry and radiology information systems. 

Patients who underwent immediate craniotomy, transferred 

from other hospitals, or who did not have head CT within 6 

hours of ED arrival were excluded. We recorded the change 

on the repeat CT examination (improved/unchanged/worse), 

the timing of surgery, the type of surgery (evacuation of 

hematoma, decompressive craniectomy), and whether the 

repeat CT was performed routinely or as the result of other 

findings/symptoms (raised ICP, deteriorating or ongoing 

poor neurologic status). 

RESULTS 

Eight thousand five hundred and five pediatric patients were 

evaluated in the ED following blunt trauma of whom 3,586 

(42%) had at least one head CT and 1,235 (15%) had at least 

2 head CTs. After exclusions (603 transfer patients, 50 with 

initial CT > 6 hours after admission, 38 immediate craniotomy), 

544 patients were analyzed (mean age: 7.3 years (4.7 

SD), 62% male). Two hundred and seventeen (41%) patients 

had severe (ED Glasgow Coma Scale 3-8), 52 (10%) moderate 

(GCS 9-12) and 261 (49%) mild head injury (GCS 13-15) 

(14 patients had no ED GCS recorded). On average, the first 

head CT was performed 58 minutes after ED arrival and the 

second head CT was completed 12.9 hours later. There was no 

evidence that the utilization of initial or repeat head CTs 

increased over time; however the delay between first and second 

head CT diminished in the final 5 years of the study period 

(14.2 hours vs 11.4 hours; p = 0.01). Eleven (2%) of 544 

patients had evacuation of epidural, subdural, or intracranial 

hematomas following the second or subsequent head CTs and 

8 (1%) other patients had decompressive craniectomies. All 19 

patients had symptoms [neurologic decline (n = 7), seizure (n 

= 1)] or elevated intracranial pressure (n = 11) prior to surgery. 

Seven of the 19 patients (37%) had unchanged or improved 

CT findings immediately prior to surgery. 

240 

CONCLUSION 

We found that repeat head CT in pediatric patients with blunt 

head trauma is followed by craniotomy in 3% of cases. In 

these surgical cases, it appears that the repeat head CT was 

requested due to adverse clinical symptoms or high intracranial 

pressure. We found no cases where a worsening serial 

head CT, in the absence of symptoms or other findings of 

high intracranial pressure, led directly to surgery. Unchanged 

or improved CT findings did not appear to obviate the need 

for surgery in about one third of surgical cases. The frequent 

use of serial head CT involves costs and risks to the patient; 

these need to be balanced carefully against the clinical 

impact of repeat CT results on surgical and nonsurgical 

patient care. 

KEY WORDS: Brain injuries, CT, pediatrics 


Volumetric Assessment of Injury to the Anterior 

Commissure following Closed Head Trauma in Children 

Hunter, J. V. 1 · Wilde, E. 1 · Haider, J. 1 · Levin, H. S. 1 · Bigler, E. 2 

1 Baylor College of Medicine, Houston, TX, 2 Brigham Young 

University, Provo, UT 

PURPOSE 

The structure and function of the human anterior commissure 

(AC) are not well characterized, and few reports exist of 

pathologic change in the human AC following injury to the 

brain. The aim of this study is to test the hypotheses that AC 

volume would be decreased following traumatic brain injury 

(TBI) in children and that there also would be a relation 

between AC volume and 1) temporal lobe white matter 

(WM) volume and 2) temporal lesion volume. 


Sixteen children, (8 boys and 8 girls, mean age at time of 

scanning = 12.9 years, range 9-16.8, S.D.-2.5 years) who had 

sustained moderate to severe TBI, (initial Glasgow Coma 

Scale score 3-12), were imaged a mean of 3.1 years postinjury, 

(range 1-10.1, S.D.= 2.4 years). The TBI children were 

matched for age (within 6 months), sex, handedness, and 

maternal education with 16 typically developing children. 

No child had a preexisting history of head injury, neurologic 

or psychiatric disorder, or history of child abuse. Threedimensional 

sagittal T1- and T2-weighted imaging, (reconstructed 

voxel size 1 x 1 x 1 mm) was performed in addition 

to axial proton density and T2-weighted and coronal FLAIR 

imaging acquired on a 1.5 T Philips Intera, (Best, 

Netherlands) MR scanner. Area measurements of the AC 

were made electronically on the midline sagittal image and 

one parasagittal scan immediately to the left and right of 

midline respectively, using the AGFA PACS software tools. 

Four trials were performed on each slice by each of two 

trained observers with high inter and intrarater reliability, (r 

< .9). The areas were summed and multiplied by slice thickness 

to give a volume. In addition,WM volume measurements 

were performed off-line using Analyze 6.0 and previously 

established protocols, and temporal lesion volume was 

assessed on the coronal FLAIR images. Univariate analysis 

of variance was used to examine for group differences in AC 

volume. The relation in TBI children between AC section 

volumes and temporal WM and temporal lobe lesion volume

also was assessed using regression analyses which included 

total intracranial volume as a covariate. Finally, we examined 

the relation between AC size and the Metacognition 

Index of the Behavior Rating Inventory of Executive 

Function (BRIEF), a measure of working memory and the 

ability to initiate, monitor, and organize behavior in TBI 

children using a regression model. 

RESULTS 

Anterior commissure volumes as measured were significantly 

smaller in TBI patients as compared with the typically 

developing children, [F(3,28) = 5.254, p = 0.030]. 

Regression analysis revealed a statistically significant relation 

between AC size and total temporal WM, [(F(2,13) = 

5.704, p = 0.0.17, r 2 = .467, larger volumes associated with 

larger AC], as well as temporal lobe lesion volume, [F(2,13) 

= 11.176, p = 0.005, r 2 = .444, larger lesions associated with 

smaller AC]. AC volume also predicted performance on the 

Metacognition Index in TBI children [F(2,15) = 5.155, p = 

.031, r 2 = .147] where smaller AC volume predicted greater 

behavioral disturbance. 

CONCLUSION 

This is the first systematic examination of AC atrophy in 

humans following TBI and appears to demonstrate a relationship 

between white matter changes in the AC and temporal 

lobes, in association with behavioral changes, following 

TBI in children. 

KEY WORDS: TBI, anterior commissure, volumetric analysis 


Utility of Three-Dimensional Cranial CT for Evaluation 

of Nonaccidental Trauma 

Tong, K. A. 1 · Harder, S. L. 2 · Piantini, R. E. 1 · Sheridan, C. 1 

1 Loma Linda University Medical Center, Loma Linda, CA, 

2 University of Saskatchewan, Saskatoon, SK, CANADA 

PURPOSE 

To determine the value of three-dimensional (3D) cranial CT 

for distinguishing between fractures and unusual variant calvarial 

sutures in the evaluation of nonaccidental trauma. 


Two infants, ages 2 months and 8 months, presented with 

injuries suspicious for nonaccidental trauma. Due to uncertainty 

regarding the etiology of curvilinear defects in the 

squamous occipital bone on routine head CT in both cases, 

3D cranial CT was recommended. It was suspected that the 

defects might represent unusual sutures rather than fractures. 

RESULTS 

The 3D volume-rendered views better delineated the full 

extent and configuration of the defects. Interparietal bones 

were well demonstrated in both cases. One infant demonstrated 

several adjacent small sutural bones. Discussion: 

There are two components of the squamous occipital bone at 

birth; a superior or interparietal portion and an inferior or 

supraoccipital portion (1). These are separated by the mendosal 

suture(s) (1). The superior portion forms in membrane 

(2). Two paired ossification centers develop at approximately 

7 weeks of gestation and usually are fused by approxi- 

241 

mately 12 weeks of gestation (1). An unossified midline cleft 

may persist after birth and be mistaken for a fracture (3). The 

mendosal suture(s) begins to fuse at approximately the third 

fetal month and complete fusion usually occurs at approximately 

2 years of age (1). The mendosal suture(s) may 

remain open in adult life and may be mistaken also for a 

fracture (4). Three-dimensional CT has been documented as 

a useful tool in evaluating craniosynostosis and fractures (5). 

These two cases document a useful application in the setting 

of nonaccidental trauma in order to distinguish between 

sutural variants and fractures. 

CONCLUSION 

A false-positive diagnosis of nonaccidental trauma can have 

devastating consequences on the child and family. Threedimensional 

cranial CT can be a useful adjunct to conventional 

CT in the evaluation of nonaccidental head trauma, 

particularly if there is question regarding the presence of 

fractures versus sutures. 

REFERENCES 

1. Madeline LA, Elster AD. Suture closure in the human chondrocranium: 

CT assessment. Radiology 1995;196(3): 747-756 

2. Goss CM, ed. Gray’s anatomy. 29th American Edition. 

Philadelphia, Pa: Lea and Febiger 1973;158-162 

3. Franken EA, Jr. The midline occipital fissure: diagnosis of 

fracture versus anatomic variants. Radiology 1969;93:1043- 

1046 

4. Shapiro R, Robinson F. Embryogenesis of the human occipital 

bone. AJR Am J Roentgenol 1976;126:1063-1068 

5. Medina LS. Three-dimensional CT maximum intensity projections 

of the calvaria: A new approach for diagnosis of 

craniosynostosis and fractures. AJNR Am J Neuroradiol 

2000;21:1951-1954 

KEY WORDS: 3D cranial CT, nonaccidental trauma, developmental 


Quantitative in Vivo Proton (1H) MR Spectroscopy and 

Apparent Diffusion Coefficient Measurements in Pineal 

Germinomas 

Moore, K. R. · Panigrahy, A. · Nelson, M. · Krieger, M. · 

Finlay, J. · Bluml, S. 

Children’s Hospital of Los Angeles 


PURPOSE 

The quantitative MR spectral appearance of pineal germinoma 

with correlation to apparent diffusion coefficient (ADC) 

measurements has not been published previously. We characterize 

pediatric pineal germinomas using in vivo shortecho 

time (TE) single voxel 1H-MR spectroscopy with 

absolute quantitation and diffusion-weighted imaging with 

quantitative ADC measurements to identify characteristic 

features. 


Multiplanar brain MR imaging including diffusion-weighted 

imaging with ADC map and short TE proton MR spectra (1.5 

T) in 5 pediatric patients with pathology-proven intracranial 

pineal germinoma were reviewed retrospectively. Short TE 

spectra acquisition permits detection of N-acetyl aspartate 

Thursday

Thursday 

(NAA), choline (Cho), and creatine (Cr) as well as fast 

decaying components (glutamate, glutamine, myoinositol 

(mI), taurine (Tau), and lipids/macromolecules). Voxels were 

sized to include as much lesion as possible while minimizing 

inclusion of adjacent brain. Water signal was used as an 

internal reference and peak intensities were corrected for tissue 

amount within the region of interest (ROI). Automated 

quantitative analysis was performed using commercially 

available software (LC model) to generate absolute metabolite 

concentrations that were compared to institutional gray 

matter normative data. Quantitative ADC values were computed 

from ROIs placed over solid tumor regions within the 

voxel location used for MRS. Scatter plots were derived 

comparing ADC to specific metabolite concentrations. 

RESULTS 

The MRS pattern of germinoma was similar in many ways to 

observations in other brain tumors: 1) Low (due to partial volume 

average)/absent NAA peak. 2) NAA/Cr reduction and 

Cho/Cr elevation. However, several unexpected features 

were noted. Creatine concentrations were below normal; thus 

absolute Choline concentration was only mildly (1/5) to moderately 

elevated (4/5). Lipids/macromolecular peaks were 

elevated in all tumors. Taurine was identified in all 5 tumors. 

Relatively low ADC values reflected the primitive cellular 

nature of germinoma. No correlation between ADC values 

and individual metabolite concentrations was detected. 

CONCLUSION 

Pineal germinomas are characterized by elevated choline, 

lipid/macromolecular, and taurine peaks with reduced or 

absent NAA peak. Short-echo MR spectroscopy provides 

additional characterization of pediatric pineal germinoma 

with respect to taurine and lipid metabolites. The significance 

of these observations needs to be evaluated by comparison 

with the MRS pattern of other tumors (such as supratentorial 

PNET) traditionally included in the differential of 

pineal lesions, which is an on-going project. Although ADC 

measurements reflect the primitive cellular nature of these 

tumors, no correlation was found with individual metabolite 

concentrations. 

KEY WORDS: Germinoma, pineal gland, MR spectroscopy 

242 


Utility of Apparent Diffusion Coefficients in 

Differentiating Posterior Fossa Tumors in Children 

Camacho, D. L. A. 1 · Lake, D. R. 2 · Castillo, M. 1 · Rumboldt, Z. 2 

1 University of North Carolina at Chapel Hill, Chapel Hill, 

NC, 2 Medical University of South Carolina, Charleston, SC 

PURPOSE 

Three common pediatric posterior fossa tumors (pilocytic 

astrocytoma, medulloblastoma, and ependymoma) may have 

similar characteristics on contrast-enhanced MR imaging but 

differ in prognosis. The utility of diffusion-weighted imaging 

in characterizing these tumors has not been established. 

We sought to determine whether diffusion-weighted imaging 

with apparent diffusion coefficient (ADC) maps could be 

used to differentiate tumors of the posterior fossa in pediatric 

patients. 


Apparent diffusion coefficient values in regions-of-interest 

within the solid, enhancing portion of 17 pilocytic astrocytomas, 

four ependymomas, and six medulloblastomas were 

computed and compared with that of normal white matter. 

All tumors were in pediatric patients and confined to the posterior 

fossa. Apparent diffusion coefficient ratios of each 

tumor type were compared using a two-sample, two-tailed ttest. 

A P value of less than 0.05 was considered significant. 

RESULTS 

Pilocytic astrocytomas demonstrated significantly greater 

diffusion than ependymomas (P = 0.0003) and significantly 

greater diffusion than medulloblastomas (P < 0.0001). 

Ependymomas demonstrated significantly greater diffusion 

than medulloblastomas (P = 0.0005). Apparent diffusion 

coefficient ratios were 2.13 ± 0.37 (mean ± SD) (1.58-2.99 

range) in the 17 pilocytic astrocytomas, 1.33 ± 0.13 (1.15- 

1.44) in the four ependymomas, and 0.81 ± 0.13 (0.67-1.04) 

in the six medulloblastomas. The range of ADC ratios within 

each tumor type did not overlap with the range of ADC 

ratios of any other tumor type. 

CONCLUSION 

Increasing water diffusion observed in medulloblastomas,

ependymomas, and pilocytic astrocytomas likely reflects 

decreasing cellularity in these tumors. Based on our results, 

we suggest that ADC values may play an important role in 

the presurgical management of children with posterior fossa 

tumors. If high ADC values are present, a patient may go 

directly to surgery without additional imaging, given that 

pilocytic astrocytomas are unlikely to metastasize. Low 

ADC values suggest that the tumor is either a medulloblastoma 

or an ependymoma, and imaging of the spine is warranted 

to exclude metastases and appropriately stage the 

patient. 

KEY WORDS: Diffusion, tumor, pediatric 


Asymptomatic Brain Cystic Changes in Children after 

Cranial Irradiation: Frequency, Latency, and 

Relationship to Patient Age 

Kitajima, M. 1 · Hirai, T. 1 · Maruyama, N. 1 · Yamura, M. 1 · 

Hayashida, Y. 1 · Murakami, R. 1 · Ymashita, Y. 1 · Korogi, Y .2 · 

Nakamura, H. 1 

1 Kumamoto University, Kumamoto, JAPAN, 2 University of 



PURPOSE 

We have sometimes encountered asymptomatic cystic 

changes of the brain on MR imaging in patients who underwent 

cranial irradiation. To our knowledge, however, their 

MR finding has not been reported. The purpose of this study 

was to determine the frequency, latency, relationship with 

patient age, and serial changes of brain parenchymal cystic 

changes in children after cranial irradiation. 


We studied follow-up MR imaging of 6 months or more in 

33 children (20 boys, 13 girls, age range: 0-19 years, mean: 

12 years) who had undergone cranial irradiation for their primary 

brain tumors between 1990 and 2001. Primary brain 

lesions included germ cell tumor (n = 9), medulloblastoma 

(n = 8), brain stem glioma (n = 3), ependymoma (n = 3), pilocytic 

astrocytoma (n = 3), neuroblastoma (n = 1), and glioma 

arising from sites other than brain stem (n = 6). The radiation 

dose ranged from 24 to 60Gy (mean 49.3 Gy). 

Postirradiation brain cystic change on MR imaging was 

defined as a well demarcated, oval-shape, CSF-like signalintensity 

area without contrast enhancement and surrounding 

abnormal signal intensity. The frequency, distribution, size 

of the cystic lesions, and their relationship to the radiation 

dose and patient age at the time of irradiation were evaluated. 

Serial changes of the lesions on MR imaging and symptoms 

related to the cystic changes were recorded also. 

RESULTS 

Six of 33 patients (18.2 %) had one or more cystic lesions. 

The cystic change appeared within 18-84 months after cranial 

irradiation (mean 34.5 months). The cystic changes were 

found in the subcortical white matter (4 cases), periventricular 

white matter (4 cases), and the basal ganglia or internal 

capsule (2 cases). The size of the lesions ranged from 3-40 

mm (mean 8 mm), and 8 lesions gradually increased in size. 

All patients with cystic change received irradiation at the age 

243 

of 6 years or younger. They were 37.5 % of the patients who 

underwent irradiation at the age of 6 years or younger. The 

cystic change occurred within the radiation field with radiation 

dose of 36 Gy or more. Symptoms related to the cystic 

changes were not observed in any patients. There was no 

case with disappearance of the cystic lesions. No cystic 

lesions disappeared during follow-up period. 

CONCLUSION 

Asymptomatic brain parenchymal cystic changes appear to 

occur in about one fourth of children who underwent cranial 

irradiation, and frequently occur when patients received 

radiation at the age of 6 years or younger. These lesions may 

develop in various sizes. 

KEY WORDS: Radiation-induced brain damage, pediatric 

brain tumor 


Evidence of Inflammation in the Brain and Inner Ear in 

Neonatal Onset Multisystem Inflammatory Disease 

Canna, S. 1 · Goldbach-Mansky, R. 1 · Jones, J. 1 · Butman, J. A. 2 

1 National Institute of Arthritis and Musculoskeletal and Skin 

Diseases, National Institutes of Health, Bethesda, MD, 

2 Warren G. Magnuson Clinical Center, National Institutes of 

Health, Bethesda, MD 

PURPOSE 

To describe the CNS MR imaging findings in neonatal onset 

multisystem inflammatory disorder (NOMID), an autoimmune 

disorder caused by a mutation in CIAS1, resulting in 

upregulation of the IL-1 pathway. Central nervous system 

manifestations include developmental delay, chronic meningitis, 

elevated intracranial pressure (ICP), and sensorineural 

hearing loss (SNHL). 


Eighteen patients with clinically defined NOMID (urtcarial 

Thursday

Thursday 

rash, CNS manifestations, characteristic bony overgrowth of 

the epiphyses) were evaluated with inner ear and brain MR 

imaging including pre and postcontrast FLAIR, fourteen 

patients with FIESTA. Of our patients 65% had mutations in 

CIAS1. 

RESULTS 

Evidence of inner ear inflammation was marked by the presence 

of inner ear enhancement which was identified on postcontrast 

FLAIR in 14 patients. No structural ear abnormalities 

were present. Evidence of meningitis manifest as 

enhancement of the dura or leptomeninges was present in 5 

patients, best seen on postcontrast FLAIR. Arachnoid webs 

were identified on FIESTA sequences in 10 of the 14 cases 

in which FIESTA was performed. These webs were not well 

seen on corresponding FSE T2-weighted sequences. Cortical 

venous enhancement was present in three cases. In one case, 

extensive enhancement of deep parenchymal veins in the 

white matter was present. On delayed imaging in this case, 

diffuse enhancement of the white matter was present. 

Papilledema could be identified on FLAIR in 6 of 18 cases. 

Two patients had been shunted. Of the remaining 16, eight 

had ventriculomegaly. Heterotopic gray matter was present 

in one case. A Chiari I malformation was present in one case. 

Patients with and without mutations in CIAS1 did not differ 

in regards to the findings described above. 

CONCLUSION 

Although chronic meningitis has been reported in NOMID, 

our studies expand the spectrum of CNS MR imaging findings 

in NOMID to include dural and leptomeningeal 

enhancement, cortical venous enhancement, parenchymal 

venous enhancement, arachnoid webs, and enhancement of 

the labyrinth. Many of these findings were demonstrated 

best on sequences not considered “routine,” including postcontrast 

FLAIR of the inner ear and FIESTA of the subarachnoid 

space. Since NOMID patients develop progressive 

SNHL, the observed enhancement of the labyrinth suggests 

an inflammatory etiology. Furthermore, the presence of 

arachnoid webs suggests that chronic meningeal inflammation 

may obstruct CSF flow, and hence account for increased 

intracranial pressure. Whether these findings are predictive 

of clinical course (e.g., cognitive function, sensorineural 

hearing loss) or may be used to assess the effectiveness of 

therapy, are currently under investigation. 

KEY WORDS: Cochlea, papilledema, FLAIR 

244 


3:00 PM - 4:30 PM 

Room 205 

(72d) PEDIATRICS: Vascular and 

Miscellaneous 




Miscellaneous 


Techniques 


Moderators: Nancy K. Rollins, MD 

Thomas P. Naidich, MD 


Safety of Cerebral Digital Subtraction Angiography in 

the Pediatric Age Group: Retrospective Analysis of 188 

Consecutive Diagnostic Cerebral Angiograms 

Burger, I. · Jordan, L. · Murphy, K. J. · Gailloud, P. 


Baltimore, MD 

PURPOSE 

Noninvasive imaging techniques have reduced the number 

of cerebral angiograms performed for purely diagnostic reasons. 

Digital subtraction angiography (DSA) remains, however, 

the most accurate technique for the evaluation of the 

cerebrovascular system. However, DSA in children is often 

perceived as particularly challenging or dangerous by referring 

physicians, and frequently delayed, if obtained at all, 

even though the information it provides could be invaluable 

for timely diagnosis and decision-making. The frequency of 

periprocedural complications often is brought as a question 

at the time of management decision. Although older studies 

have reported relatively low complication rates for pediatric 

cerebral angiography, no recent estimation of the complication 

rate for modern cerebral DSA is available. This report 

looks at the complication rate observed in a series of 188 

consecutive diagnostic cerebral angiograms. 


This report is based on the analysis of an IRB-approved database 

of all pediatric patients who underwent diagnostic cerebral 

angiography at a single institution from January 1999 to 

November 2004. The data selected for this review consist of 

188 consecutive diagnostic cerebral angiograms (602 vessels 

selectively studied) performed in 155 children (mean age: 13 

years, range: 0 to 18). Cerebral angiograms obtained as the 

initial diagnostic component of a therapeutic procedure were 

not included.

RESULTS 

Of the total 188 angiograms, 186 (99%) were uneventful. 

One adverse event (0.5%) and one mortality (0.5%) were 

observed. An 18-year-old girl with cerebral vasculitis and 

Raynaud phenomenon involving both hands and feet, suffered 

vasospasm/thrombosis of her right foot arteries necessitating 

partial amputation 7 days after an unevenftul 

angiogram. A 7-year-old girl investigated for a grade IV 

dural arteriovenous fistula presenting with intracranial hemorrhage 

reruptured a posterior fossa varix and died 3 hours 

after an uneventful angiogram. No cerebral thromboembolic 

(ischemic) events and no intraprocedural deaths were noted. 

CONCLUSION 

Digital subtraction angiography is safe in the pediatric population. 

No intraprocedural adverse event, in particular no 

cerebral thromboembolic complication, was documented in 

our series. 

KEY WORDS: Cerebral angiography, safety, pediatric 


Contribution of MR Digital Subtraction Angiography to 

the Assessment of Brain Abnormalities in Children 

Griffiths, P. D. 

University of Sheffield 


PURPOSE 

There are several practical difficulties in performing catheter 

cerebral angiography in children. One major goal of current 

research is to evaluate alternative methods to detect or 

exclude high-flow vascular abnormalities. We describe our 

early experience in combining a dynamic MR digital subtraction 

angiographic (MR DSA) method with routine MR 

imaging and time-of-flight angiography in children. 


This is a retrospective, descriptive study of 12 children (ages 

ranging from 5 days to 16 years) with suspected intracranial 

vascular abnormalities referred for clinical MR imaging to a 

tertiary/quaternary pediatric center in the UK. Routine MR 

imaging, time-of-flight MR angiography and MR DSA were 

performed in all cases. The dynamic sequence is a thick (6- 

10 mm) slice selective RF spoiled fast gradient-echo 

sequence (RF- FAST). Sixty frames (1/sec) were acquired in 

two or three planes, before and during passage of Gd-DTPA 

bolus. 

RESULTS 

MR DSA was performed successfully in all cases and yielded 

useful information in each case. High-flow lesions detected 

on MR DSA in three cases (vein of Galen aneurysm, 

transverse sinus dural fistula, incompletely treated arteriovenous 

malformation). Low flow was found in three lesions 

(ependymoma, and two cases of cavernomas). No flow was 

confirmed in two treated lesions (a treated AVM, and a treated 

vein of Galen aneurysmal malformation). 

CONCLUSION 

Our experience suggests that MR DSA is a useful supplementary 

examination in pediatric neuroimaging due to its 

245 

ability to provide temporal flow characteristics, which previously 

was only available through catheter angiography. 

KEY WORDS: MR DSA 


Intracranial Aneurysms in Children: Single-Center 

Experience of 75 Aneurysms in 59 Consecutive Patients 

Cullen, S. P. 1,2 · Wuppalapati, S. 3 · Alvarez, H. 3 · Rodesch, G. 3 

· Ozanne, A. 3 · Lasjaunias, P. 3 

1 2 Brigham and Women’s Hospital, Boston, MA, Children’s 

Hospital, Boston, MA, 3Hopital Bicetre, Kremlin-Bicetre, 

FRANCE 

PURPOSE 

To review the single-center experience in diagnosis and 

treatment of a consecutive series of intracranial aneurysms 

seen in children 15 years of age or under. 


Retrospective review of imaging, clinical, and treatment data 

that had been prospectively entered into a departmental database. 

RESULTS 

Fifty-nine patients with 75 separate lesions were identified. 

There were an equal number of children in four age groups: 

below 2 years (22%), 2-5 years (24%), 6-10 years (24%), 

and 11-15 years (30%). Thirty-three children had dissecting 

aneurysms, two had chronic posttraumatic aneurysms, 8 had 

infectious aneurysms, and 16 patients had saccular lesions. 

Twenty-seven percent of lesions were in the posterior circulation, 

and 21% developed on the middle cerebral artery. 

Most dissecting lesions were seen in the vertebrobasilar system, 

while saccular lesions were present mostly in the anterior 

circulation. Half of all cases presented with hemorrhage. 

Hemorrhage in patients below 2 years of age was due to dissecting 

aneurysms, while saccular aneurysms were responsible 

for hemorrhage in patients above 5 years of age. Five 

children had familial disease and 9 presented with multiple 

aneurysms. We were referred 48 children for treatment. 

Thirty-two underwent either surgical (21.9%), endovascular 

(62.8%), or combined (9.3%) treatment. Eleven patients 

were treated conservatively and in 5 patients the aneurysms 

had thrombosed spontaneously at admission. Overall, complete 

or partial spontaneous thrombosis was seen in 10 

patients (16.9%). Dissecting aneurysms were frequent in 

children of all ages with either associated thrombosis or arterial 

tear with repeated acute hemorrhage and poor outcome. 

Two types of dissection seem identifiable despite the small 

number of cases collected: (1) acute segmental arterial tear 

without thrombosis, acute SAH and recurrence before 5 

years, and (2) subacute focal dissection with partial thrombosis 

(or mural hematoma), rare SAH and no early recurrence. 

The former was felt to warrant aggressive management 

whereas the latter were treated conservatively. The 

mortality in our series of aneurysms was low in the treated 

group (10.4%). The overall tolerance to hemorrhage seems 

better than in adults as already stressed in the literature. 

CONCLUSION 

Intracranial aneurysms in children comprise a heterogenous 

Thursday

Thursday 

group of lesions. Treatment strategy depends on aneurysm 

etiology. Outcome can be favorable with therapy appropriate 

for aneurysm subtype. 

KEY WORDS: Aneurysm, pediatric 


Spinal Arteriovenous Shunts Presenting before 2 Years of 

Age: Analysis of 13 Cases 

Cullen, S. P. 1,2 · Alvarez, H. 3 · Rodesch, G. 3 · Ozanne, A. 3 · 

Lasjaunias, P. 3 

1 Brigham and Women’s Hospital, Boston, MA, 2 Children’s 

Hospital, Boston, MA, 3 Hopital Bicetre, Kremlin-Bicetre, 

FRANCE 

PURPOSE 

A minority of patients with spinal arteriovenous malformations 

(SAVM) are symptomatic as neonates or infants. We 

analyzed the clinical and anatomical factors associated with 

this early presentation, and reviewed our experience treating 

patients with these lesions. 


A retrospective review of clinical records and imaging studies 

was performed for patients with SAVMs who presented 

before 2 years of age and who were evaluated by the diagnostic 

and interventional neuroradiology service at our institution. 

Clinical, imaging, and treatment data had been 

entered prospectively into a department database. 

RESULTS 

Thirteen patients were identified with SAVMs that were 

either diagnosed or that became symptomatic before 2 years 

of age (9 male, 4 female). These represented 13% of the 

SAVMs seen during the same period of time. The mean age 

at presentation was 6.9 +/- 6.4 months. Eleven of 13 patients 

had neurologic symptoms attributable to the spinal lesion. 

Presentation was nonhemorrhagic in 9 patients, and associated 

with hemorrhage in 4. Ten lesions were fistulae (SAVF) 

(77%); 2 were the nidus type of malformation. There was a 

syndromic association in 8 patients: hereditary hemorrhagic 

telangectasia (HHT) in 6, all lesions but one, were intradural 

high-flow peri-medullary SCAVFs (46% of overall SAVM 

and 56% of SCAVF), and spinal arteriovenous metameric 

syndrome (SAMS) in 2. One patient had Hirschprung’s 

anomaly. Eight patients underwent endovascular treatment 

alone, one had surgery and embolization, and four were not 

treated. In all patients undergoing embolization, n-BCA liquid 

adhesive was used. Of those patients who underwent 

endovascular treatment, in 7 the lesion was completely obliterated, 

and in the remaining 2, a 90% reduction in nidal size 

was achieved. There was one treatment complication (infection) 

which resolved with medical therapy. No procedurerelated 

permanent morbidity or mortality was seen. Follow 

up was available (mean 23.7 +/- 18.7 months) in 10 patients 

(9 treated). All treated patients were either stable or 

improved, with none showing further deterioration following 

treatment. 

CONCLUSION 

Factors associated with early presentation in neonates and 

infants with spinal arteriovenous shunts include the presence 

246 

of high-flow, solitary fistulae, and HHT. Specifically, the 

presence of SCAVF in a child before 2 years of age is highly 

suggestive of HHT. Despite the aggressive nature of these 

lesions, many are amenable to endovascular treatment, and 

this is associated with a favorable posttreatment course in 

most cases. 

KEY WORDS: Spinal AVM, pediatric 


In Vivo Quantitative Proton MR Spectroscopy of the 

Neonatal Brainstem: Correlation with Glutamate 

Receptor Binding Development in Vitro 

Panigrahy, A. 1 · Nelson, M. D. 1 · Liu, X. 1 · Friedlich, P. 1 · Seri, 

I. 1 · Kinney, H. C. 2 · Bluml, S. 1 

1 Children’s Hospital Los Angeles, Los Angeles, CA, 2 Boston 

Children’s Hospital, Boston, MA 

PURPOSE 

Glutamate is involved in mechanisms of perinatal hypoxicischemic 

brain injury. The purpose of this study was (1) 

determine “normal” developmental changes in vivo of 

selected metabolites ( including glutamate) in the neonatal 

brainstem using quantitative short echo MR spectroscopy in 

contrast to occipital gray matter and (2) to correlate the findings 

with previously published human brainstem glutamate 

receptor binding data (1). 


One hundred and twenty-three MR spectra of 98 patients 

with “normal” MR imaging were analyzed (49 brainstem 

spectra and 74 spectra of occipital gray matter). MR spectroscopy 

was performed on a 1.5 T GE clinical scanner. 

Neonates were studied using an MR-compatible incubator 

with integrated head coil (Advanced Imaging Research Inc., 

Cleveland, OH). Single voxel 1H spectra were obtained 

using PRESS sequence with short echo time of TE = 35 ms, 

repetition time TR = 1.5 s, and 128 or 192 signal averages. 

Spectra were processed using automated LCModel software 

(Stephen Provencher Inc., LCModel V6). Concentrations 

were corrected for the fraction of CSF volume within the 

ROI. Data were correlated with previously published glutamate 

receptor subtype binding data performed in frozen 

autopsy tissue using receptor autoradiography (1). 

RESULTS 

A typical MR spectrum of the neonatal brainstem is shown 

in the figure. In contrast to occipital cortex, metabolite concentrations 

in the brainstem exhibited less significant agedependent 

changes (Table). Glutamate concentrations 

remained stable across development in the brainstem, in contrast 

to a dramatic increase in glutamate in the occipital cortex. 

Despite relative changes in glutamate receptor binding 

subtypes in vitro between the neonatal period and maturity, 

the concentration of glutamate in vivo remains relatively stable.

Comparison of metabolite concentration (mmol/kg) (+SD) of brainstem and 

occpital gray matter 

REGION AGE (PCA= NAA Creatine Choline Myo- Glx Glu Gln 

postconception- inositol (glutamate (glutamate) (glutamine) 

al age) & glutamine) 

brainstem

Thursday 


Functional MR Imaging, Diffusion Tensor Fiber 

Tracking, and Surgical Cortical Mapping of Pediatric 

Tumors and Focal Cortical Lesions 

Lee, B. C. P. 1,2 · Leonard, J. R. 1,2 · Smyth, M. 1,2 · Mori, S. 3 

1Saint Louis Children’s Hospital, St. Louis, MO, 

2Washington University School of Medicine, St. Louis, MO, 

3Johns Hopkins University Medical School, Baltimore, MD 

PURPOSE 

To evaluate the role of functional MR imaging (fMRI) and 

diffusion tensor fiber tracking (DTI FT) in surgical management 

of pediatric brain tumors and focal cortical lesions. 


Nine patients, ages 18 months to 10 years (5 hemispheric, 1 

thalamic, 1 brainstem tumors and 2 focal cortical dysplasias), 

were studied. Scans were performed on Siemens 1.5 

Sonata (6 patients) and 3 T Trio (3 patients) scanners. 

Functional MR images were obtained in 8 and DTI FT in 4 

patients. Three-dimensional MPRAGE were obtained in all 

the patients. Motor fMRI using finger-thumb tapping was 

performed in 6 patients. Expressive language fMRI using the 

word generation paradigm was performed in two patients 

(one also had motor fMRI). One sedated patient was studied 

using passive tactile hand stimulation. Diffusion tensor 

imaging fiber tracking was performed in 3 patients after 

fMRI, one patient did not have fMRI. Functional MR imaging 

data were processed using Brain Voyager and the activated 

regions were displayed on surface-rendered 3D images 

of the brain in 6 patients. The functional areas shown on 

fMRI were used as “seed points” for fiber tracking; DTI FT 

was processed using the program developed by one of the 

authors (Dr. Mori). Surgical mapping was performed in four 

patients and operative photographs were taken. The relevant 

sulci and gyri were identified on both the operative pictures 

and the 3D surface images. The operative picture was placed 

over the 3D fMRI image which was manipulated until the 

landmark sulci/gyri in both exams were superimposed. 

RESULTS 

Motor, sensory, and expressive language regions were demonstrated 

clearly in all patients. The hand sensory cortex was 

shown in the sedated patient. Direct comparison of the motor 

and language areas on fMRI with surgical mapping showed 

very accurate correlation (Fig. A). Fiber tracking of the hand 

and feet motor function clearly demonstrated the position of 

these fibers relative to the margin of the tumor (Fig. B). 

248 

CONCLUSION 

Functional MR imaging and fiber tracking can be performed 

in conscious and sedated children. Functional MR imaging is 

useful for detecting eloquent cortical functional regions and 

fiber tracking for assessing the positions of the projection 

fibers relative to the tumor within the white matter. Using 

both methods for surgical planning may result in more precise 

excisions of tumors and other lesions in the children. 

KEY WORDS: Functional MR imaging, diffusion tensor fiber 

tracking, pediatric tumors 


Down Regulation of PTEN and P27 Tumor Suppressor 

Genes in Pediatric Brain Tumors Correlates with 

Preoperative Lower Levels of Myoinositol/Creatine 

Ratio Using Short-Echo Time MR Spectroscopy 

Altinok, D. 1 · Altinok, G. 2 · Rabe, A. 1 · Mody, S. 1 · Zerin, M. 

J. 1 · Rabah, R. 1 

1Children’s Hospital of Michigan, Wayne State University, 

Detroit, MI, 2Hacettepe University School of Medicine, 


PURPOSE 

Proton nuclear MR spectroscopy (MRS) is a noninvasive 

technique that allows in vivo measurements of certain tissue 

metabolites. A recent study in our radiology department suggested 

that MRS using measurements of the 

myoinositol/creatine (M/C) ratio would be helpful in differentiating 

low- and high-grade brain tumors in children. The 

PTEN tumor suppressor gene plays an important role in the 

phosphatidyl-inositol-3-kinase signaling pathway and contributes 

to cell cycle regulation by blocking entry of cells 

into S phase. Inactivation of the PTEN gene has been 

described in several cancers including glioblastoma multiforme, 

but not in low-grade tumors. The aim of this study is 

to compare preoperative MRS results with the tumor morphology 

and the expression of PTEN and p27 in pediatric 

brain tumors. 


Preoperative levels of myoinositol/creatinin ratio were 

obtained from 19 patients using short-echo time single-voxel 

MRS. Histologic features and expression of PTEN and p27 

of the same cases were reviewed. Intense staining of vessel 

walls with PTEN was used as an internal control. 

RESULTS 

The tumors included 11 medulloblastoma and 8 low-grade 

astrocytoma. Using MRS the median value of (M/C) ratio 

was 0.81 in normal brain, 1.10 in medulloblastoma, and 1.97 

in low-grade astrocytoma cases. The M/C ratio in low-grade 

astrocytoma was higher than in medulloblastoma and normal 

brain. Nuclear p27 staining and cytoplasmic PTEN staining 

were both diffuse in low-grade astrocytoma, with a median 

of 80% (range: 80% to 90%). However, among medulloblastomas 

distributions of both markers were patchy; p27 

stained with a median of 20% (range: 1% to 60%), and 

PTEN with a median of 50% (range: 30% to 60%) and both 

markers were expressed mainly in the areas of neuronal differentiation.

CONCLUSION 

In low-grade astrocytoma the M/C ratio was higher than in 

medulloblastoma and both p27 and PTEN were diffusely 

expressed. Decreased immunostaining with p27 and PTEN 

in medulloblastoma cases correlated with lower preoperative 

M/C ratios. These results suggest that inactivation of PTEN 

and p27 tumor suppressor genes play an important role in the 

development of high-grade brain tumors, and MRS using 

M/C ratios appears helpful in differentiating low-grade from 

high-grade in pediatric intraaxial neoplasms. 

KEY WORDS: MR spectroscopy, pediatric neoplasms 


Diffusion Tensor Imaging in Children with 

Periventricular Leucomalacia: Variability of Injuries to 

White Matter Tracts 

Nagae-Poetscher, L. M. 1,2 · Hoon, A. H. 2 · Stashinko, E. 2 · 

Levey, E. 2 · Wakana, S. 2 · Jiang, H. 1 · van Zijl, P. C. M. 1 · 

Johnston, M. V. 2 · Mori, S. 1 

1Johns Hopkins University School of Medicine, Baltimore, 

MD, 2Kennedy Krieger Institute, Baltimore, MD 

PURPOSE 

To apply a grading system based on visual inspection of 26 

white matter tracts identified on color-coded maps, generated 

from diffusion tensor imaging (DTI) acquisition, to a 

group of patients with periventricular leukomalacia (PVL). 

The tracts were scored 0 when normal; 1, when clearly 

reduced in size; and 2, when severely abnormal or almost 

absent. The purpose of this study is to evaluate white matter 

injury pattern in PVL patients as a potential tool to better 

classify cerebral palsy (CP) patients, and hopefully, in the 

future, provide specific treatment options based on the white 

matter pattern injury presented by an individual patient. 


Nine patients with diagnosis of PVL (mean age: 5 years 2 

months, age range: 16 months-13 years 3 months) were 

scanned. Diffusion tensor imaging was acquired with sensitivity 

encoding (SENSE) technique (reduction factor of 2.5): 

[SE single-shot EPI; TR/TE of 6.2-9.4 s/80 ms; b max of 700 

s/mm 2 ; 30 different gradient directions with 5 b0s; 50 slices; 

field of view (FOV) adjusted to the brain size with imaging 

matrix of 80 x 80 to 96 x 96, zerofilled to 256 x 256; original 

in-plane imaging resolution of 2.0 - 2.5 mm; scan time: 

4 min 18 s-6 min 34 s/sequence; 3 repetitions]. Criteria were 

established for identification of 26 white matter tracts using 

in-house software (DtiStudio). Comparison was made to 

normal controls. Scoring was performed by two raters in 

consensus. 

RESULTS 

Nineteen tracts were found to be involved in this population 

of PVL patients, including: corticopontine/corticospinal 

tracts (Fig. B), posterior limb (Fig. A, arrow heads) and 

retrolenticular part of the internal capsule (Fig. A, asterix), 

cerebral peduncles, posterior thalamic radiation, middle 

cerebellar peduncles, decussation of the superior cerebellar 

peduncles, thalamus, inferior fronto-occipital/inferior longitudinal 

fasciculi, superior longitudinal fasciculus, 

uncinate/inferior fronto-occipital fasciculus, inferior and 

249 

superior cingulum, superior fronto-occipital fasciculus, inferior 

fornix, superior corona radiata, corpus callosum, tapetum, 

and anterior commissure. There was marked heterogeneity 

on the pattern of white matter injury pattern, with the 

most affected fibers being the retrolenticular part of the 

internal capsule, tapetum, posterior thalamic radiation, body 

of the corpus callosum, and superior corona radiata. 

CONCLUSION 

Use of color-coded DTI maps is feasible for identification 

and visual inspection of white matter tracts. This approach 

can benefit PVL and other patients with neurologic deficits, 

improving detection of injury to specific tracts. Information 

on white matter injury pattern potentially can contribute to 

treatment plan. 

KEY WORDS: Diffusion tensor imaging, periventricular 

leukomalacia, children 


Tractography in Pediatric Brain Stem Glioma Patients: 

Wallerian Degeneration of Corticospinal and Sensory 

Pathways 

Helton, K. J. · Weeks, J. · Phillips, N. · Zou, P. · Li, C. · Kun, 

L. · Khan, R. · Gajjar, A. · Ogg, R. 

St. Jude Children’s Research Hospital 

Memphis, TN 

PURPOSE 

While diffusion tensor imaging (DTI) can detect early wallerian 

degeneration of pyramidal tracts after stroke, its utility has 

not been evaluated systematically in brain tumors. We hypothesized 

there would be diminished fractional anisotropy (FA) 

distal to the tumor, providing evidence of wallerian degeneration 

in pediatric brainstem glioma patients. 


A retrospective study of 7 brainstem glioma patients compared 

to 8 normal controls was performed. Diffusion tensor 

imaging data acquired were normalized to Talairach space to 

provide reference points for the three-dimensional proportional 

grid system. Regions of interest (ROIs) were carefully 

selected in precise locations of the bilateral motor and sensory 

tracts at the pons, midbrain, internal capsule/VPL, corona 

radiata, and subcortical white matter utilizing FA and 

color maps, which identified locations using x,y,z coordi- 

Thursday

Thursday 

nates of each voxel. An iterative process, using Talairach’s 

Co-Planar Stereotaxic Atlas of the Human Brain and the 

Thalamic Connectivity Atlas, was utilized to ensure precise 

placement within the coordinates with the highest probability 

of motor and sensory connections. Fractional anisotropy 

and ADC values were calculated for each ROI. Additional 

regions were sampled in the upper pons of tumor patients, 

for a total of 336 patient and 320 control data points. 

Nonparametric statistical methods (Wilcoxon-Mann- 

Whitney test) were used for all data analyses. 

RESULTS 

There were statistically significant differences (p < 0.05) in 

FA between the patient and control groups flowing up and 

down both motor (bilateral pons, upper pons, left midbrain, 

left subcortical white matter, and right corona radiata) and 

sensory (bilateral pons, left upper upper pons, bilateral midbrain, 

thalamus, internal capsule, and subcortical white matter) 

tracts. There was a trend toward more pronounced alterations 

in motor FA closest to the tumor with gradual increases 

in FA distal to the tumor. There were also scattered areas 

of significant ADC differences between the two groups. 

CONCLUSION 

Our data support the hypothesis of altered FA distal to the 

tumor, providing evidence of wallerian degeneration in 

motor and sensory pathways in this patient cohort. These 

findings are consistent with a growing body of evidence that 

FA is a useful method for evaluating wallerian degeneration. 

The rigorous methods we utilized may be useful for investigating 

FA changes in other central nervous system processes. 

KEY WORDS: Diffusion tensor imaging, brainstem tumor, 

tractography 


Reduced Myo-Inositol Ratio in the Acute Phase of Acute 

Disseminated Encephalomyelitis 

Ben-Sira, L. · Miller, E. · Constantini, S. · Ben Bashat, D. 

Souraski Medical Center 

Tel Aviv, ISRAEL 

PURPOSE 

Acute disseminated encephalomyelitis (ADEM) is a 

demyelinating disorder that includes varying degrees of 

mental state changes. Often associated with a preceding systemic 

infection or immunization in children and young 

250 

adults. MR lesions demonstrate high-signal asymmetrical 

bilateral patchy areas on T2/FLAIR sequences mostly in the 

cerebral, subcortical white matter of the posterior fossa, usually 

with no mass effect. Differential diagnosis with other 

demyelinating disorders like multiple sclerosis may be difficult 

using conventional MR imaging alone. Advanced MR 

techniques such as proton MR spectroscopy (MRS) imaging 

might have an important role in understanding the underlying 

pathology. 


Six patients with ADEM were scanned either in the acute 

and/or chronic phase of the disease. Patients were 3 - 14 

years of age (mean = 9 years). MR spectroscopy was 

obtained on a 1.5 T GE MR scanner (GE Signa Horizon, 

Echo speed, LX MR scanner, Milwaukee, WI). MR imaging 

protocols included conventional imaging and proton spectroscopy 

- single voxel PRESS sequence of both lesions and 

NAWM using short TE (TE = 35 msec). Metabolite values 

and ratio were calculated automatically from the area under 

each metabolite peak by using the standard commercial software 

program provided by the manufacturer. 

RESULTS 

In the chronic phase of the disease all spectra in the lesions 

demonstrated a reduction in the NAA/Cr ratio and elevated 

Cho/Cr ratio as previously reported. Also noted in those 

spectra, elevated myo-inositol (MI) to Cr ratio, similar to 

those detected in MS patients. During the acute phase of the 

disease, a reduction of NAA/Cr and elevated Cho/Cr ratio 

was observed, but there was a new finding - a significant 

reduction in the MI/Cr ratio. Figure 1 shows data of one 

patient. Similar results were obtained for all other patients. 

Fig. 1: Seven-year-old boy, with involuntary movements and 

loss of consciousness. Fig. 1a spectrum taken in the acute 

phase and Fig. 1b shows spectrum obtained in the chronic 

phase. During the acute phase, MI/Cr ration is reduced (0.50 

compared to 0.73 ± 0.7 in normal) and elevated in the chronic 

phase (0.98 compared to 0.73 ± 0.7 in normal). 

CONCLUSION 

Reduction in the MI/Cr ratio is detected during the acute 

phase of ADEM disease compared to elevated ratio during 

the chronic phase. This finding can be helpful in the acute 

episodes to differentiate between ADEM and other disorders 

such as MS. 

KEY WORDS: ADEM, myo-inositol, white matter disease


4:40 PM - 6:10 PM 

Room 105/106 

(73) Diffusion and Other Techniques in 

Spinal MR Imaging (ASSR) 

(457) In-Phase, Out of Phase Imaging: Can You 

Differentiate Metastases from Osteoporotic 

Fractures? 

⎯ William K. Erly, MD 

(458) Diffusion Imaging in the Spine and Spinal 

Cord 

⎯ Roland Bammer, PhD 

(459) Diffusion Tensor Imaging in the Spinal Cord 

⎯ Eric D.Schwartz, MD 

Moderators: William K. Erly, MD 

Carl E. Johnson, MD 

In-Phase, Out of Phase Imaging: Can You Differentiate 

Metastases from Osteoporotic Fractures? 


Diffusion Imaging in the Spine and Spinal Cord 

Roland Bammer, PhD 

Dr. Roland Bammer currently is leading the pediatric MR 

imaging research program at the Lucille Packard Children’s 

Hospital at Stanford University. He is an Assistant Professor 

of Radiology at Stanford University. He received his 

Doctoral Degree in Electrical Engineering with a special 

focus on Biomedical Engineering from the Technical 

University in Graz, Austria. He holds also the VENIA 

DOCENDI (a professorship) for Medical Physics and 

Biophysics at the Medical University of Graz, Austria. 

Before he started his position as Assistant Professor at 

Stanford Dr. Bammer worked for 4 years as Research 

Associate in Radiology at the Richard Lucas Center, 

Department of Radiology, at Stanford University. He has 

written over 35 articles, 6 book chapters, and more than 100 

abstracts of which 6 articles and chapters deal specifically 

with diffusion-weighted MR imaging of the spine and spinal 

cord. He is an expert in all aspects of diffusion imaging and 

rapid imaging methods, including parallel imaging. 

251 



able to: 

1) Understand the principles of diffusion-weighted imaging 

(DWI) and diffusion tensor imaging (DTI). 

2) Compare different acquisition methods for DWI and identify 

the optimal method for a specific clinical question. 

3) Illustrate the potential strengths and weaknesses of DWI 

for abnormalities in the spinal column and the spinal cord. 


Basic principles of diffusion-weighted imaging (DWI) and 

diffusion tensor imaging (DTI) will be addressed. The 

advantages and disadvantages of various pulse sequences for 

DWI of the spine will be discussed and potential pitfalls will 

be shown. The role of DWI in the assessment of abnormalities 

of the spinal column, as well as spinal cord will be 

addressed. Emphasis will be on the utility of DWI in evaluating 

vertebral compression fractures and protocols will be 

reviewed briefly. Features which may aid in distinguishing 

benign from metastatic lesions will be reviewed. The added 

value of DWI in assessing ischemic and inflammatory diseases 

as well as traumatic injury of the spinal cord also will 

be emphasized. The utility of fiber tracking in the spinal cord 


REFERENCES 

1. Basser PJ. Inferring microstructural features and the physiological 

state of tissues from diffusion-weighted images. 

NMR Biomed 1995;8:333-344 

2. Bammer R, Augustin M, Prokesch RW, Stollberger R, Fazekas 

F. Diffusion-weighted imaging of the spinal cord: interleaved 

echo-planar imaging is superior to fast spin-echo. J Magn 

Reson Imag 2002;15:364-373 

3. Bammer R, Herneth AM, Maier SE, et al. Line scan diffusion 

imaging of the spine. AJNR Am J Neuroradiol 2003;24:5-12 

4. Bammer R, Fazekas F. Diffusion imaging of the human spinal 

cord and the vertebral column. Top Magn Reson Imag 

2003;14:461-476 

Diffusion Tensor Imaging in the Spinal Cord 




able to: 

1) Discuss the histologic factors thought to result in 

anisotropic water diffusion in the spinal cord white matter. 

2) Discuss how measures of water diffusion and anisotropy 

obtained with diffusion-weighted MR imaging and diffusion 

tensor MR imaging reflects white matter integrity. 

3) Discuss the potential clinical applications of diffusionweighted 

MR imaging and diffusion tensor MR imaging in 

evaluating spinal cord pathologies. 

4) Discuss some strategies for optimizing diffusion tensor 

imaging of the spinal cord. 


Conventional MR imaging techniques have had only limited 

success in the evaluation of diseases that affect spinal cord 

white matter tracts, such as trauma (1) or demyelination (2), 

possibly because these techniques are limited to evaluating 

gross changes in the spinal cord, such as edema, hemorrhage 

Thursday

Thursday 

or blood-spinal cord barrier breakdown. The degree of preserved 

white matter and its location, however, are what primarily 

determine subsequent function. Therefore, another 

imaging technique is required to evaluate the integrity of the 

white matter. The anisotropic nature of water diffusion in 

central nervous system white matter makes diffusionweighted 

imaging (DWI) and diffusion tensor imaging (DTI) 

a promising modality for evaluating spinal cord white matter 

(3, 4). Experimental evidence has shown that perturbations 

in diffusion and anisotropy within spinal cord white matter 

water are more sensitive than conventional MR techniques. 

These changes in diffusivity also have been shown to correlate 

with histologic measures of axon loss and behavioral 

deficits. Other experimental data has suggested that DWI 

and DTI can provide measures of axon density, as well as 

other pertinent histologic parameters (5). Fiber tractography 

generated from DTI data also may be helpful in the assessment 

of white matter connectivity in the spinal cord. With 

advanced techniques such as parallel imaging, DTI in the 

human spinal cord has become more feasible and early 

results are promising. 

REFERENCES 

1. Shepard MJ, Bracken MB. Magnetic resonance imaging and 

neurological recovery in acute spinal cord injury: observations 

from the National Acute Spinal Cord Injury Study 3. 

Spinal Cord 1999;37(12):833-837 

2. Lycklama G, Thompson A, Filippi M, et al. Spinal-cord MRI in 

multiple sclerosis. Lancet Neurol 2003;2(9):555-562 

3. Clark CA, Werring DJ. Diffusion tensor imaging in spinal 

cord: methods and applications - a review. NMR Biomed 

2002;15(7-8):578-586 

4. Schwartz ED, Hackney DB. Diffusion-weighted MRI and the 

evaluation of spinal cord axonal integrity following injury 

and treatment. Exp Neurol 2003;184(2):570-589 

5. Schwartz ED, Cooper ET, Fan Y, et al. MRI diffusion coefficients 

in spinal cord correlate with axon morphometry. 

Neuroreport 2005;16(1):73-76 

252 


4:45 PM - 6:15 PM 

Theatre 


Parallel & High Field 

(460) Parallel Imaging Basics 

⎯ Klaas P. Pruessmann, PhD 

(461) Parallel Imaging: Practical Considerations 

⎯ Timothy P.L. Roberts, PhD 

(462) High-Field MR Imaging Advances 

⎯ William G. Bradley, MD, PhD, FACR 


Moderators: Timothy P.L. Roberts, PhD 


Parallel Imaging Basics 

Klaas P. Pruessmann, PhD 

Since 2002 Klaas Pruessmann has been an Assistant 

Professor of Biomedical Engineering at the Swiss Federal 

Institute of Technology (ETH), Zurich, Switzerland. In 2003 he 

also was appointed as an Adjunct Professor at the Department 

of Radiology of the University of Minnesota, Minneapolis. He 

holds a Physics Diploma from the University of Bonn, 

Germany, and a PhD in Physics from the Swiss Federal 

Institute of Technology. His work focuses on parallel MR 

imaging, dynamic MR imaging, image reconstruction, and 

high-field MR technology. Klaas Pruessmann received the 

Gunther Laukien Prize of the Experimental NMR Conference 

in 2001, the Philips Innovation Award in 2001, and the 

European Magnetic Resonance Award in 2004. He serves on 

the executive board of the European Society for Magnetic 

Resonance in Medicine and Biology and on the editorial 

board of Magnetic Resonance in Medicine. 



able to: 

1) Describe the physical foundations of parallel MR imaging. 

2) Describe how parallel MR imaging differs from standard 

Fourier MR imaging. 

3) Name the key benefits of parallel MR imaging. 

4) Appreciate the synergies of parallel MR imaging and high 

baseline field strength.


The goal of this presentation is to explain the basics of parallel 

MR imaging, starting from its physical foundations. Parallel 

MR imaging is based on the fact that the spatial variation of 

receiver coil sensitivity inherently acts as a complementary 

mechanism of spatial signal encoding. The simultaneous 

acquisition of MR imaging signals with multiple receiver coils 

hence permits reducing gradient-driven Fourier encoding. The 

added encoding mechanism entails numerous fundamental 

deviations from standard Fourier MR imaging, which will be 

addressed in detail. Most importantly, parallel MR imaging 

requires suitable receiver hardware, modified strategies for 

data acquisition, and specific algorithms for image reconstruction. 

The latter have major impact on the SNR behavior of parallel 

techniques. Higher encoding efficiency with parallel 

acquisition translates into numerous benefits, such as scan time 

reduction, enhanced resolution and coverage, and artifact 

reduction. These will be illustrated by examples from the realm 

of neuroimaging. Parallel MR imaging is particularly promising 

at high field strength. In fact parallel acquisition and high 

field form a number of substantial synergies, both at the level 

of scanning strategy and in terms of fundamental physics. The 

specific potential of parallel MR imaging at high field will be 

addressed in the last part of the presentation. 

REFERENCES 

1. Sodickson DK, Manning WJ. Simultaneous acquisition of spatial 

harmonics (SMASH): Fast imaging with radiofrequency 

coil arrays. Magn Reson Med 1997;38(4):591-603 

2. Pruessmann KP, Weiger M, Scheidegger MB, Boesiger P. 

SENSE: Sensitivity encoding for fast MRI. Magn Reson Med 

1999;42(5):952-962 

3. Griswold MA, Jakob PM, Heidemann RM, Nittka M, Jellus V, 

Wang JM, Kiefer B, et al. Generalized autocalibrating partially 

parallel acquisitions (GRAPPA). Magn Reson Med 

2002;47(6):1202-1210 

Parallel Imaging: Practical Considerations 


Imaging with multiple RF coil/receivers using the principles 

of sensitivity encoding offers a radical new approach to reducing 

scan times, via the acquisition of fewer k-space lines per 

image acquisition. Under the names, SENSE, ASSET and 

iPAT, reconstruction of the full matrix unfolded image can be 

achieved from the combination of the multiple RF coil 

images. This presentation will focus on practical considerations 

for parallel imaging, addressing not only increased 

patient throughput, but also the advantages of reduced echo 

train length (ETL) in multi-echo sequences, such as FSE and 

EPI. In particular its role in increasing anatomic coverage, 

reducing blurring and magnetic susceptibility artifact will be 

discussed. On the other hand the impact of parallel imaging on 

signal to noise ratio and potential additional artifacts will be 

evaluated. Special attention will be paid to the role of parallel 

imaging at high field (3T), where its additional significance in 

reduction of SAR will be emphasized. 

253 

High-Field MR Imaging Advances 



Thursday

Thursday 

Notes: 

254

Friday Morning 

8:00 AM - 9:43 AM 

Room 105/106 

(75a) SPINE: Trauma, Degenerative 

and Interventional 

(Scientific Papers 463 - 475A) 


(75b) PEDIATRICS: Miscellaneous 

(75c) ADULT and PEDIATRIC BRAIN: Neoplasms 

(75d) ADULT BRAIN: Miscellaneous 

Moderators: Andrew L. Wagner, MD 

Jeffrey G. Jarvik, MD, MPH 


Cement Volumes Used in Vertebroplasty: Contributing 

Factors and Outcomes 

Trout, A. T. 1 · Kallmes, D. F. 2 · Kaufmann, T.J. 2 

1Medical School, Mayo Clinic College of Medicine, 

Rochester, MN, 2Mayo Clinic, Rochester, MN 

PURPOSE 

A review of the vertebroplasty literature reveals that there is 

significant variability in the volume of cement used in the 

treatment of vertebral compression fractures. Moreover, few 

studies have explored the variables that influence the volume 

of cement used in a vertebroplasty session. Thus, there is 

scant data available to guide practitioners with respect to 

appropriate cement volumes. We describe an analysis of 

patient and procedural factors that influence cement volumes 

and the impact of cement volume on vertebroplasty outcomes. 


Retrospective review of 158 patients treated with vertebroplasty 

at a single vertebral level. Comparisons of median 

cement volume across binary variables (sex, cyst activity, 

uni/bipedicular approach, presence of cement extravasation) 

were performed using the Wilcoxon rank-sum test. Linear 

regression was used to define the relationship between age, 

percent compression, outcomes, and injected volume. 

Outcomes measured included pain at rest and pain with 

activity, both using verbal 0-10 scales, and change in med- 

255 





ication use (increased, same, decreased, none). Analysis of 

improvement in outcomes across time was performed using 

either a paired t-test comparison or the Wilcoxon signedrank 

test. 

RESULTS 

One hundred and fifty-eight patients (mean age 73.7 years) 

were treated at vertebral levels between T4 and L5. Fiftyseven 

patients were men (36%) and 110 were women. A unipediculate 

approach was used in 100 patients (63%). Mean 

vertebral compression was 30% +/- 20% and cyst activity 

was described in 53 patients (34%). Extraosseous extravasation 

of cement was observed in 33 patients (21%). Median 

cement volume injected was 3.00 cc (0.5-10.3 cc) for all 

patients. Median cement volumes were 3.00 (0.5-7.0 cc) and 

3.33 cc (1.5-10.3 cc) for uni- and bipedicular approaches 

respectively and 2.5 (0.8-6.1cc) and 3.5 cc (0.5-10.3 cc) for 

thoracic and lumbar vertebrae respectively. Volume of 

cement injected significantly correlated with uni- vs bipedicular 

approach (p = 0.018), with bipedicular injection resulting 

in placement of an additional 0.83 cc of cement. Male 

patients were treated with, on average, 0.96 cc more cement 

than females (p = 0.032). Lumbar vertebrae were treated 

with 1.43 cc more cement than thoracic vertebrae (p < 

.0001). Percent compression and age at time of vertebroplasty 

were found to significantly associate with cement volume 

(p = 0.049 and 0.039 respectively). Cyst activity did not correlate 

significantly with cement volumes but was related significantly 

to percent compression and age at time of vertebroplasty 

(p = .026 and p = .0004 respectively). Specifically, 

cyst activity was more frequent in older patients and in more 

compressed vertebrae. Scores for rest pain and activity pain 

improved significantly by 1 week (p < .0001) and remained 

improved through maximal follow up. Medication use 

declined significantly at all time-points relative to the prior 

time-point (p < .0001). Cement volumes did not correlate 

significantly with improvement in any of the outcome measures 

or with the presence of extravasation of any type. 

CONCLUSION 

Our series demonstrates that small volumes of cement are 

adequate to achieve acceptable pain relief and reduction in 

medication use. Larger volumes were used in male patients, 

bipedicular access cases, older patients, and in mildly compressed 

vertebrae. These data may assist practitioners in predicting 

adequate volumes of cement needed for good outcomes. 

KEY WORDS: Vertebroplasty, volume, outcomes 

Friday

Friday 

256 

Paper 464 Starting at 8:08 AM, Ending at 8:16 AM this debate. Specifically, if vertebroplasty is to be implicated 

Quantification of Cement Filling in Patients Having 

Undergone Percutaneous Vertebroplasty Utilizing 3D 

Volumetric Analysis and its Relation to Refracture 

Rates: A Retrospective Review 

as causative of new adjacent level fractures, then one would 

reason that these fractures should occur earlier than nonadjacent 

fractures which may represent the natural history of the 

disease. 

Ponzo, J. A. 

Scott and White Clinic 

Temple, TX 

PURPOSE 

In vitro studies suggest that greater volumes during cement 

augmentation of the spine predispose to new fractures, 

although this has not been studied in vivo. The purpose of 

this study was to correlate relative injected volume of 

cement and refracture rates in patients who underwent percutaneous 

vertebroplasty procedures for benign compression 

fractures. 


All patients who had a single vertebral level treated over a 

19-month period and at least 1 year of follow up were 

included in the study. Volume analyses of the percent filling 

of a vertebral body treated with injected PMMA (polymethylmethacrylate) 

bone cement using 3D CT volumetric 

measurements of vertebral body and injected cement utilizing 

thresholding techniques were performed. These data then 

were correlated with the presence or absence of a new fracture. 

Twenty-two patients fit these criteria, of which 10 

patients had new fractures and 12 did not. No patient had a 

new fracture if less than 28 percent of the vertebral body was 

filled during vertebroplasty (n = 7). 

RESULTS 

Of patients with greater than 28 percent filling of the vertebral 

bodies, two thirds had new fractures during this same 

time period. 

CONCLUSION 

These data suggest that higher percent cement filling during 

vertebroplasty increases the risk of refracture. 

KEY WORDS: Vertebroplasty, refracture, cement filling 


New Fractures after Vertebroplasty: Adjacent Fractures 

Occur Significantly Sooner 

Trout, A. T. 1 · Kallmes, D. F. 2 · Kaufmann, T. J. 2 



PURPOSE 

Despite greater than 10 years of clinical application of vertebroplasty, 

there remains ongoing uncertainty as to whether 

cement deposition increases the risk of new, adjacent vertebral 

fractures. Both biomechanical and clinical studies suggest 

that vertebroplasty may increase the risk of new fracture, 

beyond the expected natural history incidence. 

However, compelling data have not been put forth yet to settle 

this difficult issue. We hypothesized that study of the time 

course of occurrence of new fractures developing in patients 

previously treated with vertebroplasty might shed light on 


We performed a retrospective review of 57 patients who 

were treated with percutaneous vertebroplasty and subsequently 

presented with additional painful vertebral compression 

fractures. New fractures were diagnosed on the basis of 

MR imaging or bone scan. Survival analysis and the log rank 

statistic were used to compare time to diagnosis of a new 

vertebral fracture between two groups: those with fractures 

adjacent to the vertebroplasty and those not adjacent to the 

vertebroplasty. The null hypothesis was that, following vertebroplasty, 

adjacent level fractures occur at similar timepoints 

as nonadjacent level fractures. 

RESULTS 

One hundred thirty-two new vertebral fractures occurred in 

57 patients (median new fractures = 2, range 1-10). Forty-six 

(34.8%) of 132 fractures occurred in vertebrae adjacent to 

the vertebral body most recently treated by vertebroplasty. 

Median time to diagnosis of a new, adjacent-level fracture 

was 70 days (range 3-1139 days). Median time to diagnosis 

of a new, nonadjacent level fracture was 184 days (range 3- 

1304 days). Time to fracture was significantly different 

between the two groups (Log rank = 0.018) (Fig. 1), and we 

therefore reject the null hypothesis that the time course of 

fractures was similar between groups. 

CONCLUSION 

Our data regarding the frequency of new-onset fractures following 

vertebroplasty are in line with previous reports. 

Notably, our data also show that new-onset fractures of vertebral 

bodies adjacent to those treated by vertebroplasty 

occur significantly earlier than nonadjacent level fractures. 

While this timing phenomenon may simply reflect a “clustering” 

pattern of osteoporotic fractures that is independent

of vertebroplasty, we consider these data provocative. Future 

studies should analyze not only the incidence but also the 

timing of new-onset fractures following vertebroplasty. 

KEY WORDS: Vertebroplasty, new fracture, adjacent 


Effect on Partial Pressure of Oxygen in Arterial Blood in 

Percutaneous Vertebroplasty 

Uemura, A. · Numaguchi, Y. · Kobayashi, N. · Yamashita, S. 

· Matsusako, M. 

St. Luke’s International Hospital 

Tokyo, JAPAN 

PURPOSE 

The purpose of this studyis to evaluate the potential effect of 

percutaneous vertebroplasty (PVP) on pulmonary function 

during PVP. 


We retrospectively reviewed pre and postprocedural partial 

pressure of oxygen (PaO2) of 35 consecutive patients who 

underwent PVP between November 2003 and October 2004 

(8 males and 27 females, age range 50-93 years with a mean 

age of 74.2 years). Thirty-one patients had been treated for 

osteoporosis-related fractures and 4 had malignant disease. 

Percutaneous vertebroplasty was performed in a conventional 

manner using fluoroscopic guidance under local anesthetics 

and conscious sedation. Arterial blood was drawn from 

the radial or femoral artery to measure PaO2 in a supine 

position and on room air. Pre and postprocedural blood 

drawing was performed on the day before PVP and also at 30 

minutes after the injection of bone cement. The difference 

between pre and postprocedural data of PaO2 was correlated 

with patients’ age, number of treated levels, presence of 

cement leakage, and presence of malignancy for each 

patient. 

RESULTS 

Mean pre and postprocedural PaO2 were respectively 79.7 

mmHg and 68.6 mmHg with a mean decrease of 11.1 mmHg 

(95%CI 7.6-14.6). Two patients presented with slight breathing 

difficulty on effort. Patients’ age, presence of cement 

leakage, and presence of malignancy were not correlated 

with the decrease of PaO2. On the other hand, number of 

treated levels was a predictor of more severe decrease in 

PaO2 (p = .04). 

CONCLUSION 

Pulmonary function is impaired during PVP regardless of 

presence of clinical symptoms. Although there is no clear 

explanation of the mechanism, number of treated levels is a 

predictor of more severe decrease of PaO2. 

KEY WORDS: Percutaneous vertebroplasty, pulmonary function 

257 


Minimally Invasive Vertebroplasty: Experience with a 

Coaxial System 

Simko, M. 

Centre d’Imagerie Medicale 

Fribourg, SWITZERLAND 

PURPOSE 

The purpose of the present study is to evaluate the clinical 

interest of a coaxial vertebroplasty system and its advantages 

or eventual risks compared to a conventional vertebroplasty 

system. 


The coaxial vertebroplasty system is designed for a minimally 

invasive positioning of a vertebroplasty needle over a 

guiding needle. The guiding needle allows repeated local 

anesthesia along the needle track, within the bone, and during 

different phases of needle placement. The coaxial system 

was used in 33 patients. 

RESULTS 

The coaxial system allowed safe placement of the vertebroplasty 

needle without any additional complications. It was 

very well tolerated and additional local anesthesia could be 

delivered throughout the procedures, whenever required. 

CONCLUSION 

The coaxial vertebroplasty system helps in performing vertebroplasty 

as a relatively painless, comfortable, and minimally 

invasive procedure. The use of this system could 

reduce the dose of systemic medications and allows realizing 

vertebroplasty even under local anesthesia. It also avoids 

unnecessary tracks of the vertebroplasty canula. 

KEY WORDS: Vertebroplasty 


Technical Aspects of Sacroplasty 

Erdem, E. · Akbas, T. · Beck, J. · Angtuaco, E. J. C. 

University of Arkansas for Medical Sciences 

Little Rock, AR 

PURPOSE 

The same principle used with vertebroplasty can be applied 

to focal lesions and insufficiency fractures involving sacrum 

to alleviate pain, improve the quality of life in affected individuals. 

Due to the complex three-dimensional (3D) anatomy 

of sacrum and its neural foramina, detailed anatomical 

knowledge is needed in performing sacroplasty. We present 

the pertinent anatomical landmarks, needle placement, 

cement deposition techniques, and our clinical experience. 


CT data of 40 human pelvis were studied to define a safe and 

effective needle placement site in sacrum. Five human dry 

sacrum specimens were imaged also under rotational angiography 

and 3D reformats were obtained with markers to define the 

anterior and posterior sacral foramina. The findings were studied 

Friday

Friday 

to identify the best flouroscopic projections during cement deposition. 

Using this technique six patients with painful sacral fractures 

and lesions were treated. 

RESULTS 

CT reconstruction and 3D rotational angiography images 

reveal data that can be used consistently for safe needle 

placement and cement injection. Our clinical experience, 

although limited, show that sacroplasty can be performed 

under biplane flouroscopy. 

CONCLUSION 

Sacral anatomical landmarks are crucial for needle and 

cement placement during sacroplasty. The procedure can be 

done safely under flouroscopic guidance without additional 

CT localization. 

KEY WORDS: Sacroplasty, sacral insufficiency fractures, 

sacral metastasis 


Accuracy of Initial Needle Placement during CT 

Fluoroscopic-Guided Epidural Steroid Injection 

Wagner, A. L. 1,2 

1 Rockingham Memorial Hospital, Harrisonburg, VA, 

2 University of Virginia School of Medicine, Charlottesville, 

VA 

PURPOSE 

To determine the accuracy of initial needle placement during 

CT fluoroscopic-guided epidural steroid injections (ESI) 

using the loss of resistance technique, as well as the reasons 

for initial false-positive needle placement. 


Over a 3-month period, 436 sequential ESIs were evaluated 

prospectively. Needles were placed on or just short of the ligamentum 

flavum using CT fluoroscopic guidance and a loss 

of resistance technique then was used to place the needle into 

the epidural space. Contrast was injected and images 

acquired to determine the exact needle placement. 

RESULTS 

The needle was placed successfully into the epidural space 

on the initial attempt in 428 of 436 cases (98%). There was 

1 intraligament placement, 1 case due to a very thin ligamentum 

flavum, and the other 5 cases were due to the needle 

being placed short of the ligamentum flavum during the 

initial guidance. 

CONCLUSION 

CT fluoroscopy is a highly accurate means of guiding needle 

placement during epidural steroid injections, although there 

are still some false-positive needle placements due to inherent 

inaccuracies of the loss of resistance technique. Ensuring 

that the needle is touching the ligamentum flavum prior to 

beginning the loss of resistance technique will eliminate 

most incidences of initial incorrect placement. 

KEY WORDS: CT fluoroscopy, epidural injections, spine 

258 


Intervertebral Disk Degeneration: Quantitative 

Assessment with T2 Measurements 

Perry, J. · Haughton, V. · Wu, Y. · Anderson, P. · Mistretta, C. 

· Du, J. 

University of Wisconsin Hospitals and Clinics 

Madison, WI 

PURPOSE 

To grade intervertebral disk degeneration, systems have been 

developed based on the MR appearance of the disk. In theory, 

degeneration or regeneration could be monitored more 

accurately with a technique that accurately measured water 

content in the disk. T2 measurements have been proposed for 

this purpose, since average T2 of the disk correlates with 

water content, changes with age and with the presence of 

degenerative changes. However, the problem of averaging 

T2 values over the inhomogeneous intervertebral disk structure 

has not been resolved. We investigated the homogeneity 

of the T2 throughout the intervertebral disk and developed 

an alternative to the average T2 for studying disk degeneration. 


An FSE acquisition was modified to provide T2 calculations 

from the 16 echoes in the echo train. T2 maps were calculated 

automatically by a least squares fitting routine. A sagittal 

7 mm thick FSE slice was obtained in 3 normal volunteers 

and 2 patients with back pain and disk degeneration. All 

were between the ages of 25 and 35. T2 values for each 

voxel in the entire disk were displayed as a contour plot and 

plotted in order of decreasing T2 value (Fig). These plots 

were compared between normal and abnormal disks. 

RESULTS 

The T2 maps of intervertebral disks showed variation due to 

the peripheral annulus fibrosus and intranuclear cleft. The 

plot of T2 values in order of descending values showed a 

roughly linear relationship of T2 values in normal disks 

(Fig). In degenerating intervertebral disks, the T2 values 

were diminished. The water content of the degenerative 

disks, estimated as the sum of the T2 values, was diminished 

markedly with respect to the normal disks. The distribution 

T2 values in the degenerative disks suggested a larger effect 

in regions which normally have higher water content. 

Fig. T2 values in 5 lumbar intervertebral disks plotted in 

order of decreasing value in a subject with diminished disk 

height and signal intensity at L3-4, L4-5, and L5-S1. Note

that the T2 values for L1/2 and L2/3 fit closely to a linear 

trendline. For L3-4, L4/5, and L5/S1, the values are lower 

and conform closely to a linear trendline with a slope that is 

different from the normal disks in this patient. 

CONCLUSION 

Plotting T2 values of the intervertebral disk provides a 

means of estimating water content and theoretically of grading 

intervertebral disk degeneration and measuring degeneration 

or regeneration. 

KEY WORDS: Disk degeneration, MR imaging, T2 relaxation 


Intraosseous and Paraspinal Postlaminectomy 

Pseudomeningocele 

Lis, E. · Krol, G. · Souweidane, M. M. · Bilsky, M. H. · 

Dunkel, I. J. · Cooperman, S. 


New York, NY 

PURPOSE 

To report an unusual case of a expansile, cystic lesion replacing 

T2 vertebra which was determined to be a large 

intraosseus and paraspinal pseudomeningocele. CT, MR, 

myelographic, and intraoperative findings are presented. 


A 13-year-old female was referred to our institution for percutaneous 

biopsy of lytic lesion of T2 vertebra, suspicious 

for neoplasm. At the age of two she underwent partial 

laminectomy at T1 and T2 levels for intraspinal shunting of 

a lower cervical/upper thoracic arachnoid cyst. She did well 

until she was 11 years old when she began to experience 

intermittent headaches. An MR image of the brain at that 

time was reportedly normal. The headaches continued and 

she also developed some neck pain while looking down. An 

MR image of the spine was obtained, revealing an expansile, 

lytic, multiseptated, minimally enhancing process involving 

entire T2 vertebral body and posterior elements with an associated 

paraspinal component. After the biopsy needle was 

advanced into expanded left transverse process, predominately 

clear fluid return was obtained, which was observed 

to have respiratory oscillations. About 5 ml omnipaque 180 

was hand injected via biopsy needle, revealing cystic, multicompartmental 

structure occupying entire T2 vertebra, communicating 

with posterior paraspinal collection and the thecal 

sac. 

RESULTS 

At the time of surgery a dural tear was identified and 

repaired. 

CONCLUSION 

A unusual case of an expansile benign lesion of upper thoracic 

vertebra, mimicking neoplasm is reported. Diagnosed 

by “biopsy-myelography,” it was confirmed surgically to 

represent an intraosseus pseudomeningocele, likely due to a 

dural tear related to prior spinal surgery. 

KEY WORDS: Pseudomeningocele, cystic lesion 

259 


Imaging Features with Histopathologic Correlation of a 

Spinal Angiolipoma 

Pace, M. T. · Jain, R. 


Detroit, MI 

PURPOSE 

Extradural angiolipoma is a rare tumor of the spine that consists 

primarily of adipose and vascular elements. 

Angiolipomas usually are found in the soft tissues of the 

neck, trunk, or extremities. Less than 100 cases of this spinal 

tumor have been reported in the literature until now. Most of 

these occur in the thoracic level. The MR findings and pathologic 

features of this tumor will be presented. 


A 58-year-old female presented with mild back pain. MR 

imaging showed an extradural mass with extension into the 

paravertebral space as described below. Neurologically the 

patient was found to have brisk lower extremity reflexes 

with no motor or sensory deficits. The patient underwent surgical 

resection of the mass. Histopathology revealed the 

mass to be a soft tissue proliferation composed of large vessels 

with thin muscular walls, capillary-type vessels with 

focal proliferation of perivascular smooth muscle cells, and 

a component of adipose tissue. All these pathologic findings 

are consistent with an angiolipoma. 

RESULTS 

Plain films demonstrate a left paravertebral soft tissue mass 

at T6-7. MR imaging of the thoracic spine revealed a 4 cm 

paravertebral mass abutting the T6 vertebral body on the left. 

It extends through the neural foramen at T6-7 and continues 

into the extradural space which compresses and displaces the 

spinal cord. There is no cord edema. The mass is mostly 

hypointense on T1 and is mostly hyperintense on T2. In retrospect, 

there is fat signal on T1 and T2 at the periphery of 

the mass. After gadolinium administration, the mass nearly 

strongly enhances. 

Friday

Friday 

CONCLUSION 

Spinal angiolipomas are rare benign tumors of the axial 

skeleton consisting of fat and vascular elements. The primary 

differential diagnosis is a nerve sheath tumor among 

others. However, the presence of fat signal on MR imaging 

or CT should suggest the diagnosis of angiolipoma. 

KEY WORDS: Angiolipoma, spine, lipoma 


CT and MR Diagnosis of Bronchopleural-Subarachnoid 

Fistula with Pneumorrhachis and Pneumocephalus after 

Lung Lobectomy: Case Description and Literature 

Review 

Stecco, A. · Leone, M. · Ferrari, A. · Puppi, A. M. · Sacco, M. 

· Monaco, F. · Carriero, A. 

AO Maggiore Carita’-Università Piemonte Orientale - 

Novara 

Novara, ITALY 

PURPOSE 

Describe and review the literature of an unusual case concerning 

a patient with lung cancer who developed a postsurgical 

bronchopleural-subarachnoid fistula, pneumorrhachis, 

and pneumocephalus. 


We report a case of a 70-year-old patient affected by adenocarcinoma 

of right upper lobe lung carcinoma, treated with 

upper right lobectomy. A month after surgery, patient presented 

daily frontal bilateral headache; two months later he 

was diagnosed and treated for a pleural empyema. Later on 

he showed confusion, lethargy, aphasia and right hemiparesis. 

Brain CT showed a massive pneumocephalus and head 

and spine CT and MR evaluation were performed to find the 

site of fistula. Patient died after 3 days. 

260 

RESULTS 

Brain CT scan showed pneumatic distension of cerebral ventricles 

of the brain with fluid-to-air levels, and air in subarachnoid 

space and basal cisterns (Fig. 1 A). To exclude the 

presence of spontaneous cerebro-spinal fistula a spiral acquisition 

was conducted through the skull base, without finding 

any skull bony defects. Spinal MR imaging was performed 

on a 0.5 T magnet, with FSE T2-weighted and GRE* T2weighted 

multiplanar pulse sequences. Coronal T2-weighted 

imaging of the spine showed a hyper-intense fluid collection 

in the right upper paramediastinal area, to be referred to fluid 

liquor-like collection (Fig. 1 B). Axial GRE* T2-weighted 

scans revealed an hypo-intense signal alteration across the 

right D2-D3 neural foramen. Another small finding with 

similar features was reported in the anterior subarachnoid 

space, at D1 level. Sagittal T2-weighted MR imaging 

showed, an air-to-gas level at the same level (D1). Spiral CT 

at C7-D4 level allowed to recognize small bubbles of gas at 

D1-D3 levels, though not revealing exactly the fistula path. 

CONCLUSION 

Pneumorrhachis occurs following trauma, barotraumas, or 

invasive procedures; air may enter the epidural space or may 

penetrate further into the subarachnoid space. These conditions 

are usually self-limited, without symptoms. When pulmonary 

air enters the subarachnoid space, it easily travels 

cephalad to the head (pneumocephalus). These patients 

experience severe headaches and neurologic stroke-like 

symptoms. Pneumorrhachis usually is demonstrated best on 

CT scanning but may be seen on MR scan too. It is mandatory 

to search for the exact location of the fistula, to attempt 

a surgical repair. In this case the sensitivity of GRE pulses to 

the presence of air made the MR more affordable than TC in 

finding the exact “entrance door” of air, and CSF external 

leakage, and this aspect never has been stressed in literature. 

KEY WORDS: Fistula, pneumorrhachis, pneumocephalus

261 

Paper 475A Starting at 9:35 AM, Ending at 9:43 AM rior, and 13% holo-vertebral fractures, respectively, p < 

Vertebral End Plate Fractures: An Indicator of the 

Abnormal Forces Generated in the Spine following 

Vertebroplasty 

.0001). Incident fractures immediately below a treated level 

showed a disproportionate number of superior end plate fractures 

when compared to prevalent fractures (84% superior, 

12% inferior, and 4% holo-vertebral fractures, p = 0.002). 

Trout, A. T. 1 · Kallmes, D. F. 2 · Thielen, K. R. 2 · Layton, K. F. 2 



PURPOSE 

Biomechanical simulations indicate that, following vertebroplasty, 

increased stress is localized to the end plates of adjacent 

vertebral bodies (1, 2). Among spontaneous osteoporotic 

fractures, isolated superior end plate fractures are 

substantially more common than isolated inferior end plate 

fractures. We hypothesized that, if new-onset fractures of 

vertebral bodies immediately above a cemented vertebrae 

localized disproportionately to the inferior end plate, one 

might implicate the cement as causative of this unusual fracture 

pattern. 

METHODS & MATERIALS 

We performed a retrospective review of 86 patients with 

documented new vertebral fractures following vertebroplasty 

for osteoporosis induced fractures. Fractures were diagnosed 

on the basis of MR imaging or bone scan. We define 

“prevalent” fractures as those present prior to vertebroplasty, 

and “incident” fractures as new-onset fractures following 

vertebroplasty. We defined fracture location as “end plate” 

(superior or inferior) if the marrow along the opposite end 

plate was normal on MR imaging, if a distinct fracture line 

could be identified on MR imaging, or if plain films showed 

a clear end plate deformity. Fractures not meeting these criteria 

were considered holo-vertebral. Chi squared analysis 

was used to compare the observed proportion of inferior end 

plate fractures in both prevalent and incident fractures. Our 

null hypothesis was that incident fractures above a treated 

level would demonstrate the same proportion of inferior end 

plate fractures as other groups, including 1) prevalent fractures, 

2) incident fractures immediately below a treated 

level, and 3) incident fractures not adjacent to treated levels. 

RESULTS 

The 86 patients selected for analysis presented with 313 vertebral 

fractures and were treated at a total of 137 vertebral 

levels. Following vertebroplasty, 186 new fractures developed 

(median number of new fractures = 2, range 1-10). MR 

images were available for 274 of the 313 prevalent fractures 

and 128 of the 186 incident fractures. In prevalent (baseline) 

fractures, superior end plate fractures were significantly 

more common than either inferior end plate or holo-vertebral 

fractures (57% superior end plate, 11% inferior end plate, 

and 32% holo-vertebral, p < .0001). Of the 186 incident 

(new-onset) fractures, 77 (41%) were adjacent to treated vertebrae. 

Nonadjacent, incident fractures showed a fracture 

distribution that was different from prevalent fractures only 

in terms of the proportion of holo-vertebral fractures; superior 

end plate fractures still predominated over inferior end 

plate fractures (69%, 25%, 6% for superior end plate, inferior 

end plate, and holo-vertebral fractures, respectively, p < 

.0001). Incident fractures immediately above treated levels 

showed a disproportionate number of inferior end plate fractures 

compared to all other groups (30% superior, 57% infe- 

CONCLUSION 

The typical location for vertebral compression fractures in 

osteoporotic patients is along the superior end plate. As predicted 

by biomechanical simulations, vertebroplasty disrupts 

this typical fracture distribution through its effect on adjacent 

vertebral bodies. New-onset fractures that occur immediately 

above a cemented vertebral body demonstrate a striking 

localization to the inferior end plate. This pattern was not 

seen in prevalent fractures, in non-adjacent incident fractures, 

or in incident fractures immediately below treated levels. 

In addition, new-onset fractures immediately below a 

treated level demonstrate a predominance of superior end 

plate fractures when compared to prevalent fractures. These 

data add additional circumstantial evidence that vertebroplasty 

may play a causative role in new fractures. 

REFERENCES 

1. Baroud G, Heini P, Nemes J, et al. Biomechanical explanation 

of adjacent fractures following vertebroplasty. Radiology 

2003;229(2):606-607; author reply 607-608 

2. Baroud G, Nemes J, Heini P, et al., Load shift of the intervertebral 

disc after a vertebroplasty: a finite-element study. 

Eur Spine J 2003;12(4):421-426 

KEY WORDS: Vertebroplasty, spine, interventional 

This paper was selected to receive the 2005 Mentor Award 

by the American Society of Spine Radiology (ASSR) at its 

Annual Symposium held February 24-27, 2005. 

Friday

Friday 


8:00 AM - 9:30 AM 

Theatre 




(75a) SPINE: Trauma, Degenerative and 

Interventional 



Moderators: Caroline D. Robson, MB, ChB 

Alena Horska, PhD 


Quantitative Evaluation of Brain Involvement in Ataxia 

Telangiectasia by Diffusion-Weighted MR Imaging 

Firat, A. K. · Karakas, H. M. · Firat, Z. Y. · Yakinci, C. · 

Alkan, A. · Altinok, M. T. 

Inönü University, Turgut Ozal Medical Center 

Malatya, TURKEY 

PURPOSE 

Spatiotemporal sequence of the pathologic brain changes in 

ataxia telangiectasia is controversial. Microstructural 

changes of the brain parenchyma were evaluated with diffusion-weighted 

imaging (DWI) in order to answer the question 

of whether the disease begins diffusely or selectively. 


Six patients with ataxia-telangiectasia (9-13 years) were 

examined on 1.5 T scanner. In addition to conventional 

images, DWI were performed using B values of 0, 500, and 

1000 s/mm2. Trace ADC values were measured from different 

supra and infratentorial locations. Nine age-matched 

healthy children also were included as controls. 

RESULTS 

On conventional images the only abnormal radiologic finding 

was cerebellar atrophy in all patients. T2 signal pattern 

was normal in all patients and controls. For cerebral structures 

mean ADC values of patients and controls were similar. 

The cerebellum had abnormal ADC values in patients, 

contrary to cerebrum. Apparent diffusion coefficient values 

of the cerebellar white matter (p < 0.011) and superior, lateral, 

and inferior cortex (p < 0.001-0.0001) were significantly 

higher in patients than in controls. The highest significance 

was produced by ADC values of middle cerebellar cortex. 

Patients and controls were classified with 100% accuracy 

using ADC values of cerebellar white matter and cortex 

262 

(logistic regression, p < 0.013). The cut-off ADC value of 

0.694 mm2/sn for middle cerebellar cortex produced 100% 

sensitivity and specificity. 

CONCLUSION 

In ataxia telangiectasia a selective rather than the diffuse 

involvement is favored. The diagnostic value of DWI in 

detecting the disease by showing structural disintegrity of 

cerebellar tissue is established over conventional techniques 

that mainly depend on morphologic changes. 

KEY WORDS: Ataxia telangiectasia, MR diffusion-weighted 

imaging, apparent diffusion coefficient quantification 


Diffusion Tractography and Functional MR Imaging for 

Preoperative Planning in Children 

Jones, B. V. · Crone, K. · Myseros, J. · Dardzinski, B. 

Children’s Hospital Medical Center 


PURPOSE 

We present our experience in the use of functional MR imaging 

(fMRI) and diffusion tensor imaging (DTI) techniques in 

the preoperative evaluation of 8 children with focal brain 

lesions. 


All fMRI and DTI studies were performed on a Siemens 3 T 

Trio MR scanner. Right- and left-hand finger-tapping paradigm 

was performed in all cases, with single-shot echo-planar 

BOLD imaging. Diffusion tensor imaging was acquired 

in six directions. All data were analyzed on a Siemens 

Leonardo postprocessing workstation. Diffusion anisotropy 

maps and tractography were created from the DTI data using 

software developed at Massachusetts General Hospital. 

RESULTS 

In each case, confident identification of the motor cortex was 

achieved on fMRI. Tractography then was performed on 

regions of interest correlating to the motor cortex as identified 

by fMRI. Additional tractography was performed typically 

on relevant ipsilateral structures such as the internal 

capsule and cerebral peduncle. Findings were reviewed with 

the neurosurgical team to identify an ideal angle of approach 

for surgery. Five of the eight cases had surgery shortly after

the studies, and decisions regarding surgical approach in all 

five cases were influenced by the tractography findings. In 

the three cases that did not proceed to surgery, the tractography 

findings did not influence the decision not to operate. 

One surgery was performed using specialized minimal 

access resection technique (SMART), in which a surgical 

tract was created by placing a balloon catheter through a 

small burr hole and slowly inflating the balloon over several 

days. Gross total lesion resection was achieved in four of the 

five operated cases, with the fifth surgery performed as a 

biopsy only. No new postoperative neurologic deficits developed 

in any of the operated cases. 

CONCLUSION 

Diffusion tractography can be combined with fMRI to provide 

a clinically useful guide for surgical approaches to brain 

lesions in children. Findings at tractography may result in an 

alteration of surgical approach to minimize postoperative 

deficits, and can be used to guide innovative surgical techniques 

such as SMART. Our preliminary experience shows 

the potential of these techniques in children, although further 

study is required to assess their accuracy. 

KEY WORDS: Tractography, functional MR imaging, tumor 


Diffusion Tensor Imaging and Fiber Tracking in 

Neurologically Devastated Children with Callosal 

Agenesis 

Rollins, N. 1,2 · Booth, T. 1,2 · Veltkamp, D. 1,2 · Reyes, T. 1 

1 Children’s Medical Center, Dallas, TX, 2 University of 

Texas, Southwestern Medical Center, Dallas, TX 

PURPOSE 

Diffusion tensor imaging (DTI) and fiber tracking (FT) provide 

a novel way to evaluate white matter. White matter 

tracts of the limbic system, which is important in declarative 

memory and learning and known to be abnormal in many 

cerebral malformations, are readily apparent on DTI/FT. We 

evaluated white matter tracts of the limbic system using 

DTI/FT in 9 patients with callosal agenesis and reviewed 

medical records for neurologic status. 


The investigation was HIPPA compliant. Routine MR imaging 

(1.5 T, Philips Medical Systems, Best, the Netherlands) 

was followed by DTI [SENSE = 2.5, EPI = 35, b = 700, 

7357/98 (TR/TE), scan matrix 112 x 256, 246 x 246 mm 

FOV, 55 slices, 2.2 mm slice thickness, no gap, 15-32 directions, 

7-13 min acquisition times]. The raw data was transferred 

to an offline PC having PRIDE software V4.1V3. 

Automated motion correction was performed; corrected data 

sets were averaged together and saved as a new raw data file. 

Fiber tracking was generated using the FACT algorithm, 

manually placing ROIs over the posterior frontal, parietal, 

and temporal white matter on the color maps and using a 

“brute force” approach. Tract propagation was terminated at 

FA value = 0.15, internal angle = 0.75. Images derived from 

tractography were assessed qualitatively with respect to the 

presence and continuity of the cingulum, forniceal columns, 

crura, and fimbria. The appearance of the fornices and cingulum 

were compared to age-matched normal patients. 

263 

RESULTS 

In term infants, the fimbria, crura, columns, and precommissural 

and postcommissural portions of the fornix are visible by 

FT as continuous structures separate from the cingulum; the 

columns appear fused. The frontoparietal segments of the cingulum 

are seen at birth while the preseptal and temporal portions 

of the cingulum are visible by FT by 3 months of age. 

There were 9 patients, ages 3 day-14 years, 5 males/6 females 

with callosal agenesis. Six patients had total callosal agenesis; 

3 had rudimentary callosal fibers. No patient had hydrocephalus; 

5 had colpocephaly. As seen by FT, the columns of 

the fornices were similar in appearance to age-matched controls 

in 9/9 patients but were widely separated in 8/9. The fornices 

were completely formed and symmetric in 5 patients; in 

4 patients 1 or both fornices were formed incompletely. The 

crura and temporal portions of the fornices were underdeveloped 

or absent in 6/9. The cingulum were completely formed 

in 3 patients and absent in 1. In 5 patients the frontoparietal 

segments were present while temporal portions were absent. 

In 8/9 patients the cingulum were thickened and continuous 

with Probst’s bundles. In 6 patients, the forniceal crura were 

inseparable from the cingulum. Eight of 9 patients had severe 

seizures; 7/9 had pervasive global developmental delay while 

2 were too young to assess. 

CONCLUSION 

Patients with callosal agenesis who are devastated neurologically 

have striking abnormalities of the white matter of the 

inner and outer arches of the limbic system by DTI/FT. In 

this group of patients, abnormalities of the limbic arches 

could not be predicted on the basis of findings by routine 

MR imaging. 

KEY WORDS: Agenesis of the corpus callosum, fiber tracking, 

pediatrics 


Unsedated Pediatric MR Imaging: A Retrospective 

Review of Clinical Use of a Multiplanar Single-Shot Fast 

Spin-Echo T2 (Quick Brain) Protocol for the Evaluation 

of Ventriculomegaly 

Hayek, R. A. · Mamourian, A. C. · McIntyre, J. J. · Quebada, 

P. B. · Duhaime, A. 

Dartmouth-Hitchcock Medical Center 

Lebanon, NH 

PURPOSE 

To assess image quality, interobserver agreement, and technical 

success of our "quick brain" protocol in 44 unsedated pediatric 

patients undergoing evaluation for ventriculomegaly. 


We performed a retrospective review of 54 random MR scans 

over an 18-month time period in 44 patients (aged 3 months to 

18 years) with possible ventriculomegaly. Of the 54 exams 

reviewed, 44 were initial exams and 10 were follow-up studies. 

Imaging was performed on a 1.5 T MR scanner using single-shot 

fast spin-echo T2-weighted images (TR/TE - 966/90) 

in three planes. Acquisition time was less than 20 seconds per 

plane, with total scan time of approximately 3 minutes. 

Patients frequently were accompanied in the scanner bore by 

a caregiver. No patients received sedation for imaging. All 

images were interpreted retrospectively and independently on 

Friday

Friday 

a PACS workstation by two staff neuroradiologists and one 

neuroradiology fellow. The readers were blinded to the final 

diagnosis. Initial exams were evaluated for image quality 

(adequate or not) and for the presence or absence of ventriculomegaly. 

If present, ventriculomegaly was graded as mild, 

moderate, or severe. Follow-up exams were graded for image 

quality and for interval change (none, worse, or better). 

Objective measurements of technical success were determined 

by electronic chart review of the subjects' imaging history. 

Technical success was determined by: 1) the number of successfully 

completed scans and 2) the number of repeated 

sequences obtained to achieve a complete study. As part of the 

chart review, note was made of presence of additional incidental 

findings on initial imaging interpretation. 

RESULTS 

All 54 exams were rated adequate for evaluation of ventricular 

size by all 3 readers. For the initial exams, interobserver 

agreement among all 3 readers for the presence or absence 

of ventriculomegaly (N = 44) was 84% (38/44). All discrepancies 

were between no enlargement and mild enlargement. 

For patients in whom ventriculomegaly was determined 

present by all three radiologists (N = 38) interobserver agreement 

for degree of severity was 92% (35/38). One hundred 

percent interobserver agreement was present between at least 

two of the three radiologists. Of the three discrepancies: 2 

cases were graded mild by one radiologist and moderate by 

two radiologists; 1 case was graded severe by one radiologist 

and moderate by two radiologists. For follow-up exams (N = 

10), 100% agreement was present between all three radiologists. 

No patients failed unsedated imaging and no patients 

required additional imaging. Sequence repeat rate for excessive 

motion was low at 7% (12/174). 

CONCLUSION 

This study revealed 100% success and high interobserver agreement 

using multiplanar single-shot FSE T2-weighted images 

for assessment of ventricular size and change in ventricular size 

in unsedated pediatric patients. Because of the inherent advantages 

compared with CT or conventional MR, (i.e., no radiation 

or sedation), the quick brain protocol has become the primary 

method for imaging suspected ventriculomegaly and shunt failure 

in the pediatric age group at our institution. 

KEY WORDS: Ventricular size, unsedated MR imaging, 

quick brain 


Improved Measurements of Cervical Cerebrospinal 

Fluid Flow Using Fully Balanced Steady-State Free- 

Precession Phase-Contrast MR Imaging 

Wagshul, M. E. · Egnor, M. R. · McCormack, E. J. · Roche, 

P. E. 

Stony Brook University 

Stony Brook, NY 

PURPOSE 

Measurements of cerebrospinal fluid (CSF) flow with MR 

imaging have improved understanding of intracranial disorders. 

Typically, velocity imaging of CSF flow utilizes phasecontrast 

(PC) MR imaging, but signal from CSF acquired with 

this technique varies throughout the cardiac cycle and produces 

low signal intensity. A new type of velocity imaging with fully 

264 

balanced steady-state free-precession (b-SSFP) generates high 

signal from flowing spins (1). The techniques presented by 

Markl, et al. (2) for combining PC MR imaging with b-SSFP 

(PC SSFP) are implemented in this study for improving the 

speed and accuracy of CSF flow measurements. 


Seven volunteers were imaged on a Philips Intera 1.5 T to compare 

PC SSFP (TE/TR 6.5 = 13, FOV = 15-18 cm, 18 cardiac 

phases, 1NEX) with PC MR imaging (TE/TR = 8/14, FOV = 

15-18 cm, 18 cardiac phases, 2NEX). Both sequences were 

acquired in the transverse orientation to visualize flow through 

the aqueduct of Sylvius and at the CC junction. Velocities were 

measured with inverted gradient switching in the slice-select 

direction. Signal variation throughout the cardiac cycle, signalto-noise 

ratio (SNR), and accuracy of velocity waveforms were 

analyzed for both imaging methods. 

RESULTS 

Phase-contrast SSFP greatly enhanced the signal from CSF, 

even though the flow is pulsatile. Images at the CC junction 

demonstrated high, uniform signal easily distinguished from 

the surrounding tissue, whereas PC MR images showed varied 

signal throughout the cardiac cycle and were isointense 

or dark compared to surrounding tissue. Phase-contrast 

SSFP images yielded CSF signal which were distinguished 

easily from surrounding tissue and maintained throughout 

the cardiac cycle, yielding a higher, more uniform SNR than 

the PC MR images. In addition, PC MR imaging often 

showed decreased signal in the posterior SAS, most likely 

due to turbulence caused by the confined space and the confluence 

of subarachnoid and fourth ventricle outflow, but PC 

SSFP yielded high signal for the entire SAS. This enabled 

accurate total flow calculations, which were not always possible 

with PC MR imaging. At an optimum flip angle of 50 

degrees, high SNR was obtained with minimal signal variation 

as a function of the cardiac cycle. Total CSF signal for 

PC SSFP averaged three times higher. On average, SNR 

increased 4.2 times over PC MR imaging. Comparable SNR 

gain was observed for all volunteers. 

CONCLUSION 

With enhanced flow region depiction throughout the cardiac 

cycle and reduced scan time, PC SSFP can allow more efficient 

scanning and more accurate measurements, since net 

flow waveforms at CC include the entire SAS. With PC 

SSFP acquisition times of less than 1 minute, the addition of 

this sequence to routine clinical brain imaging protocols is 

feasible. Measuring intracranial flow for evaluating intracranial 

disorders with PC SSFP may offer better diagnostic 

capabilities due to improved efficiency and flow region 

delineation for quantification. 

REFERENCES 

1. Overall WR, et al. Fast phase-contrast velocity measurement 

in the steady state. Magn Reson Med 2002;48:890-898 

2. Markl M, et al. Balanced phase-contrast steady-state free precession 

(PC-SSFP): a novel technique for velocity encoding 

by gradient inversion. Magn Reson Med 2003;49:945-952 

KEY WORDS: Brain 

Acknowledgment: We would like to thank Philips Medical 

Systems for assistance with sequence development and the 

Brainchild and Dana Foundations for financial support.


Optimizing Phase-Contrast MR Imaging for 

Cerebrospinal Fluid Flow Analyses 

Wu, Y. · Fain, S. · Wentland, A. · Korosec, F. · Vigen, K. · 

Haughton, V. 

University of Wisconsin Hospitals and Clinics 

Madison, WI 

PURPOSE 

While phase-contrast MR imaging has been used to characterize 

CSF flow in the foramen magnum in patients with a 

Chiari I malformation, studies at multiple levels with the 

neck flexed, extended, and in neutral position have not been 

possible because of excessive acquisition time. 

Hypothetically, an undersampled radial projection MR 

method, PIPR, applied to the imaging of CSF flow improves 

acquisition time sufficiently to permit multiple slices and 

multiple patient positions. The purpose of this study was to 

assess image quality and accuracy of the PIPR method for 

CSF flow imaging. 


A version of PIPR was modified for CSF flow measurement 

and tested. Data were acquired in a phantom with PIPR and 

phase-contrast MR imaging. Velocities calculated by PIPR 

were plotted against velocities measured by phase-contrast 

MR imaging and regression coefficient and standard error 

calculated. Images were acquired in a normal human subject 

with the PIPR and conventional phase-contrast MR techniques 

and the quality of the images compared by inspection. 

RESULTS 

With the modified PIPR, 20 flow images were acquired 

through the cardiac cycle in 29 seconds, compared to 14 

images in about 120 seconds with a fast phase-contrast MR 

imaging method. In the phantom, velocities calculated from 

PIPR images correlated within a few percent of the velocities 

calculated from the phase-contrast MR data. In the 

human subject, the PIPR images demonstrated CSF flow 

effectively, with little interference from the expected artifacts 

due to undersampling. 

Fig.1. Single CSF flow image from a PIPR acquisition (A) 

and a phase-contrast MR acquisition (B) in the volunteer. 

Note the artifacts in A due to undersampling, which do not 

obscure CSF flow information. The PIPR acquisition 

required 29 sec and the phase-contrast MR imaging 2 minutes. 

265 

CONCLUSION 

PIPR, an accurate and faster imaging technique than phasecontrast 

MR imaging, facilitates the measurement of CSF 

flow. 

KEY WORDS: CSF flow, Chiari I malformation, PIPR 


Stroke Volume Ratio and Aqueductal Phase Shifts as 

Measures of Cerebrospinal Fluid Flow Dysfunction in 

Hydrocephalus 

Egnor, M. R. · McCormack, E. J. · Roche, P. E. · Wagshul, 

M. E. 

Stony Brook University 

Stony Brook, NY 

PURPOSE 

Hydrocephalus is often characterized by increased cerebrospinal 

fluid (CSF) pulsatility in the cerebral aqueduct (1). 

However, the question of the most valuable flow measure 

indicative of intracranial flow disorder still lingers. The purpose 

of this study was to investigate two measures of intracranial 

flow and their relationship to a model of hydrocephalus 

developed recently by Egnor, et al. (2). The measures were: 

1) the ratio of pulsatile stroke volume between the ventricular 

and subarachnoid spaces as measured in the aqueduct and 

at the craniocervical junction (CC) and 2) the phase of CSF 

flow in these two regions relative to arterial flow. 


Fifteen healthy volunteers were studied on a 1.5 T Philips 

Edge scanner. Phase-contrast MR images were acquired perpendicular 

to the aqueduct and to the cervical subarachnoid 

space. Both sequences used the following parameters: craniocaudal 

velocity encoding, Venc 4-6 cm/s, peripheral gating, 

FOV 14-16 cm (aqueduct) or 22 cm (CC), 256 x 192, 2 

NEX. Intracranial arterial flow was measured using a Venc 

of 60 cm/s. Net stroke volume was extracted as the integral 

of the flow waveform times pixel area for all positive flow 

values. Flow phase was measured by fitting the flow waveform 

to a three harmonic sinusoidal function. The phase of 

the fundamental component was extracted and compared to 

the average carotid phase. 

RESULTS 

Valid flow data were obtained in 13 subjects. Results for the 

group were: mean stroke volume ratio: 5.07 ± 0.49%, mean 

aqueductal-to-arterial phase: -52.3 ± 4.2°, mean CC-to-arterial 

phase: 6.5 ± 3.6°. The stroke volume ratio showed a smaller 

variation than aqueductal stroke volume alone, although 

with a small sample size this difference may be insignificant. 

The phase data demonstrated a clear lag of aqueductal flow 

relative to the arterial input, and at CC the data were consistent 

with synchrony between the CSF and arterial flows. 

CONCLUSION 

We have presented an alternative to the standard measurement 

of flow in the aqueduct, based on the assumption that 

flow dysfunction may be related not to absolute hyperdynamic 

flow, but to the redistribution of flow between the 

ventricular and subarachnoid spaces. The results demonstrate 

the expected range for this measure in healthy individuals. 

Phase measurements show that flow in the aqueduct 

Friday

Friday 

lags the arterial input, representing a measure of the local 

compliance or inertia of the intracranial system, which may 

change in hydrocephalus patients. At CC, however, the flow 

is synchronous with arterial flow indicating a balance 

between compliant and inertial forces. The next important 

step will be to apply these methods in a clinical population 

and to correlate changes in stroke volume ratio and phase 

with changes in brain compliance, and more importantly, 

with clinical measures of patient prognosis, such as ventriculomegaly 

and shunt malfunction. 

REFERENCES 

1. Greitz D. CSF circulation and associated intracranial 

dynamics: A radiologic investigation using MRI and 

radionuclide cisternography. Acta Radiol Suppl 1993;386:1- 

23 

2. Egnor M, et al. A model of pulsations in communicating 

hydrocephalus. Ped Neurosurg 2002; 36:281-303 

KEY WORDS: Brain, CSF flow, stroke volume 

Acknowledgments: We thank the Dana and Brainchild 

Foundations for financial support. 


Nonhuman Immunodeficiency Virus Cerebrospinal 

Dolichoectasia in Two Children 

Oka, M. · Gailloud, P. · Jordan, L. 


Baltimore, MD 

PURPOSE 

We report angiographic observations of two children with 

cerebral dolichoectasia, one of whom also showed a similar 

anomaly in the anterior spinal axis. 


Case 1: A 16-year-old girl consulted for headache. MR imaging 

showed abnormal flow voids in the left sylvian fissure 

and prominent perispinal vessels at the lumbar level. 

Cerebral angiography revealed marked fusiform enlargement 

of the right posterior cerebral artery (P3-P4 segments) 

and of the left middle cerebral artery (M1-M2-M3 segments) 

(Fig. 1). Both lesions provided normal distal branches. 

Spinal angiography demonstrated an enlarged anterior spinal 

axis without associated vascular malformation. Case 2: A 2year-old 

girl was evaluated for dizzy spells. MR imaging 

demonstrated prominent flow voids in the right sylvian fissure. 

Cerebral angiography demonstrated fusiform enlargement 

of the right MCA involving M2-M3 segments, which 

provided normal distal branches. 

RESULTS 

The etiology of dolichoectasia is poorly understood. It appears 

that prevalent theories of the development of cerebral 

dolichoectasia favor atheromatous disease in the vertebrobasilar 

circulation of older patients, and arterial dissection in the 

anterior circulation of young patients. In addition, children 

affected by human immunodeficiency virus (HIV) are known 

to have a cerebral vasculopathy with cerebral segmental ectasia 

or fusiform aneurysms. Our two patients were children 

without acute symptoms and showed no evidence of atheromatous 

disease, hypertension, history of trauma, connective or 

266 

autoimmune disorders. The absence of thrombus or wall 

thickening on MR imaging goes further against the aforementioned 

theories. One of our patients showed involvement of 

three distant vessels (MCA, PCA, anterior spinal artery), a 

finding also speaking in favor of a developmental mechanism. 

CONCLUSION 

Our cases seem to suggest an underlying developmental 

mechanism. It is possible that findings here are the earliest 

phase of the process, and that wall thickening, intramural 

thrombus, and dissections represent later stages of evolution. 

REFERENCES 

1. Day AL, Gaposchkin CG, Yu CJ, Rivet DJ, Dacey RG, Jr. 

Spontaneous fusiform middle cerebral artery aneurysms: 

characteristics and a proposed mechanism of formation. J 

Neurosurg 2003;99:228-240 

2. Doran SE, Deveikis JP, Chandler WF. Dolichoectasia of the 

anterior cerebral arteries in an adolescent. AJNR Am J 

Neuroradiol 1995;16:1548-1550 

3. Mizutani T, Miki Y, Kojima H, Suzuki H. Proposed classification 

of nonatherosclerotic cerebral fusiform and dissecting 

aneurysms. Neurosurgery 1999;45:253-259; discussion 259- 

260 

4. Nakatomi H, Segawa H, Kurata A, et al. Clinicopathological 

study of intracranial fusiform and dolichoectatic aneurysms: 

insight on the mechanism of growth. Stroke 2000;31:896-900 

KEY WORDS: dolichoectasia 


Arterial Vascular Abnormalities Accompanying 

Cerebral Cortical Dysplasia 

Saatci, I. · Tasbas, B. · Ozturk, H. · Ozturk, A. · Cekirge, H. S. 



PURPOSE 

To present the rare association of arterial abnormalities with 

neuronal migration disorder. 


First patient is a 22-month-old boy presented with left-sided 

proptosis. The eye findings have been present since birth. He 

has normal mental and motor development. He does not have

seizures. No stigmata of any neuroectodermal disorder is 

noted. The other patient is a 50-year-old female with episodes 

of left-sided numbness for last 4 years, increased lately. 

RESULTS 

The MR imaging of the first patient revealed an orbital mass, 

left frontal cortical dysplasia associated with developmental 

venous anomaly, another developmental venous anomaly 

(DVA) in the ipsilateral cerebellar hemisphere and an 

“aneurysm” in the lateral aspect of the left frontal lobe adjacent 

to cortical dysplasia. The cerebral angiography showed a 

slow-flow pial arteriovenous fistula with venous aneurysm 

and an orbital mass with capillary blush with no enlarged 

feeding artery or early draining vein, probably representing an 

hemangioma. The pial fistula was occluded with intraarterial 

n-BCA injection (Fig 1). The MR imaging of the other patient 

showed right frontoparietal cortical dysplasia in the slightly 

larger right hemisphere and enlarged ipsilateral lateral ventricle, 

consistent with unilateral megalencephaly. Also noted was 

abnormal vascular structure in the anterior interhemispheric 

fissure which was confirmed to be arterial on MRA, and a 

cerebellar DVA. The cerebral angiography performed to rule 

out AVM demonstrated dysplasia of the right pericallosal 

artery with normal parenchymal blush and no early draining 

vein as confirmed by superselective catheterization. 

Moreover, a small aneurysm was present at left MCA bifurcation 

which then was treated by endovascular means. 

CONCLUSION 

Accompanying DVA in the vicinity of cortical dysplasia is a 

well known occurrence; however, arterial abnormalities are 

very rare in migration disorder. These two cases are examples 

of in utero disturbances during the neuronal migration resulted 

in several vascular abnormalities, not only in the same 

anatomical distribution but also elsewhere in the cranium, in 

addition to cortical dysplasia. We also emphasize that ipsilateral 

arterial ectasia may be seen in unilateral megalencephaly. 

KEY WORDS: Arterial abnormalities, cortical dysplasia, neuronal 

migration disorder 


Pituitary Gland with “Figure-Eight” Morphology: A 

Normal Variant! 

Karimi, S. · Tabar, V. · Lis, E. · Holodny, A. I. 


New York, NY 

PURPOSE 

To present a case of pathologically proven normal pituitary 

gland with suprasellar extension in a “figure-eight/snowman” 

morphology on MR imaging. 

267 


A15-year-old girl was found to have apparent enlargement 

of the pituitary gland, with suprasellar extension in a “figureeight/snowman” 

configuration, during workup of frequent 

headaches. The headaches were bilateral frontal migrating to 

the occipital region and were not associated with visual 

changes, nausea, or vomiting. The patient was otherwise 

healthy with no past medical history. She had reached 

menarche at age 13 with normal periods. The physical and 

neurologic examinations were unremarkable including normal 

visual field testing. Extensive laboratory workup, which 

included an endocrine profile, did not reveal any abnormalities. 

Transphenoidal biopsy revealed no evidence for tumor. 

Volumetric analysis using MR imaging demonstrated a 

diminutive pituitary fossa and a normal volume to the pituitary 

gland including the suprasellar component. We believe 

that this unusual configuration of the pituitary was precipitated 

by the normal enlargement of the pituitary gland in 

pubescent girls in the setting of a small pituitary fossa. 

RESULTS 

The MR imaging revealed an apparent enhancing sellar mass 

with suprasellar extension in the form of a “figure-eight” 

(Fig 1). The apparent lesion demonstrated homogeneous 

enhancement and did not appear separate from the pituitary. 

The superior margin was approximately 12 mm above the 

sellar floor and abutted the chiasm but without significant 

mass effect or displacement. The infundibulum was normal 

and the sellar floor was intact without enlargement or 

destruction. The sella turcica appeared small. The central 

skull base and the sphenoid sinus as well as the brain itself 

were otherwise unremarkable. The volume of the pituitary 

fossa was diminutive (113 mm 3 ) significantly smaller (one 

third) than established minimal expected normal sella volume 

of aged-matched individuals. Interestingly, the calculated 

pituitary volume (633 mm 3 ) was close to reported normal 

volumes of age- and sex-matched adolescents (655.3 ± 99.3 

mm 3 ). The height, however, was greater (12 mm) than the 

mean height (6.5 mm) of individuals of similar sex and age. 

CONCLUSION 

The “figure-eight” shape of the pituitary has been associated 

with pituitary lesions, mainly adenoma, but it also can represent 

a normal variant as in this case. Volume measurements 

as shown in the past are more reliable in detection of a pituitary 

abnormality and hyperplasia and should be used in 

imaging assessment of the pituitary when routing imaging 

findings are equivocal or not definitive. 

KEY WORDS: Pituitary gland, figure-eight morphology, sella 

turcica 

Friday

Friday 

268 

Paper 486 Starting at 9:11 AM, Ending at 9:19 AM Health Organization Grade 1 lesion, typically with excellent 

MR Imaging Manifestations of Pediatric Posterior 

Reversible Encephalopathy Syndrome 

prognosis. We report a case of DIG with numerous extraaxial 

metastases throughout the central nervous system (CNS). 

Feygin, T. · Simon, E. · Pollack, A. · Bilaniuk, L. · 

Zimmerman, R. A. 



PURPOSE 

Posterior reversible encephalopathy syndrome (PRES) 

attracts growing attention in the clinical literature, with the 

emphasis on necessity of prompt and proper recognition of 

this clinicoradiologic entity on cross-sectional imaging. To 

characterize the neuroimaging features of PRES in the pediatric 

population. 


An electronic search of radiology database from 1992 to 2004 

for MR images with the key words “hypertensive encephalopathy,” 

“PRES,” and “brain” yielded a list of 30 cases. Images 

were retrieved and retrospectively reviewed. Medical records 

were reviewed for relevant clinical information. 

RESULTS 

Twenty-two cases with clinicoradiologic features of PRES 

were diagnosed, utilizing a 1.5 T superconducting Siemens 

Vision System. The majority of the studies included T1-, T2weighted, 

FLAIR, DWI with apparent diffusion coefficient 

(ADC) mapping and 2D FLASH GRE sequences. All cases 

demonstrated the well described pattern of vasogenic cortical 

and subcortical edema in the characteristic distribution, 

predominantly the posterior watershed zones; in 5 cases the 

basal ganglia and cerebellum also were involved . All available 

diffusion-weighted images demonstrated increased 

water motion with increased ADC. Increased signal in the 

affected areas was demonstrated with application of magnetization 

transfer technique in 10 cases, especially notable on 

follow-up examinations, when the remainder of the imaging 

sequences had returned to normal. 

CONCLUSION 

Neuroimaging manifestations of PRES in pediatric patients 

in many features are similar to those described in adults. 

Magnetization transfer technique appears to be a helpful tool 

in the recognition of subtle damage to affected areas. This is 

especially important on follow-up examinations when these 

areas may appear normal otherwise, given the false impression 

of complete reversibility. 

KEY WORDS: PRES, hypertensive encephalopathy, MR 

imaging 


Metastatic Desmoplastic Infantile Ganglioglioma 

Bourgeois, D. · Welsh, C. T. · Rumboldt, Z. 



PURPOSE 

Desmoplastic infantile ganglioglioma (DIG) is a rare 

intracranial neoplasm of infancy categorized as a World 


A 6-month-old girl presented with enlarging head circumference 

and left-sided weakness. Imaging studies revealed a 

large cystic lesion in the right cerebral hemisphere and 

extraaxial enhancing masses and the patient was taken to 

surgery for partial resection of the mass and drainage of the 

cysts. A dissection plane was obtained easily, and communication 

between the cysts and adjacent dural-based masses 

was not grossly evident. Final pathologic diagnosis of both 

the cyst wall and dural mass was DIG. The patient’s leftsided 

weakness postoperatively improved. 

RESULTS 

MR imaging of the brain showed a large nonenhancing multicystic 

lesion without a visualized solid component in the 

right cerebral hemisphere with associated prominent mass 

effect. Multiple strongly enhancing extraaxial masses of relatively 

low T2 signal were seen also. These masses were 

arising from the dura and leptomeninges and some of them 

were adjacent to the cystic mass, but without evidence of 

direct extension. This appearance of apparently discontiguous 

lesions is very unsual for DIG. MR imaging of the spine 

revealed leptomeningeal enhancing nodules involving cervical 

and thoracic spinal cord and cauda equina, consistent 

with drop metastases. On histology, the tumor exhibited fascicles 

of elongated, spindled astrocytic cells focally admixed 

with larger, more polygonal cells with neuronal features. In 

addition, scattered foci of primitive neuroblastic cells were 

noted, typical for DIG. Immunohistochemical staining confirmed 

the presence of both neuronal and glial differentiation.Tumor 

was identified inside the Virchow-Robin spaces, 

indicating communication with the subarachnoid space. The 

Ki-67 proliferation index was generally less than 5%, with 

focal increased proliferative activity (around 13%) in blastic 

areas, which is higher than previously reported in these 

lesions.

CONCLUSION 

Desmoplastic infantile ganglioglioma may present with 

aggressive behavior and multiple metastases throughout the 

CNS, which can be associated with focally increased proliferative 

index. Tumor spread of aggressive DIG appears to 

occur predominantly through the perivascular and subarachnoid 

spaces and not by direct extension. 

KEY WORDS: Desmoplastic infantile ganglioglioma, 

metastatic 


8:00 AM - 9:30 AM 

Room 107 

(75c) ADULT and PEDIATRIC 

BRAIN: Neoplasms 







Moderators: Thomas Kim, MD 

Edmond Knopp, MD 


Cerebral Blood Volume Predicts Patient Outcome Better 

than Histopathology in Low-Grade Gliomas Using 

Dynamic Susceptibility Contrast Perfusion MR Imaging 

Law, M. · Oh, S. · Babb, J. · Wang, E. · Inglese, M. · Zagzag, 

D. · Knopp, E. · Johnson, G. 


New York, NY 

PURPOSE 

To determine whether cerebral blood volume (CBV) can predict 

patient outcome, specifically tumor progression and 

malignant transformation, in low-grade gliomas (LGGs) and 

thus overcome some of the limitations of histologic assessment. 


Thirty-five patients with histologically diagnosed LGGs (21 

low-grade astrocytomas and 14 low-grade oligodendroglioma 

and low-grade mixed oligoastrocytomas), were 

studied with dynamic susceptibility contrast-enhanced perfusion 

MR imaging. Wilcoxon tests were used to compare 

patients in different response categories (complete response, 

stable, progressive, death) with respect to baseline CBV. 

269 

Kaplan-Meier time-to-progression curves were generated. 

Weibull survival models were used to predict the association 

of CBV with survival and time to progression. Tumor volumes 

and CBV measurements were obtained at the initial 

examination and again at follow up. 

RESULTS 

Lesions with CBV < 1.75 had a median time to progression 

of 4620 ± 433 days and lesions with CBV > 1.75 had a median 

time to progression of 245 ± 62 days. Patients manifesting 

an adverse event (either death or progression) had significantly 

higher (p = 0.003) CBV than patients without 

adverse events (either complete response or stable disease). 

Using Weibull survival models, neither age (p = 0.339) nor 

gender (p = 0.90) was associated with overall survival, 

whereas CBV exhibited a significant negative association 

with survival (p = 0.001) such that low CBV values were 

associated with longer survival times. The same basic conclusion 

held for time to progression: neither age (p = 0.312) 

nor gender (p = 0.285) was associated with time to progression, 

whereas CBV exhibited a significant negative association 

with disease-free survival (p = 0.001), such that low 

CBV values were associated with longer disease-free survival 

times. Baseline CBV was not correlated with tumor 

volume changes seen in postcontrast T1 (p = 0.896) or T1 (p 

= 0.338) on follow-up imaging. 

CONCLUSION 

The current gold standard of histopathologic glioma grading 

has limitations. Partly because of this, the triage, treatment, 

and survival statistics of low-grade gliomas remain unclear. 

However, patients with misclassified gliomas will not 

receive optimum treatment. Our study strongly suggests that 

CBV measurements correlate more accurately with time to 

progression than initial histopathologic grading. 

KEY WORDS: Low-grade glioma, perfusion, survival 


Significance of Cerebral Blood Volume Measurements in 

Human Gliomas with 1p and 19q Molecular Deletions 

Law, M. 1 · Livshiz, J. 1 · Rosenblum, M. 2 · Zagzag, D. 1 

1New York University Medical Center, New York, NY, 

2Memorial Sloan-Kettering Cancer Center, New York, NY 

PURPOSE 

Loss of heterozygosity of chromosomes 1p and 19q appears 

to confirm responsiveness to chemotherapy and is associated 

with improved survival in human gliomas. There is 

histopathologic and molecular evidence to support increased 

neovascularity in gliomas with deletion at the 1p19q locus. 

The purpose of this study is to determine if there is a correlation 

between the 1p19q deletion and increased perfusion 

using dynamic susceptibility contrast perfusion MR imaging 

(DSC MRI) in human gliomas. 


Sixteen patients who underwent DSC MRI and also had 

molecular studies for 1p19q chromosomal deletion were 

reviewed. Allelic status was assessed by loss of heterozygosity 

using PCR. DNA was extracted from paraffin curls of 

brain section and nail clippings. Relative CBV measurements 

were obtained from regions of maximum perfusion as 

Friday

Friday 

determined from rCBV color maps within the glioma. These 

measurements then were correlated with the presence of a 1p 

19q chromosomal deletion in patients with gliomas. 

RESULTS 

Patients with 1p and 19q deletions (n = 7) demonstrated 

rCBV values of 10.40 ± 5.63. Patients without the deletion 

(n = 9) had rCBV values of 5.91 ± 4.13 (p = 0.12). Even 

though there was not a significant difference in the rCBV 

measurements, rCBV was found to be almost twice as high 

in patients with 1p and 19q deletions. 

CONCLUSION 

There is histopathologic and molecular evidence to support 

increased neovascularity in gliomas with 1p 19 q deletions. In 

this small sample, we found elevated rCBV in patients with 1p 

19q chromosomal deletions. Thus rCBV may serve as an 

important physiologic imaging marker for identifying gliomas 

with 1p 19q chromosomal deletions potentially conferring 

high chemosensitivity. A prospective study of a larger cohort 

is underway to determine if it may be advantageous to obtain 

molecular studies in lesions demonstrating high rCBV. 

KEY WORDS: Perfusion, molecular, gliomas 


Contrast Enhancement after Gadolinium 

Administration: Results from Double-Blind, 

Randomized, Intraindividual Studies Comparing 

Gadobenate Dimeglumine with Standard Gadolinium- 

Based MR Contrast Agents 

Essig, M. 1 · Tartaro, A. 2 · Tartaglione, T. 3 · Knopp, M. V. 4 

1 German Cancer Research Center, Heidelberg, GERMANY, 

2 G. d'Annunzio University, Chieti-Pescara, ITALY, 3 Catholic 

University "Sacra Cuore", Rome, ITALY, 4 Ohio State 

University, Columbus, OH 

PURPOSE 

Gadobenate dimeglumine (MultiHance®, Bracco Imaging, 

Milan, Italy) is a gadolinium-based MR imaging contrast 

agent possessing an in vivo relaxivity approximately twice 

that of standard gadolinium chelates, due to a weak and transient 

interaction with serum proteins. We performed a blinded, 

independent off-site evaluation of the quantitative and 

qualitative enhancement obtained after a dose of 0.1 

mmol/kg gadobenate dimeglumine, as compared to the 

enhancement obtained after the same dose of other gadolinium 

chelates (Gd-DTPA or Gd-DOTA). The study was 

designed to evaluate whether the higher signal seen anecdotally 

in early studies with this agent was detectable in a clinical 

population. 


Patients with suspected glioma or cerebral metastasis (n = 

45) were studied in separate imaging sessions after 0.1 

mmol/kg gadobenate dimeglumine (MultiHance) and the 

same dose of either gadopentetate dimeglumine (Magnevist, 

n = 23) or gadoterate meglumine (Dotarem, n = 22). Imaging 

parameters and equipment were identical for the two examinations 

for each patient, and the patients were randomized as 

to which agent they received first. Each contrast agent was 

administered by power injector at 2 mL/s with the second 

agent administered between 24 hours and 14 days later. 

270 

Images were acquired predose (T1-weighted SE, T2-weighted 

FSE sequences) and postdose (sequential T1-weighted SE 

sequences at 2, 4, 6, 8, 10, 15 min) at either 1 T or 1.5 T 

using a dedicated head coil. Two independent off-site readers 

blinded to clinical information and contrast agent administered 

performed quantitative assessments corrected for precontrast 

values [lesion-to-brain ratio (L/B), contrast-to-noise 

ratio (C/N) and % lesion enhancement (%En)] and qualitatively 

evaluated lesion enhancement using a five point scale 

(0-4). Differences in mean quantitative parameters were 

assessed using paired t-tests while differences in qualitative 

evaluations were tested using the Wilcoxon signed rank test. 

RESULTS 

Quantitative assessments could be performed on images 

from 43 of 45 patients. Both readers noted significantly 

greater L/B (p < 0.003), C/N (p < 0.03) and %En (p < 

0.0001) for gadobenate dimeglumine-enhanced images at all 

time-points from 2 min postcontrast. Qualitative matchedpairs 

assessment of all 90 examinations revealed significant 

preferences for gadobenate dimeglumine over the comparator 

agents for lesion border delineation (p < 0.004, both readers), 

lesion internal morphology (p < 0.008, both readers), 

global contrast enhancement (p < 0.0001, both readers) and 

global diagnostic preference (p < 0.0005, both readers). 

Several additional lesions were detected on postdose images 

after gadobenate dimeglumine that were not seen on the corresponding 

comparator images (reader 1: 75 vs 72; reader 2: 

77 vs 72). In addition, interreader agreement was improved 

significantly on the gadobenate dimeglumine studies 

(weighted kappa for contrast enhancement = 0.244; 95% C.I. 

0.065, 0.422 after gadobenate dimeglumine vs 0.094, 95% 

C.I. -0.078, 0.267 after comparator). 

CONCLUSION 

Gadobenate dimeglumine showed improved enhancement 

relative to two comparator agents in this independent blinded 

reading of images from two randomized, double blind, 

crossover studies with gadobenate dimeglumine and either 

gadopentetate dimeglumine or gadoterate meglumine. The 

improved contrast enhancement seen with gadobenate 

dimeglumine contributed to the improved detection, delineation, 

and conspicuity of enhancing intracranial lesions, 

improved interreader agreement, and allowed the detection 

of additional enhancing lesions. In clinical practice 

improved enhancement with gadobenate dimeglumine may 

facilitate presurgical planning and postsurgical follow up. 

KEY WORDS: Gadobenate dimeglumine, gadolinium, relaxivity 


Glucocorticoid Analogues Influence MR Imaging-Based 

Cerebral Perfusion and Blood-Tumor Barrier Kinetics 

Wilkinson, I. D. · Jellineck, D. A. · Levy, D. · Giesel, F. L. · 

Miller, B. A. · Griffiths, P. D. 

University of Sheffield 


PURPOSE 

Administration of a glucocorticoid analogue (dexamethasone) 

is common following initial diagnosis of an intracranial 

tumor. The underlying physiologic response that gives

ise to often dramatic clinical improvement remains unclear. 

This study sought to investigate localized vascular perfusion 

plus brain-tumor barrier and blood-brain barrier (BTB/BBB) 

integrity before and after administration of dexamethasone 

in patients with enhancing cerebral mass lesions. 


Seventeen patients (11 astrocytoma IV, 2 anaplastic oligodendroglioma, 

1 sarcoma, 1 meningioma, 1 demyelinating 

plaque and 1 unknown) underwent MR imaging at 1.5 T 

(Eclipse, Philips Medical Systems, Cleveland, Ohio) before 

and 3 days after initiation of high-dose dexamethasone. Two 

exogenous contrast-based methods were used. Contrast “T1uptake” 

time curves were calculated from dynamic, T1weighted, 

RF-spoiled FAST datasets [50 frames over 162 

sec, 10 ml bolus of Gadovist (Schering AG, Germany)]. T1uptake 

was defined as maximum signal change relative to 

baseline signal from normal-appearing contralateral white 

matter. Regional CBV and first moment mean Transit Time 

(TTFM) were assessed using a dynamic, T2*-weighted single-shot 

EPI technique (1) (70 frames over 98 sec; 10 ml 

bolus Gadovist). Results obtained before and after initiation 

of dexamethasone were compaired using paired t-tests. 

RESULTS 

Following initiation of steroid treatment: tumor rCBV was 

reduced by 27% (p < 0.005); contralateral normal-appearing 

white matter rCBV was reduced by 16% (p < 0.05); contralateral 

TTFM increased by 6% (p < 0.05) and maximum 

uptake of tumor contrast was found to have decreased by 

28% (p < 0.005). 

CONCLUSION 

Suggested mechanisms underlying symptom resolution following 

dexamethasone therapy include a decrease in vasogenic 

edema following tightening of the BTB or changes in 

rCBV with subsequent lowering of ICP. Our results are supportive 

of both of these possible mechanisms. In addition to 

changes localized to tumor, we found that rCBF within normal-appearing 

white matter also is lowered following drug 

initiation. This work highlights the need to control for steroid 

therapy in MR studies involving cerebral neoplasms, particularly 

in those which assess or may be influenced by cerebral 

perfusion, BTB, or BBB integrity. 

REFERENCES 

1. Wilkinson ID, et al. Short-term changes in cerebral microhemodynamics 

following carotid stenting assessed by MR 

perfusion imaging. AJNR Am J Neuroradiol 2003;24(8):1501- 

1507 

KEY WORDS: Steroids, MR perfusion, blood-brain barrier 

271 


Characterization of Diffusion Tensor Abnormalities in 

Peritumoral White Matter: An Image-Guided Biopsy 

Study 

Price, S. J. · Dean, A. F. · Jena, R. · Hutchinson, P. J. A. · 

Burnet, N. G. · Pickard, J. D. · Gillard, J. H. 

University of Cambridge 

Cambridge, UNITED KINGDOM 

PURPOSE 

The propensity of gliomas to infiltrate surrounding white 

matter tracts is a major factor in the failure of current treatments. 

Current imaging methods are unable to identify the 

tumor margin accurately and this has major implications in 

both surgical and radiotherapy planning. Diffusion tensor 

imaging (DTI) is sensitive to subtle white matter disruption 

and can detect a larger abnormality around gliomas than T2weighted 

imaging (1). Using diffusion tissue signatures can 

further classify white matter involvement into disruption, 

infiltration, and displacement (2). The aim of this study is to 

correlate these markers with tumor histology using imageguided 

biopsies. 


Sixteen patients (mean age 48.5) requiring biopsy of an 

intracranial presumed glioma were recruited. Patients were 

imaged preoperatively at 3 T using a T2-weighted, a singleshot, 

spin-echo, echo-planar DTI sequence (5 b-values, 12 

directions) and a gadolinium-enhanced 3D SPGR sequence. 

The DTI data were processed using an in-house program 

implemented in MATLAB. Maps of the isotropic component 

p and the anisotropic component q were generated and 

coregistered to the SPGR sequence used for image guidance. 

All patients underwent image-guided biopsies with samples 

taken from the selected target and then at centimeter intervals. 

Regions of interest that corresponds to the area sampled 

(10 x 4 mm) were determined at each of the biopsy sites and 

the T1- and T2-weighted appearances noted. Diffusion tissue 

signatures were calculated for each site and compared to a 

region from the normal contralateral hemisphere. All biopsy 

samples were examined by an experienced neuropathologist 

and classified as purely tumor, normal tissue infiltrated with 

tumor, or normal brain. 

RESULTS 

A total of 63 regions was biopsied. Half of the biopsy tracts 

included normal brain. Diffusion tensor imaging could accurately 

identify tumor in all of the 37 biopsy sites that contained 

pure tumor. T1- and T2-weighted sequences failed to 

identify tumor in two cases. In the 13 samples of tumor infiltrated 

brain, T1-weighted imaging could identify 6/13 and 

T2-weighted could identify 7/13. Diffusion tensor imaging 

tissue signatures correctly identified tumor infiltration in 

12/13 cases. There was one false-positive in an area of normal 

brain with a perivascular mononuclear infiltrate. Overall 

the tissue signatures had a sensitivity of 96% and specificity 

of 85%. 

CONCLUSION 

Diffusion tensor tissue signatures are more sensitive at identifying 

tumor infiltration than conventional T1- or T2weighted 

sequences. Using the tissue signature method, you 

can define a region around a tumor with a reduction in the 

Friday

Friday 

anisotropic component (q) that is due to tumor disruption, 

and an area outside this of increased isotropic component (p) 

that is due to tumor infiltration. This may provide a useful 

technique for more individualized radiotherapy planning. 

REFERENCES 

1. Price SJ, Burnet NG, Donovan T, et al. Diffusion tensor imaging 

of brain tumours at 3T: A potential tool for assessing 

white matter tract invasion? Clin Radiol 2003;58:455-462 

2. Price SJ, Pena A, Burnet NG, et al. Tissue signature characterisation 

of diffusion tensor abnormalities in cerebral 

gliomas. Eur Radiol 2004;14:1909-1917 

KEY WORDS: Diffusion tensor, gliomas, white matter 


Quantitative Proton MR Spectroscopy of Untreated 

Pediatric Astrocytoma 

Liu, X. 1 · Panigrahy, A. 1 · Gonzalez-Gomez, I. 1 · Nelson, M. D. 1 

· Krieger, M. D. 1 · Gilles, F. H. 1 · McComb, J. G. 1 · Bluml, S. 1,2 

1 Children’s Hospital Los Angeles, Los Angeles, CA, 2 Rudi 

Schulte Research Institute, Santa Barbara, CA 

PURPOSE 

Astrocytomas are among the most common types of brain 

tumors in children. Differentiation between different subclasses 

of astrocytomas is based on the histologic evaluation. 

In addition to morphologic alterations, these tumors may differ 

also in their metabolic profiles. It has long been speculated 

that biochemical alteration may be used to assess tumor 

malignancy and contribute to clinical decision making. 

However, most of the earlier in vivo studies focused on the 

prominent peaks of the 1H spectrum N-acetyl-aspartate 

(NAA), creatine (Cr), and choline (Cho) and only ratios of 

metabolite intensities were reported. In this study it was 

investigated whether absolute quantitation of metabolites of 

untreated brain tumors in pediatrics improves the differentiation 

of astrocytomas with different grades of aggressiveness. 


MR spectra from 26 subjects with histologically confirmed 

astrocytomas (17 pilocytic astrocytoma, 5 nonanaplastic 

astrocytoma, 4 anaplastic astrocytoma) were evaluated. MR 

imaging and spectroscopy were performed on a 1.5 T GE 

clinical scanner. Single voxel 1H spectra were obtained 

using a PRESS sequence with short echo time TE = 35 ms, 

repetition tome TR = 1.5 s, and 128 signal averages. Regions 

of interest (ROIs) were selected carefully to maximize tumor 

tissue and minimize partial volume of surrounding 

normal/edemous appearing tissue (Fig.1). Spectra were 

processed using fully automated LCModel software 

(Stephen Provencher Inc, LCModel V6). Since metabolites 

are deemed to be intracellular, concentrations were corrected 

for the fraction of necrotic/cystic volume within the ROI. 

Concentrations are reported in mmol/kg tissue. 

RESULTS 

Pilocytic astrocytoma exhibited reduced total Cr concentration 

when compared with all other astrocytoma cases (p < 

0.0001). Cr/Cho also separated pilocytic from other astrocytoma 

but was a less significant differentiator (p < 0.005). 

Cho concentration was lower in pilocytic astrocytoma than 

272 

in other astrocytoma but did not reach statistical significance. 

Cr and myo-inositol concentrations were higher in 

anaplastic than in nonanaplastic astrocytoma (Table 1), however, 

the significance of this finding is uncertain due to the 

small number of subjects studied. 

Table 1: Metabolite concentrations (mmol/kg) and concentration ratios 

of pediatric astrocytomas 

[NAA] [Cr] [Cho] [mI] NAA/Cho Cr/Cho mI/Cho 

Anaplastic 2.43±1.49 7.61±1.75 4.47±3.93 14.79±4.010.93±0.73 

2.38±1.15 4.38±1.65 

astrocytoma 

Non-anaplastic 1.61±1.59 3.74±0.77 3.45±0.78 6.96±3.37 0.52±0.60 

1.21±0.48 2.48±1.85 

astrocytoma 

Pilocytic 2.06±1.28 0.96±0.83 2.29±1.26 4.21±2.34 1.23±1.47 

0.38±0.26 1.94±1.24 

astrocytoma 

CONCLUSION 

Astrocytomas of varying malignancy do exhibit different 

metabolic profiles. Low absolute Cr is specific for pilocytic 

astrocytoma. These findings demonstrate the importance of 

absolute quantitation of metabolites. 

KEY WORDS: MR spectroscopy, pediatric, brain tumor 


Hyperintense Rim Sign on FLAIR Images in 

Dysembryoplastic Neuroepithelial Tumors 

Parmar, H. A. · Ozelame, R. · Chuang, S. · Hawkins, C. · 

Rutka, J. · Blaser, S. 



PURPOSE 

Dysembryoplastic neuroepithelial tumors (DNET) are 

benign cortical-based lesions affecting young adults and 

characterized by drug-resistant seizures and normal neurologic 

examination. On MR imaging, superficial location, triangular 

appearance with internal “septations,” and relative 

absence of contrast enhancement are well described features. 

Unfortunately, both low-grade glioma and oligodendroglioma 

can show similar neuroimaging features and differentiation 

between these tumors is important due to their 

markedly different treatment regime. The finding of inner 

table remodeling is useful, but is dependent upon location of 

the tumor. We describe the hyperintense rim on FLAIRweighted 

images in patients with DNET tumors and determine 

its role in preoperative diagnosis. 


We retrospectively analyzed imaging features in seven 

patients with DNETs. All these patients had undergone surgery 

for refractory seizures and had a pathologic confirma-

tion of DNET. All patients had MR study performed on a 1.5 

T system and a fast FLAIR sequence was performed in all 

patients. Detailed radiopathologic correlation will be shown, 

with a particular emphasis at the findings at the periphery of 

the tumor, corresponding to the site of abnormality on neuroimaging. 

RESULTS 

There were seven patients who matched our inclusion criteria; 

four girls, three boys. The age of presentation varied 

from 5 years to 18 years (average 10 years, 4 months). Three 

tumors were located in the frontal lobe while two each were 

located in the temporal and parietal lobes. All tumors were 

hypointense on T1- and hyperintense on T2-weighted 

images, with no perifocal vasogenic edema. In five patients 

(sensitivity of 72%) FLAIR images showed a well defined, 

thick hyperintense rim around these tumors, forming either a 

complete or incomplete rim. Rim on FLAIR was considerably 

more distinct than on PD images. Pathologic evaluation 

of these tumors suggested presence of subtle cortical dysplasia 

at the periphery of these tumors with mild to moderate 

amount of balloon cells. 

CONCLUSION 

With the exception of inner table remodeling, imaging features 

in DNET are usually nonspecific. Hyperintense rims 

have been described in one report on proton density images. 

The hyperintense rim sign on FLAIR images is considerably 

more distinct, is a fairly sensitive sign and is a helpful adjuvant 

to preoperatively diagnose these tumors. Further study 

with a large patient population is, however, warranted to substantiate 

our claims. 

KEY WORDS: Neoplasms, seizures, dysembryoplastic neuroepithelial 

tumor 

273 


Clinical and MR Characteristics of Dysembryoplastic 

Neuroepithelial Tumors in Children 

Rodriguez, D. P. 

Children’s Hospital 

Boston, MA 

PURPOSE 

To review the clinical and MR imaging features of dysembryoplastic 

neuroepithelial tumors in children. 


A retrospective analysis of pediatric patients who had the 

pathologic diagnosis of DNET between 1995 and 2004 was 

done. Clinical and imaging findings were reviewed. The following 

MR imaging characteristics were recorded including 

location, size, lesion signal characteristics, pattern of 

enhancement, presence of hydrocephalus, and remodeling of 

the calvarium. 

RESULTS 

A total of fifteen children, eight girls and seven boys with a 

proven pathologic diagnosis of DNET were identified. The 

age range was 11 months to 18 years (mean age 9.2 years). 

Twelve children presented with seizures and one patient presented 

with papilledema. Two lesions were detected incidentally, 

one from a work up for developmental delay and the 

other for head trauma. The tumors were in the following 

locations: five in the temporal lobe, five in the parietal lobe, 

four in the frontal lobe, and one lesion in the occipital lobe. 

The lesions typically involved the cortex and subcortical 

white matter. The maximum dimension at presentation of the 

smallest lesion was 0.8 cm, while the largest dimension at 

presentation was 9.4 cm. On T1-weighted images, all of the 

tumors were relatively isointense to hypointense to gray matter 

and hyperintense on T2-weighted images. Five patients 

had multilobulated, cystic lesions. An enhancing nodule was 

demonstrated in five patients, with heterogeneous enhancement 

in one patient. Central calcification was present in one 

lesion. Clear remodeling of the calvarium was present in 3 

cases. None of the patients had associated hydrocephalus. 

Three of these patients had evidence of regrowth of their 

tumors and required a second operation. No adjuvant treatment 

was required for any of these patients. 

CONCLUSION 

Dysembryoplastic neuroepithelial tumors are rare tumors in 

childhood with variable and unique imaging characteristics 

on MR imaging. The majority of children have good longterm 

outcomes. 

KEY WORDS: Pediatric, brain, neoplasms 

Friday

Friday 

274 

Paper 496 Starting at 9:04 AM, Ending at 9:12 AM Paper 497 Starting at 9:12 AM, Ending at 9:20 AM 

Comparison of Imaging Features of Atypical Primary Intracranial Atypical Teratoid Rhabdoid 

Teratoid/Rhabdoid Tumor with that of Primitive Tumors in Children: MR Imaging Features and Patient 

Neuroectodermal Tumor/Medulloblastoma in Children Outcomes 

Koral, K. 1,2 · Gargan, L. 2 · Weprin, B. 2 · Rollins, N. K. 1,2 

1University of Texas Southwestern Medical Center, Dallas, 

TX, 2Children’s Medical Center of Dallas, Dallas, TX 

PURPOSE 

Atypical teratoid/rhabdoid tumor (AT/RT) is a highly malignant 

tumor whose imaging features may resemble that of 

primitive neuroectodermal tumor/medulloblastoma 

(PNET/MB). We compared imaging findings of AT/RT to 

those of PNET/MB to determine if any distinguishing imaging 

features exist. 


Search of neurosurgery database from our institution yielded 

18 patients presenting with AT/RT and 103 patients with 

PNET/MB, including 16 patients with PNET/MB who were 

< 5 years of age at presentation. All patients have had their 

surgeries performed at our institution. Two pediatric neuroradiologists 

evaluated preoperative MR studies of 18 

patients with AT/RT and 16 patients with PNET/MB for 

presence of hydrocephalus, tumor location, hemorrhagic 

changes, cystic changes within the tumor, signal intensity of 

the solid portion, enhancement characteristics and presence 

of leptomeningeal disease. Findings on preoperative MR 

imaging for AT/RT were compared to PNET/MB as were 

survival rates at 1 year. 

RESULTS 

Atypical teratoid/rhabdoid tumor patients presented earlier 

than PNET/MB patients (mean age at presentation 1.16 

years vs 6.46 years). Survival at 1 year was 29.5% for AT/RT 

vs 80.77% for PNET/MB. Although majority of tumors were 

infratentorial in both groups (66.7% for AT/RT, 68.75% for 

PNET/MB), hydrocephalus was more common in 

PNET/MB (87.5% vs 55.6%). Hemorrhage was more common 

in AT/RT (22.2% vs 12.5%). Intense enhancement was 

more common in PNET/MB (%62.5 vs 38.9%). There were 

more mixed tumors (with approximately equal amounts of 

solid and cystic components) in AT/RT group (27.8% vs 

6.25%). PNET/MB was more homogenously solid than 

AT/RT (50.0% vs 33.33%). The majority (72.2%) of AT/RT 

and all of PNET/MB had solid components that were moderately 

to severely hypointense on T2-weighted images. Two 

(11.1%) AT/RTs were extremely hypointense on T2-weighted 

imaging. Leptomeningeal/spinal involvement was more 

common in AT/RT (33.33%) than PNET/MB (18.75%). 

CONCLUSION 

Imaging findings of AT/RT and PNET/MB are similar, but 

not identical. Atypical teratoid/rhabdoid is more likely to be 

heterogenous and hemorrhagic than PNET/MB. Presence of 

extreme hypointensity of T2-weighted imaging is a feature 

seen in AT/RT, but not seen on PNET/MB. 

KEY WORDS: Atypical teratoid/rhabdoid tumor, primitive 

neuroectodermal tumor, medulloblastoma 

Meyers, S. P. 1 · Khademian, Z. P. 2 · Chuang, S. H. 3 · Biegel, 

J. A. 2 · Korones, D. N. 1 · Zimmerman, R. A. 2 

1University of Rochester Medical Center, Rochester, NY, 

2Children’s Hospital of Philadelphia, Philadelphia, PA, 

3Hospital for Sick Children, Toronto, ON, CANADA 

PURPOSE 

Primary atypical teratoid/rhabdoid tumors (ATT/RhTs) are 

rare malignant intracranial neoplasms usually occurring in 

young children. The objectives of this study were to characterize 

the MR imaging features and locations of primary 

intracranial ATT/RhTs, determine the frequency of disseminated 

disease in the CNS at diagnosis and postoperatively, 

and assess patient outcomes. 


The preoperative cranial MR images of 13 patients with 

ATT/RhTs were reviewed retrospectively for evaluation of 

lesion location, size, MR signal and enhancement characteristics, 

and presence of disseminated intracranial tumor. 

Postoperative MR images of the head and spine for 17 

patients were reviewed for the presence of locally recurrent 

or residual tumor, and disseminated neoplasm. Imaging data 

were correlated with patient outcomes. 

RESULTS 

At diagnosis, patients ranged in age from 4 months to 15 

years (median = 2.9 years). Primary ATT/RhTs were intraaxial 

in 94%. The single primary extraaxial lesion was located 

in the cerebello-pontine angle cistern. ATT/RhTs were 

infratentorial in 59%, supratentorial in 29%, and both infraand 

supratentorial in 12%. A germline mutation of the INI1 

tumor suppressor gene was responsible for the simultaneous 

occurrence of an intracranial ATT/RhT and a malignant renal 

rhabdoid tumor in a 4-month-old patient. Mean tumor size 

was 3.6 x 3.7 x 3.9 cm. On short TR images, ATT/RhTs typically 

had heterogeneous intermediate signal, as well as 

zones of low (46%), high (15%), or both low and high (31%) 

signal from cystic and/or necrotic regions, hemorrhage, or 

both, respectively. On long TR/long TE images, solid portions 

of ATT/RhTs typically had heterogeneous intermediate 

to slightly high signal with additional zones of high (54%), 

or both high and low signal (38%) secondary to cystic and/or 

necrotic regions, prior hemorrhage and/or calcifications. All 

but one ATT/RhT showed contrast enhancement. The fraction 

of tumor volume showing enhancement was greater than 

two thirds in 54%, between one third and two thirds in 31%, 

and less than one third in 8%. Disseminated tumor in the leptomeninges 

was seen with MR imaging in 23% of patients at 

diagnosis/initial staging, and occurred in another 46% from 

4 months to 2.8 years (mean = 1.1 years) after surgery and 

earlier negative imaging examinations. The overall 1-year 

and 5-year survival probabilities were 71% and 28%, respectively. 

Patients with MR imaging evidence of disseminated 

leptomeningeal tumor had a median survival of 16 months 

compared to 149 months for those without disseminated 

tumor, (p < 0.004, log rank test).

CONCLUSION 

Atypical teratoid/rhabdoid tumors are usually intraaxial 

lesions of which more than half are infratentorial. The unenhanced 

and enhanced MR imaging features of ATT/RhT are 

often variable secondary to cystic/necrotic changes, hemorrhage, 

and/or calcifications. Poor prognosis was associated 

with MR imaging evidence of disseminated leptomeningeal 

tumor. 

KEY WORDS: Rhabdoid, teratoid, children 

Discussion 


8:00 AM - 9:32 AM 

Room 205 








Moderators: Erin M. Simon, MD 

Charles R. Fitz, MD 


Feasibility of a Standardized Multimodal MR Protocol 

for Acute Stroke 

Rovira, A. · Persiva, O. · Arenillas, J. F. · Grivé, E. · 

Schorlemmer, C. · Soto, S. · Sánchez, E. · Alvarez-Sabín, J. 

Vall d`Hebron Hospital 


PURPOSE 

Accumulated evidence from recent years has documented 

the superiority of MR imaging over CT for proper selection 

of hyperacute stroke patients who might benefit from reperfusion 

therapy. Nevertheless, there are still substantial 

doubts regarding the feasibility of MR imaging in hyperacute 

stroke because of its variable availability and the time 

constraints involved in stroke evaluation. The purpose of the 

present study is to test the feasibility of multimodal MR 

imaging in the setting of hyperacute stroke. 


From November 2000 to March 2004, 120 patients with 

hyperacute stroke were referred to our department for MR 

assessment within 6 hours after the onset of symptoms. 

275 

Inclusion in this study required accurate reporting of the following 

time intervals: 1) symptoms duration: time between 

symptoms onset and entering the MR suite; 2) delay time: 

time between arrival to the emergency room and arrival to 

the MR unit; 3) waiting time: interval between arrival to the 

MR unit and entering the MR suite; 4) scanning duration: 

time between entering and exiting the MR suite; and 5) total 

imaging duration: time between patient arrival and discharge 

from the MR unit. From November 2000 to September 2002 

a single MR unit (Magnetom Vision) was available to examine 

hyperacute stroke patients, whereas from October 2002 a 

second MR unit (Magnetom Symphony) also was available. 

The following images were obtained: 1) transverse T2weighted 

EPI GE sequence, 2) transverse diffusion-weighted 

EPI SE sequence, 3) 3D time-of-flight MR angiograms, 

and 4) transverse perfusion-weighted EPI GE sequence. 

RESULTS 

One hundred patients (54 women; mean age, 67.4 years) 

were included in the study. The mean time intervals were as 

follows: symptoms duration, 148.1 min; delay time, 36.8 

min; waiting time, 8.6 min; scanning duration, 16.6 minutes; 

and total imaging duration, 33.1 minutes. A significant 

decrease in the following time intervals was observed after 

two MR units became available: symptoms duration 167.7 

vs 139.1 minutes (P = .020); delay time 88.4 vs 41.2 minutes 

(P = .020); scanning duration 20.9 vs 14.5 minutes (P < 

.001), and total imaging duration 46.3 vs 26.5 minutes (P < 

.001). Scanning duration and total imaging duration were 

significantly lower in patients scanned with the Symphony 

system as compared to those scanned with the Vision system 

(20.6 vs 11.2 minutes for scanning duration, P < .001; 41.4 

vs 21.9 minutes for total image duration, P < .001), a fact 

that can be explained by the faster reconstruction time of the 

Symphony system and the increase in practice and experience 

of the stroke team. 

CONCLUSION 

The total time devoted to MR imaging in the setting of acute 

stroke should be within the time constraints required for 

reperfusion therapy. In our experience, this goal was 

achieved only when two fully-equipped MR units were readily 

available at any time, and in the hands of a well trained 

and experienced stroke team. Under these circumstances, 

multimodal MR imaging for acute stroke is feasible and 

applicable before therapy decisions. 

KEY WORDS: Cerebral ischemia, MR imaging 


Cerebral Venous Sinus Thrombosis: “Pseudophlebitic 

Pattern” on Fluid Attenuation Inversion Recovery 

Sequence 

Mantilla-Martin, T. 

Clínica Colsanitas S.A. 

Bogotá, D.C., COLOMBIA 

PURPOSE 

The purpose of this review is to illustrate the 

“pseudophlebitic pattern” on fluid attenuated inversion 

recovery (FLAIR) in cerebral venous sinus thrombosis. 

Friday

Friday 


Fifteen cases of dural sinus thrombosis collected from July 

2003 to July 2004 and an additional one from 2001 were 

reviewed in order to assess the MR findings. Only three of 

these cases had had previous CT head. 

RESULTS 

The main signs and symptoms of these patients included: 

headache (11/15), seizures (3/15), focal neurologic deficit 

(3/15), vomiting (2/15), vertigo (2/15), and hypertensive crisis 

(1/15). In this group of patients oral contraceptives 

(2/15), malignancy (1/15), trauma (1/15), oral corticosteroid(1/15) 

and thrombotic thrombocitopenic purpura 

(1/15) were found to be predisposing factors. There was no 

correlation between the severity of the imaging findings and 

the clinical manifestation of the thrombosis. The most common 

venous sinuses affected were the transverse (6/15) and 

the superior sagittal (4/15). There were four additional cases 

in which both the superior sagittal and the transverse were 

compromised. There was one case of sigmoid sinus thrombosis 

with extension to the intern jugular vein. The intraluminal 

thrombus (equivalent to the cord sign) was seen in all 

cases. It appeared as a hyperintense signal, on FLAIR, within 

the involved sinus. MR venography was done in six cases, 

all of which showed absence of flow of the compromised 

sinus. Brain edema was detected in six cases. It showed high 

signal intensity on T2 and FLAIR, in both the gray and the 

white matter; associated to sulcal effacement. In these six 

cases the FLAIR sequence showed linear increased signal 

intensity in the pial region next to the effaced sulci. I think 

this is caused by the slow flow within the venous collateral 

circulation; which should be considered as the equivalent to 

the “pseudophlebitic pattern” described on the venous phase 

of the angiogram. This corresponds to the second pattern on 

MR imaging described by Yuh, in which the venous pressure 

had exceeded the venous bed compliance and thus could 

cause a shift in the pressure gradient. The empty delta sign 

was seen in one case, because only in one case enhanced MR 

imaging was done. A subdural collection was found in one 

case. 

CONCLUSION 

The diminished flow of the venous collateral circulation is 

seen as linear increased signal intensity on FLAIR, in the 

pial region next to the effaced sulci. It should be taken as the 

equivalent to the “pseudophlebitic pattern” described on the 

venous phase of the angiogram. 

KEY WORDS: Cerebral venous sinus thrombosis, dural sinus 

thrombosis, MR imaging 


2D and 3D Time-of-Flight MR Angiography Using 

Parallel and Nonparallel Imaging Techniques for the 

Evaluation of Carotid Stenosis 

Sunenshine, P. J. · Pramanik, B. · Law, E. M. · Hecht, E. M. 


New York, NY 

PURPOSE 

Two-dimensional and 3D time-of-flight (TOF) MR angiography 

(MRA) are well established, noninvasive ways to 

assess carotid artery stenosis. With the advent of parallel 

276 

acquisition techniques and multichannel neck coils, examination 

time can be decreased significantly and/or spatial resolution 

can be increased significantly without significant 

compromise of the other. In general, 3D TOF MRA has 

excellent spatial resolution but is relatively time consuming 

where as 2D TOF MRA permits better evaluation of slow 

flow, but with poorer spatial resolution. The purpose of this 

study is to compare 3D and 2D TOF MRA with and without 

parallel acquisition techniques, to determine if parallel imaging 

can be used to decrease the time of sequence for conventional 

3D TOF without compromising interpretation of 

carotid stenosis, and to determine if parallel imaging can be 

used to improve the spatial resolution of conventional 2D 

TOF to a level comparable to 3D TOF. 


Prospectively, five consecutive patients presenting for 

carotid MRA underwent 2D and 3D TOF MRA with and 

without parallel acquisition techniques (GRAPPA, acceleration 

factor 3) at 1.5 T. Conventional 3D imaging was performed 

with a voxel size of 0.7 x 0.5 x 0.9 mm. Threedimensional 

TOF with parallel imaging was performed using 

similar parameters to maintain spatial resolution; however 

time of acquisition was decreased from 6:21 minutes to 2:40 

minutes. Conventional 2D TOF was performed using a voxel 

size of 0.9 x 0.8 x 3.0 mm. Using parallel imaging (GRAP- 

PA, R = 3) resolution was improved to achieve a voxel size 

of 0.4 x 0.4 x 3.0 mm, while time of acquisition was kept 

constant at 3:40 minutes. Quantitative assessment was performed 

by calculating signal-to-noise ratio (SNR) and contrast-to-noise 

ratios (CNR) in the distal common carotid 

artery using standard methods. Qualitative assessment was 

performed by two neuroradiologists blinded to imaging 

method. Subjective assessment of image quality, noise, 

stenosis delineation, intravascular signal intensity, degree of 

stenosis (NASCET criteria), and level of confidence in diagnosis 

was performed using a 4 point scale. 

RESULTS 

Signal-to-noise ratio was decreased for both 3D and 2D TOF 

MRA with parallel acquisition techniques, but was still adequately 

high for both groups. Qualitative parameters such as 

overall quality are not significantly different as shown in the 

table. 

TABLE 1 Conventional TOF vs. TOF with Parallel Imaging 

3D 3D IPAT P-value 2D 2D IPAT P-value 

Wilcoxin 

Average SNR 106.7 82.3 p ≤ 0.0625 67.0 56.1 p ≤ 0.125 

Average CNR 81.7 63.3 p ≤ 0.125 57.0 43.2 p ≤ 0.125 

Overall Quality 2.2 2.4 p ≤ 0.5 2.6 2.6 p ≤ 0.5 

*Note: Average overall quality on a scale of 1-4. (1 = 

Excellent, 2 = More than adequate for diagnosis, 3 = 

Adequate for diagnosis, 4 =Nondiagnostic). 

CONCLUSION 

Two-dimensional and 3D TOF MRA with and without parallel 

imaging techniques of the carotids are comparable in 

image quality, and can be used to either reduce the time of 

the examination in 3D imaging or improve spatial resolution 

in 2D imaging, without significantly compromising image 

quality or diagnostic reliability in the qualification of carotid 

stenosis. 

KEY WORDS: MR angiography, carotid, parallel

277 


Early and Reliable Diagnosis of Intracranial Aneurysms Isolated Subdural Hemorrhage from a Middle Cerebral 

to Prevent Deadly Subarachnoid Hemorrhage 

Artery Aneurysm: A Rare Presentation of Aneurysm 

Rupture 

Medhkour, A. · Elsamaloty, H. · Herial, N. A. 

Medical College of Ohio 

Izbudak, I. · Takhtani, D. 

Toledo, OH 

Johns Hopkins School of Medicine 

Baltimore, MD 

PURPOSE 

Subarachnoid hemorrhage (SAH) with rupture of aneurysm 

is a devastating condition with about 50% mortality rate at 

first bleed. Early and reliable diagnosis at the time of warning 

leak needs to be underscored to avoid deadly outcomes. 


We present a case of SAH with a Hunt-Hess Grade I, where 

a computerized tomographic angiogram (CTA) failed to 

detect any aneurysm in the circle of Willis. After a week of 

observation, the patient was discharged from hospital in 

excellent neurologic condition. Unfortunately, the patient 

returned the next day with a massive SAH around the base 

of the brain, in the left temporal lobe and left sylvian fissure. 

The Hunt-Hess Grade on the second admission was IV and 

Fisher Grade IV. Despite aneurysm clipping and aggressive 

treatment for vasospasm, the patient’s neurologic condition 

continued to deteriorate and he finally expired few weeks 

later. 

RESULTS 

A 3D volume rendering CTA image (Fig 1A) demonstrated a 

kink above the carotid syphon. Conventional angiography 

revealed a small left posterior communicating artery 

aneurysm of about 3 mm (Fig 1B). 

CONCLUSION 

CT angiography failed to detect the aneurysm possibly due 

to the location, small size of the aneurysm, and the contrast 

bolus. These are all factors contributing to the sensitivity of 

the imaging procedure itself. CT angiography is indicated to 

be beneficial in neuroradiologic workup of subarachnoid 

hemorrhage but unfortunately, detection of aneurysms with 

< 3 mm is still suboptimal. Limitations of CTA have to be 

clearly identified to avoid fatal consequences. We present 

this case, report experiences with CTA investigations for 

aneurysms done at our institution, review the literature on 

CT angiography, and discuss its limitations. 

KEY WORDS: Subarachnoid hemorrhage, aneurysm, 

angiogram 

Dr. Elsamaloty will present the paper.. 

PURPOSE 

To share our experience with a rare subdural hemorrhage 

from a middle cerebral artery (MCA) aneurysm. To discuss 

mechanism and frequency of such a presentation. 


A 38-year-old female presented with several days history of 

worsening headache with nausea and vomiting. No history 

of trauma was provided. A CT done in the emergency room 

showed subacute isodense left subdural hematoma. A CT 

angiogram (CTA) showed right M1 segment and left MCA 

bifurcation aneurysms. Findings were confirmed on the 

catheter angiogram and in addition a small aneurysm also 

was seen in the left posterior communicating artery region. 

During the surgery no subarachnoid hemorrhage was identified. 

At surgery, few small vessels were seen entangled along 

the dome of the MCA aneurysm which necessitated couple 

of extra attempts for an eventual successful clipping. 

RESULTS 

Image on the left shows left subdural hematoma and the 

image on the right demonstrates an MCA aneurysm on the 

CTA. 

CONCLUSION 

An unusual presentation of MCA aneurysm bleeding is presented. 

Two possible theories have been proposed. First one 

relates to formation of sentinel hemorrhage which causes 

adhesions creating a tunnel through which blood enters the 

subdural compartment. Other explanation proposes tearing 

of the arachnoid membrane by the force of the bleeding. 

Clinicians should be aware of aneurysm as one of the causes 

of spontaneous subdural hematoma. 

KEY WORDS: Subdural hemorrhage, aneurysm 

Friday

Friday 

278 


An Automatic Method of Bone Removal in Brain CT Hemorrhagic Patterns in Intracranial Dural 

Angiography: 

Development 

Initial Protocol and Algorithm Arteriovenous Fistula 

Shelef, I. · Klurfan, P. · Gunnarsson,, T. · TerBrugge, K. · 

Parrish, T. B. · Sen, A. · Walker, M. T. 

Willinsky, R. 


University of Toronto 

Chicago, IL 


PURPOSE 

The purpose of this study was to implement an imaging protocol 

in conjunction with a postprocessing algorithm to eliminate 

density from bone or other tissues that remain constant 

during a contrast-enhanced CT angiography procedure without 

compromising the detection of small lesions. 


Five patients were scanned using a standard brain CTA 

acquisition protocol: kv = 120, mA = 300, thickness = 2.5 

mm, interval = 1.25 mm, contrast volume = 75 cc, delay = 

12 secs.). An identical acquisition was obtained prior to contrast 

injection. To eliminate movement-related errors, the 

data sets were aligned using a standard motion correction 

algorithm within SPM2. The precontrast data were used to 

form a mask based upon intensity thresholds that would capture 

bone (> 200 HU). Additional image processing was used 

including smoothing and a morphologic dilation operator (3 

x 3 kernel) to improve the elimination of the bone signal 

while minimizing the effects to vessel signal. Subtracting the 

precontrast data from the postcontrast data was performed to 

remove static signal. The bone mask then was applied to the 

difference data to create the final image for viewing. The 

postprocessed CTA data were transferred to a 3D workstation 

for viewing and qualitative assessment and included the 

source images and maximum intensity projections. 

RESULTS 

The imaging protocol was well tolerated by all subjects and 

postprocessing algorithm applied to all data. The motion correction 

data demonstrated minimal shift or rotation between 

exams (< 1 mm or 1 degree) in all but one subject. In that 

subject, a large amount of rotation was not correctable, 

which resulted in signal artifacts in the orbit and at the edge 

of the cortex. Bone subtraction was successful in all patients 

without qualitatively significant degradation in contrast to 

noise. 

CONCLUSION 

An automatic method to remove bone density from brain 

CTA data sets was developed and is promising. The method 

requires a pre and postcontrast CTA acquisition and thus 

increases the patients overall radiation exposure though not 

to significant levels. Additional work is required to optimize 

the algorithm to ensure vessel sharpness and to maintain routine 

CTA spatial resolution. 

KEY WORDS: CTA, bone removal, brain 

PURPOSE 

Intracranial DAVFs with leptomeningeal drainage can present 

with intracranial hemorrhage. In this study, the authors 

performed a retrospective analysis of 41 patients presenting 

with intracranial hemorrhage related to DAVFs. The type 

and location of the hemorrhage was reviewed in relationship 

to the Borden grade. 


Between 1984 and November 2004, 279 patients with 

intracranial DAVF were treated at our institution. From the 

database of the University of Toronto Brain Vascular 

Malformation Study Group, 41 patients were found to present 

with intracranial hemorrhage. Out of this group 8 (20%) 

patients had Borden type 2 DAVFs and 31 (80%) had Borden 

type 3 DAVFs. All 41 patients were studied with CT, 17 had 

MR imaging and 1 patient was studied with CTA. CT, MR 

imaging and cerebral angiograms were reviewed. The chisquare 

test was used to determine statistical significance. 

RESULTS 

The following types of intracranial hemorrhages were found: 

4 (11%) had subarachnoid hemorrhage, 7 (19%) had subdural 

hemorrhage, 8 (22%) had intraventricular hemorrhage, 

and 27 (73%) had intraparenchymal hemorrhage. A combination 

of hemorrhagic types was found in 16 (59%) patients. 

A peripheral location was found in 20 (75%) of the cases 

presenting with intraparenchymal hemorrhage. MR imaging 

or CTA detected dilated vessels in close proximity to the 

hemorrhage in 16 (89%) patients. Posterior fossa hemorrhagic 

location was observed in 12 (43%) patients. Three of 

the four SAHs presented in the posterior fossa. 

CONCLUSION 

Intraparenchymal hematoma (73.9%) is the single most 

common type of hemorrhage in patients with DAVFs while 

multiple types of hemorrhages were noted in 16 patients 

(59%). Peripheral location of the parenchymal hematoma 

(dural contact) or the presence of abnormal vessels was 

found in all patients with IPH. A DAVF should be considered 

in the presence of either of these two findings. 

KEY WORDS: DAVF


Use of Time-Resolved Echo-Shared Angiographic 

Technique Sequence in 3 T MR Imaging for Intracranial 

Vascular Lesions 

Parkinson, R. J. · Cashen, T. · Khawar, S. · Shaibani, A. · 

Hopkins, J. · Batjer, H. H. · Carroll, T. J. 


Chicago, IL 

PURPOSE 

We have used the recently developed time-resolved echoshared 

angiographic technique (TREAT) sequence with 3 T 

MR imaging (Siemens) to image intracranial vascular 

lesions such as arteriovenous malformations for comparison 

with conventional angiography. One difficulty with conventional 

digital subtraction angiography is the differentiation 

between “en passage” vessels going to adjacent brain and the 

true AVM feeding vessels, which often are branches. The 

rapid flow through an AVM makes identification of these 

important vessels difficult. 


Following informed consent for a research procedure, 

patients harboring intracranial arteriovenous malformations 

(AVMs) underwent a 3 T MR imaging study using the 

TREAT protocol. A 3D multi-phase TREAT pulse sequence 

was combined with parallel imaging (GRAPPA). The 3D 

TREAT sequence is a segmented k-space acquisition, which 

uses the TRICKS variable rate k-space acquisition. Typical 

acquisition parameters were: (spatial resolution = 1.0 x 0.8 x 

2.0 mm, temporal resolution = 3.0 s/frame, TR/TE = 3.2/1.6 

ms, flip = 15, bandwith = 1200 Hz/pixel, partial Fourier factors 

= 0.75/0.75). Gadolinium-based contrast material was 

administered as a single dose in an antecubital vein at an 

injection rate of 4.0 ml/s. 

RESULTS 

The TREAT sequence allows excellent quality 3D rendering 

of high flow intracranial vascular shunts. The rapid acquisition 

time allows differentiation between high flow feeders to the 

AVM shunt and normal vascular feeders to the adjacent brain. 

279 

CONCLUSION 

The TREAT protocol using 3D MR imaging has great promise 

for the pre and postoperative evaluation of patients with 

high flow arteriovenous shunts, allowing high-resolution 3D 

imaging and excellent temporal and spatial resolution. 

Imaging of the transit of blood through a high flow AVM has 

only been possible previously with very high frame rates on 

digital subtraction angiography. This MR imaging sequence 

offers a noninvasive method of evaluation of high-flow 

shunts in the cerebral circulation. 

KEY WORDS: MR imaging, TREAT, neurovascular 


Posterior Reversible Encephalopathy Syndrome: The 

Role of Diffusion-Weighted Imaging and MR 

Spectroscopy for Predicting Areas Which May Progress 

to Infarction 

Fatterpekar, G. M. · Delman, B. N. · Sacher, M. · Lee, B. C. 

· Naidich, T. P. 

Mount Sinai Medical Center 

New York, NY 

PURPOSE 

The posterior reversible encephalopathy syndrome (PRES) 

comprises a heterogeneous group of disorders with the common 

feature of reversible vasogenic edema within the territory 

of the posterior circulation. Rare cases progress to true 

irreversible cerebral infarction. This paper addresses imaging 

criteria for predicting which cases of PRES have greater 

likelihood to progress to infarction. 


The MR imaging studies of 16 patients with firm clinical diagnoses 

of PRES were reviewed retrospectively. These included 

5 patients with hypertensive encephalopathy, 2 with eclampsia, 

6 with immunosuppressive therapy, and 3 with uremia. 1.5 

T MR series (GE Signa, Milwaukee, WI) were obtained in 

each patient, including axial T1-weighted spin-echo, axial fast 

spin-echo T2-weighted, axial FLAIR, coronal T2*-weighted 

field echo, and axial diffusion-weighted sequences. Apparent 

diffusion coefficients (ADC) were calculated and correlated 

with the routine MR series in all patients. In 5 patients, the 

regions of signal abnormality were further evaluated by single 

voxel MR spectroscopy (MRS) (PRESS: TR 1500 ms, TE 

136-272 ms) (2 patients) and multivoxel 2D chemical shift 

imaging (3 patients). All 16 patients were followed by serial 

MR imaging for up to 4 weeks. 

RESULTS 

The full extent of the signal abnormalities was shown best by 

the FLAIR sequence in all 16 patients. The signal abnormalities 

involved the posterior circulation in all patients (100%), 

and the anterior circulation in 11 patients (69%). The lesions 

characteristically affected both the periventricular white 

matter (100%) and the subcortical white matter (100%). The 

overlying gray matter was involved in 7 patients (43%), both 

in the posterior circulation (7 of the 7 patients), and in the 

anterior circulation (4 of the 7 patients). The calculated ADC 

values identified areas of facilitated diffusion within the 

white matter in all 16 patients (100%). Apparent diffusion 

coefficient values also documented areas of restricted diffusion 

within the posterior circulation in 5 patients (32%), 

Friday

Friday 

affecting the gray matter in all 5 and the white matter in 2 of 

these 5. In those patients studied by MRS, the areas with 

facilitated diffusion showed reduced NAA with no lactate 

peak, whereas the areas of restricted diffusion showed 

reduced NAA with a prominent lactate peak. On follow-up 

MR imaging scans, the areas with facilitated diffusion 

showed recovery of normal NAA peaks, suggesting 

reversibility, whereas the areas with restricted diffusion and 

a positive lactate peak demonstrated further reduction in 

NAA, greater elevation of the lactate peak, and volume loss, 

indicating irreversibility. 

CONCLUSION 

Posterior reversible encephalopathy syndrome typically 

affects the posterior circulation (100%), but extends to 

involve the anterior circulation commonly (69%). Abnormal 

signal within the gray matter, restricted diffusion at that site, 

and the presence of a lactate peak on MRS are all indicators 

of likely progression to true cerebral infarction, with poorer 

prognosis. 

KEY WORDS: Posterior reversible encephalopathy syndrome, 

diffusion-weighted imaging, MR spectroscopy 


Optimized Voxel-Based Morphometric Analysis of Gray 

and White Matter in Drug-Naive Patients with 

Schizophrenia 

Jayakumar, P. N. 1 · Gangadhar, B. N. 1 · Desai, S. V. 1 · 

Venkatasubramanian, G. 1 · Srinivas, J. S. 1 · Keshavan, M. S. 2 

1 National Institute of Mental Health and Neurosciences, 

Bangalore, INDIA, 2 Wayne State University School of 

Medicine, Detroit, MI 

PURPOSE 

MR imaging studies in drug-naive schizophrenia patients 

using manual region of interest (ROI) method have reported 

inconsistent findings. Optimized voxel-based morphometry 

(OVBM) is an automated image-analysis technique that offers 

a rapid, sensitive, and unbiased whole brain survey. We used 

OVBM to examine gray matter (GM) and white matter (WM) 

volume abnormalities in first-episode, drug-naive schizophrenia 

patients without confounds of neuroleptic treatment. 


There were 52 drug-naive schizophrenia patients (SCID, 

DSM-IV) (age range:17-42 years, male/female = 31/21) and 

43 age-, gender- and education-matched healthy subjects 

(HS) (age range: 21-47 years, male/female = 33/10). The 

first episode and illness duration (range: 6-240 months) as 

defined by report of psychotic symptoms were assessed. 

Healthy subjects were screened using the 12-item General 

Health Questionnaire. Nineteen patients had a mean follow 

up of 1.2 years. Data were obtained on a 3D high-resolution 

T1-weighted MPR imager and analyzed with SPM99. 

Morphometric analyses of gray and white matter was done 

using PET/SPECT model. Group comparisons for regional 

GM and WM volume differences were performed using single 

subject conditions and covariates analysis within the 

framework of general linear model in SPM2 with age and 

sex as nuisance covariates. Statistical parametric maps were 

constructed to test for morphologic differences between 

patients and HS. Regional GM and WM volume differences 

280 

between the patients and HS were assessed on a voxel-byvoxel 

basis with the threshold set at p ≤0.05 (corrected). To 

study the possible relationship between the duration of illness 

(DOI) and morphologic brain changes, the patients 

were grouped into those with less than 1 year ( n = 17) and 

5 years or longer (n = 15) of DOI and compared to matched 

HS (n = 23 and 16 each). 

RESULTS 

Baseline studies in patients showed decreased gray matter 

volume in bilateral internal frontal, bilateral insula and right 

superior temporal gyrus and decreased white matter volume 

in genu and left internal frontal region compared to HS. 

Nineteen patients who had 1- to 2-year follow up did not 

show GM or WM differences. Patients with shorter DOI 

showed reduced GM volume in left insula and reduced WM 

volume in splenium. Patients with longer DOI showed 

reduced GM volume in the right visual association area and 

reduced WM volume in genu and bilateral frontal regions. 

Comparison of patients with short and long DOI showed 

reduced volume of right parieto-occipital visual association 

area and superior vermis. The two groups of HS showed no 

differences in any region. 

CONCLUSION 

Significant GM volume deficits in internal frontal, superior 

temporal gyrus, insulae support heteromodal association cortical 

abnormalities in schizophrenia. The corpus callosum was 

abnormally smaller in patients; while the splenium was smaller 

in patients with shorter DOI, the genu was smaller in 

patients with longer DOI. Collectively, these findings suggest 

limbic, heteromodal cortical and cerebellar abnormalities in 

schizophrenia. Morphologic differences between patients with 

shorter and longer DOI suggest illness-related pathology. That 

1 year follow up in the same patient group failed to detect this 

may point to an insidious nature of the process. 

KEY WORDS: Voxel-based morphometry, schizophrenia, statistical 

parametric mapping 

Acknowledgment: This work was supported by Indo-US 

project 1 R43 04370-01A1 1997-2001. 


Prevalence of Cavum Septi Pullicidi in Boxers: Is It a 

Sign of Damage? 

Aviv, R. I. 1 · Jafri, Z. 2 · Valentine, A. 3 · Thakkar, C. 4 · Kendall, 

B. E. 3 

1 Sunnybrook Hospital, Toronto, ON, CANADA, 2 Charing 

Cross Hospital, London, UNITED KINGDOM, 3 Royal Free 

Hospital, London, UNITED KINGDOM, 4 Royal London 

Hospital, London, UNITED KINGDOM 

PURPOSE 

British professional boxers require an annual MR scan to 

maintain registration. The presence of the cavum septi pellucidi 

(CSP) is of uncertain and controversial significance. A 

cavum has been cited as a marker of previous head injury in 

these and other athletes engaged in sports involving heavy 

physical contact. The prevalence of CSP on MR imaging 

amongst the general population is unknown. Debate as to

whether there is a higher prevalence in boxers continues and 

whether these professional sportsmen should be allowed to 

continue boxing if a CSP is present. 


We evaluated serial scans of 164 consecutive boxers and 

compared them with an age-matched cohort of control 

patients who underwent MR imaging. The presence, 

absence, size and extent of CSP was evaluated in addition to 

any parenchymal abnormality and Evan’s ratio. The extent 

was divided into type 1 anterior to the fornix, type 2 extending 

up to the fornix, and type 3 extending into cavum vergae. 

RESULTS 

One hundred and sixty-four boxers underwent 273 scan, 

median number of scans was 1 (range 1 to 5). Forty-three 

control patients each had a single scan. There was no difference 

between the control group and boxers for age. A CSP 

was present in 49.5% of boxers and 39.5% of control 

patients (p = 0.7). There was no significant difference 

between the two groups for coronal dimension. Boxers were 

significantly more likely to have a greater extent of CSP than 

controls (p < 0.05). There was no increase in CSP prevalence 

with number of scans, indicating duration of boxing career. 

There was a nonsignificant trend in boxers to increased 

prevalence of CSP in the < 20 and 20-29-year-old age 

groups. The prevalence of cavum was higher in boxers 

between the ages of 20-29 years than controls. Three boxers 

increased the extent of their CSP over successive scan. Two 

increased from a type 1 to type 3, while one with a type 2 

developed a type 3. 

CONCLUSION 

We have found a much higher prevalence of CSP than previously 

reported indicative of improved MR imaging detection 

compared with CT. The prevalence was not higher in boxers 

than normal controls; however boxers were more likely to 

have a greater posterior extent of cavum. Boxers have a nonsignificant 

trend to increased cavum prevalence when compared 

with controls in the 20-29-year-old age group. We conclude 

that the finding of an anterior or type 1 CSP is not of any 

significance; however a greater extent was not seen amongst 

our control population and may be a significant finding. 

KEY WORDS: Cavum septi pellucidi, boxing, MR imaging 


Contrast-Enhanced FLAIR Imaging for Infectious 

Leptomeningeal Diseases 

Parmar, H. A. 1,2 · Anand, P. 2 · Hui, F. 2 · Sitoh, Y. 2 

1 Hospital for Sick Children, Toronto, ON, CANADA, 

2 National Neuroscience Institute, Singapore, SINGAPORE 

PURPOSE 

Infectious meningitis is a serious disease with high potential 

for permanent neurologic impairment and high mortality 

rates. Hence, prompt diagnosis and early treatment is essential. 

MR imaging, especially contrast-enhanced T1-weighted 

imaging, is regarded as a reliable technique for confirming 

the diagnosis. FLAIR imaging, which yields highly sensitive 

images of several brain parenchymal lesions, has been 

reported to be a promising technique for the diagnosis of 

several extraaxial diseases, including subarachnoid hemor- 

281 

rhage and leptomeningeal carcinomatosis. The purpose of 

our study was to compare contrast-enhanced FLAIR images 

with contrast-enhanced T1-weighted images for infectious 

leptomeningitis. 


From September 2003 to June 2004, we studied 24 patients 

with clinical suspicion of infectious meningitis on a 1.5 T 

system. All MR imaging studies included unenhanced 

FLAIR and contrast-enhanced T1-weighted and FLAIR 

sequences. The timing of the contrast-enhanced T1-weighted 

and FLAIR sequences was random; in twelve studies the 

FLAIR images were obtained before the T1-weighted 

images. No supplemental oxygen was administered to any 

patient during their imaging studies. Twelve patients had 

cytologic and biochemical diagnosis of meningitis on cerebrospinal 

fluid examination obtained 48 hours before or after 

the MR study. The unenhanced FLAIR, contrast-enhanced 

FLAIR, and contrast-enhanced T1-weighted axial view 

images were reviewed separately by two neuroradiologists 

who were blinded to patient identity and clinical outcome. A 

positive rating was given when signal abnormality or 

enhancement was present in the subarachnoid space (cisterns 

and sulci) or along any pial surface or cranial nerves. 

RESULTS 

A total of 27 examinations for 24 patients (three patients 

underwent imaging twice) were performed. The mean 

patient age was 23 years (age range, 3-78 years; 15 female 

and 9 male patients). Of the twelve patients (13 studies) in 

which cytology was positive, unenhanced FLAIR images 

were positive in six cases (sensitivity 46%), contrastenhanced 

FLAIR images were positive in 111 (sensitivity 

84%), and contrast-enhanced T1-weighted MR images were 

positive in 11 patients (sensitivity 84%). Of these 13 studies, 

in three studies both contrast-enhanced FLAIR and T1weighted 

images were rated equal. Contrast-enhanced 

FLAIR was considered better in six, while T1-weighted 

images were rated better in three studies (parenchymal 

abnormalities were seen in all three studies). In one study 

both sequences were negative. Of the 12 patients (14 studies) 

in whom cerebrospinal fluid study was negative, unenhanced 

FLAIR images were negative in 13, contrast-enhanced 

FLAIR images were negative in 10, and contrast-enhanced 

T1-weighted MR images were negative in six. Thus, the 

specificity of unenhanced FLAIR, contrast-enhanced FLAIR 

and contrast-enhanced T1-weighted images was 93%, 79%, 

and 43% respectively. 

CONCLUSION 

Our results suggest that postcontrast FLAIR images have 

similar sensitivity but a higher specificity compared to contrast-enhanced 

T1-weighted images for detection of subtle 

leptomeningeal enhancement. A postcontrast FLAIR 

sequence is therefore a valuable adjunct to postcontrast T1weighted 

imaging in evaluation of patients with infective 

leptomeningitis. Contrast-enhanced T1-weighted sequence 

is more sensitive when associated parenchymal abnormality 

is present. 

KEY WORDS: Leptomeningitis, FLAIR imaging 

Friday

Friday 


A Wolf in Sheep’s Clothing: Neurologic or 

Neurosurgical? 

Prescod, D. 1 · Renowden, S. 1 · Ramnarine, D. 2 

1Frenchay Hospital, Bristol, UNITED KINGDOM, 

2Derriford Hospital, Plymouth, UNITED KINGDOM 

PURPOSE 

A young lady of leisure presents to a neurosugical service 

with possible history of physical trauma. Initial investigations 

reveal hydrocephalus, subarachnoid hemorrhage, a 

subdural and eventually fatal intracranial hemorrhage. All 

the events were investigated serially and treated; including 

biposy and consultant neurologic review, but no cause for 

neurologic decline is found premorbidly. 


A retrospective look at the clinical scenarios that are presented 

and treated as neurosurgical case. The spectrum of 

neurosurgical manifestations were disguising an inflammatory 

neurologic condition. 

RESULTS 

The clinical progression of the case is supported with the 

corresponding radiologic findings which are not typical for 

the disease presentation. 

CONCLUSION 

Acute hemorrhagic leukoencephalopathy has a variety of 

radiologic presentations, of which the combined imaging of 

the present case has not been recorded previously in the literature. 

Even in the face of a series of neurosurgical problems, 

a neurologic differential always should be considered. 

KEY WORDS: Leukoencephalopathy, hemorrhagic, 

encephalomyelitis 


10:00 AM - 11:30 AM 

Theatre 

(76) Review of Pediatric Head, Neck, 

and Spine (ASPNR) 

(511) Head and Neck Imaging 

— Caroline D. Robson, MD, ChB 

(512) Spine Imaging 

Moderator: Patrick D. Barnes, MD 

— Erin M. Simon, MD 

282 

Head and Neck Imaging 

Caroline D. Robson, MB, ChB 



able to: 

1) Review the imaging appearance of common or important 

head and neck abnormalities. 

2) Generate a differential diagnosis for these conditions. 

3) Provide imaging algorithms for head and neck disorders. 

4) Review some of the more common genetic syndromes 

associated with head and neck malformations. 


Imaging protocols for head and neck imaging will be 

reviewed briefly. The characteristic imaging appearance for 

a variety of common and/or important head and neck 

processes will be reviewed. A wide range of case illustrations 

will be discussed and these include: a) developmental 

disorders: thyroglossal duct cyst, branchial cleft cyst, ectopic 

thyroid, ectopic thymus, dermoid cyst, cephalocele, choanal 

atresia, cleft lip/palate, micrognathia; b) acute and chronic 

infections of the neck: epiglottitis, retropharyngeal, and peritonsillar 

abscesses; lymphadenitis; sinusitis and complications, 

and mastoiditis and complications; c) benign neck 

masses: vascular anomalies, teratoma, nerve sheath tumor, 

juvenile nasopharygeal angiofibroma and d) malignant neck 

masses rhabdomyosarcoma, lymphoma. The differential 

diagnosis for each example will be illustrated and discussed. 

Where relevant, common predisposing genetic disorders will 

be mentioned and, in some instances, fetal imaging also wil 

be demonstrated. 

REFERENCES 

1. Robson CD. Cysts and tumors of the oral cavity, oropharynx 

and nasopharynx in children. Neuroimag Clin N Am 

2003;3:427-442 

2. Koch, BL. Imaging extracranial masses of the pediatric head 

and neck. Neuroimag Clin N Am 2000;10(1):193-214, ix 

3. Robson CD, Barnewolt CE. Magnetic resonance imaging of 

fetal head and neck anomalies. Neuroimag Clin N Am 

2004;14(2):273-291 

Spine Imaging 

Erin M. Simon, MD


10:00 AM - 11:30 AM 

Room 105/106 

(77) Interventional Spine (ASSR) 

(513) Vertebral Body Interventions: An Update 

— Bassem A. Georgy, MD 

(514) Percutaneous Disk Interventions 

— Kurt P. Schellhas, MD 

(515) Nerve Injections and Rhizotomies 

— Blake A. Johnson, MD 

Moderators: Bassem A. Georgy, MD 


Vertebral Body Interventions: An Update 

Bassem A. Georgy, MD 



able to: 

1) Review anatomical and technical considerations for vertebral 

body interventions. 

2) Present an update of the recent advances in vertebroplasty 

and kyphoplasty. 

3) Discuss safety issues and how to avoid complications. 


Vertebral body interventions cover a large variety of procedures 

ranging from simple percutaneous biopsy to complicated 

intraarterial tumor embolization. However, percutaneous 

cement injection (Vertebroplasty and Kyphoplasty) 

presents the majority of procedures done in any radiology 

department. Cement injection is a relatively safe procedure 

with a very good success rate and low complication rates. In 

this presentation an overview of the recent advances in performing 

vertebroplasty and kyphoplasty will be reviewed. 

Different techniques to avoid complications will be discussed. 

The author will share his experience in dealing with 

certain difficult situations. Different types of needles and 

delivery devices are now available in the market. An 

overview of these delivery systems and injection devices 

will be presented. The use of newly available tools like unidirectional 

balloons and curettes (by Kyphon Inc., 

Sunnyvale, CA) as well as the Sky Bone Expander system 

(Disc Orthopedic Technologies Inc., Monroe Township, NJ) 

in certain situations will be discussed. Review of different 

needle placement techniques; transpedicular, para-pedicular 

283 

or postrolateral approaches also will be presented. The 

author also will present in his own experience certain situation 

where percutaneous injection of cement in malignant 

lesions is performed for stabilization of the fractures as well 

as for pain control. A biopsy also can be performed in the 

same time. Metastatic lesions are more technically demanding 

than benign osteoporotic fractures. More complications 

are expected due to the complexity of the lesions in particular 

if the posterior cortex is involved. The author will present 

his experience in performing radiofrequency coblation in 

combination with cement injection in certain cases to 

decrease the likelihood of complications. The benefits of 

performing myelograms before cement injection also will be 

discussed. Although vertebroplasty and kyphoplasty are relatively 

safe procedures, the potential for devastating complication 

always exist. We also will discuss ways to avoid such 

complications and optimizing the procedure techniques. A 

suggested approach for some difficult situations like vertebra 

planna and multiple levels treatment also will be presented. 

REFERENCES 

1. Mathis JM, Barr JD, Belkoff SM, Barr MS, Jensen ME, 

Deramond H. Percutaneous vertebroplasty: a developing 

standard of care for vertebral compression fractures. AJNR 

Am J Neuroradiol 2001;22(2):373-381 

2. Mathis JM. Percutaneous vertebroplasty: complication 

avoidance and technique optimization. AJNR Am J 

Neuroradiol 2003;24(8):1697-706 

Percutaneous Disk Interventions 

Kurt P. Schellhas, MD 



able to: 

1) Review technique(s) for gaining access to the disk(s) to be 

treated. 

2) Compare morbidity and efficacy of the various procedures. 


Various percutaneous radiologic spinal disk interventions 

have evolved over the years. These include intradiskal injections 

of therapeutic substances, such as local anesthetics and 

steroids, chemonucleolysis, intradiskal prolotherapy, 

intradiskal electrothermal therapy (IDET), percutaneous 

diskectomy/nucliectomy, and nucleoplasty procedures. All 

of these procedures are based upon fundamental techniques 

used for spinal diskography, whereby access to the disk and 

nuclear space in particular is achieved. Our clinical experience 

has been primarily with intradiskal steroids and nucleoplasty. 

A short review of lumbar, thoracic and cervical 

diskography technique will be discussed, as this discussion 

is fundamental to gaining access to the nuclear space of the 

desired disk(s). Furthermore, diskography is required prior 

to most of the therapeutic injections. Our extensive, 14-year 

experience with intradiskal steroid (and anesthetic) injections 

will be presented, as will our more recent experience 

with nucleoplasty. In our practice, thoracic nucleoplasty is 

the more frequent procedure (compared to lumbar) which we 

have been asked to perform. The advantages of nucleoplasty 

compared with IDET, percutaneous diskectomy, and major 

surgical procedures will be emphasized. 

Friday

Friday 

Nerve Injections and Rhizotomies 

Blake A. Johnson, MD 



able to: 

1) Review imaging and anatomical considerations for performing 

nerve blocks and rhizotomies. 

2) Review risks and potential complications associated with 

spine injection procedures. 

3) Review technique and equipment requirements for performing 

nerve blocks, median branch blocks, and RF rhizotomies. 


Minimally invasive needle procedures provide an important 

contribution to the management of patients with neck, back, 

or radicular pain symptoms. Although imaging provides 

invaluable morphologic information, the additional data provided 

by these special spine procedures through provocative 

and palliative feedback improve diagnostic accuracy for pain 

syndromes of spinal origin. In addition to providing diagnostic 

information, therapeutic spine interventions provide a 

minimally invasive method for treating affected patients. 

Selective nerve blocks are performed as both diagnostic and 

therapeutic procedures. The needle is placed within the lateral 

aspect of the neural foramen along the course of the exiting 

nerve root. Contrast (0.5-1.5 cc) is injected after needle 

placement to document accurate positioning within the nerve 

sheath. It also is important to exclude venous opacification 

which will decrease the efficacy of the injection and potentially 

induce side effects. Significant epidural reflux should 

be avoided. After filming, a therapeutic mixture consisting of 

two parts local anesthetic and one part steroid mixture is 

injected. Generally, 1.2-1.8 cc is an adequate volume to prevent 

significant epidural reflux and achieve sufficient nerve 

root anesthesia. In addition to facet joint blocks, the facet 

nerve (medial branch of the dorsal ramus) can be anesthetized. 

A facet nerve block is preferred to a joint block for 

patients who are candidates for radiofrequency neurotomy 

procedures. For this procedure, the needle is placed along 

normal course of the facet nerve. In the cervical spine, a 25gauge 

spinal needle is advanced to the center (waist) of the 

articular pillar. In the lumbar spine, a 22-gauge spinal needle 

is placed at the junction of the superior articular process and 

the transverse process. After needle placement, a small dose 

of contrast (.3-.5 cc) is injected to exclude venous opacification. 

Following injection of contrast and filming in at least 2 

planes, 0.5-1.0 cc of local anesthetic are injected. Patient is 

monitored thereafter for response and assessed at 30 minutes 

postinjection. Electrode placement for the RF neurotomy 

procedure is similar to that for the facet nerve block. Motor 

stimulation testing is performed to confirm accurate placement 

on the nerve, and to exclude contact with the motor 

nerve (or ganglion). Contact with the bony landmark is 

important, as the RF lesion profile is small, requiring direct 

contact with the nerve in order to reliably coagulate the 

nerve fibers. The trajectory of the electrode should parallel 

the course of the nerve for the same reason. Treatment of the 

cervical facet nerves is performed using RF energy for 60 

seconds at 80°C. In the lumbar spine, a slightly higher temperature 

(87°) is utilized for 60 seconds. One, two, or three 

lesions are performed at each level, depending on the electrode 

used and other individual factors. In this session, 

284 

patient selection, testing, and treatment will be discussed, 

with an emphasis on technical aspects and radiographic 

anatomy. 


11:30 AM - 11:45 AM 

Theatre 

Closing Remarks/Adjournment 

— Patricia A. Hudgins, MD, ASNR President

Scientific Posters 1-184 

Exhibit Hall B 


12:00 PM – 9:00 PM 

Tuesday, May 24 - Thursday, May 26 

6:30 AM - 9:00 PM 


6:30 AM – 11:45 AM 

A missing Scientific Poster number indicates an 

abstract has been withdrawn. 

Adult Brain 

1-81 

Poster 1 

CT Angiographic Analysis of Carotid Artery Stenosis 

Silvennoinen, H. M. · Ikonen, S. · Soinne, L. · Railo, M. · 

Valanne, L. K. 

Helsinki University Hospital 

Helsinki, FINLAND 

PURPOSE 

The aim of our study was to compare multislice helical CT 

angiography (CTA) to intraarterial digital subtraction 

angiography (DSA) in detection and grading of carotid 

artery bifurcation stenosis. In CTA we used a novel commercial 

vessel analysis (VA) software in addition to axial 

source images and minimal intensity projection (MIP) 

images. 


Thirty-seven consecutive consenting patients (73 vessels, one 

carotid artery catheterization was unsuccessful) with ultrasound 

evidence of a marked internal carotid artery (ICA) 

stenosis were enrolled in the study, undergoing carotid CTA 

and DSA within 2 weeks. CT angiography was performed 

with a GE LightspeedUltra 8-row scanner. The slice thickness 

was 2.5 mm, pitch 1.35:1. One hundred ml of iodinated contrast 

medium 300 mgI/ml was injected at a rate of 4 ml/sec, 

2.5 mm axial raw data was reformatted into 1.25 mm axial 

scans. The grade of stenosis was determined using axial 

source images, MIP, and VA. Vessel analysis provides intraluminar 

diameter measurements and cross-sectional areas of 

285 





carotid artery at selected anatomical points, calculating the 

maximum stenosis in percents in proportion to the chosen reference 

point. DS angiography was performed with selective 

catheterization of both common carotid arteries, obtaining 

three or four projections from each bifurcation. CT angiography 

and DSA were analyzed by two neuroradiologists blinded 

to the results of the other modality and the clinical data. The 

degree of ICA stenosis was graded according to NASCET. 

RESULTS 

Four carotids were occluded (5 %), 20 (27%) had a high-grade 

stenosis (70-99%), and 16 (22%) a moderate-grade (50-69%) 

stenosis. There were 16 (22%) mild (< 50%) stenoses, and no 

stenosis was detected in 17 (23%). Degree of stenosis in CTA 

was correlated closely to those in DSA (R = 0.95, regression 

analysis), and with VA method it was slightly lower (R = 0.89; 

p = .02 for paired comparison with DSA). CT angiography 

underestimated high-grade and moderate stenosis in comparison 

to DSA (78.2% vs 86.4%, 57.3% vs 63.1%, p < .05, 

Wilcoxon matched pairs test). Underestimation was greater 

with VA method (68.5% vs 83.5%, 51.8% vs 63.1%, p < .05). 

Intermethod agreement of grading was high (Cohen’s weighted 

κ = 0.93), and slightly lower for VA (weighted κ = 0.88). 

CT angiography detected all four occlusions. For a high-grade 

stenosis, sensitivity of CTA was 0.75 (95% confidence interval 

0.51-0.91) and specificity 0.96 (0.87-1.00). For moderate 

stenosis, the sensitivity was 0.88 (0.61-0.98) and specificity 

0.82 (0.69-0.92). For VA method, sensitivity and specificity 

were lower, 0.47 (0.21-0.73) and 0.96 (0.87-1.00) for highgrade 

stenosis, and for moderate stenosis 0.62 (0.35-0.85) and 

0.82 (0.69-0.92), respectively. 

CONCLUSION 

CT angiography estimate of carotid stenosis is highly correlated 

to that in standard conventional DSA, and may replace 

it in many instances as a less invasive method, although it 

tends to underestimate clinically relevant grades of stenosis. 

Its accuracy is not improved by semiautomated analysis. The 

data support the use of CTA in confirming carotid occlusion. 

KEY WORDS: CT angiography, carotid artery stenosis 

Poster 2 

Evolution of Apparent Diffusion Coefficient and 

Cerebral Perfusion in a Real-Time Endovascular MR 

Imaging-Guided Swine Stroke Model 

Lin, E. 1 · Feng, L. 2 · Dashnaw, S. 1 · Zhang, H. 1 · Oklu, R. 1 · 

Baytion, M. 1 · Gupta, G. 1 · Kim, T. 1 · Pile-Spellman, J. 1 

1Columbia University Medical Center, New York, NY, 

2University of California Los Angeles, Los Angeles, CA 

PURPOSE 

To demonstrate the feasibility of serial diffusion-weighted 

imaging (DWI) and perfusion-weighted imaging (PWI) in a 

novel real-time endovascular MR-guided swine stroke 

Posters

Posters 

model and to evaluate the evolution and relationship of 

apparent diffusion coefficient (ADC) with cerebral perfusion. 


Three endovascular methods of inducing cerebral ischemia 

using only MR guidance were studied in six domestic pigs 

utilizing two pigs for each method. Trans-femoral catheterization 

with active MR-tracking catheters and guide wires 

was performed. The three methods were: balloon catheter 

occlusion of the brachiocephalic artery; the left subclavian 

and bilateral carotids; the left common carotid and left subclavian 

artery with induced systemic hypotension. Serial 

DWI and PWI were performed for a period of 197 ± 41 minutes 

immediately after occlusion. 

RESULTS 

The brachiocephalic occlusion method induced a global pattern 

of DWI lesions. At 2 hours, the ADC ratio (lesion/baseline) 

was reduced to 0.61 ± 0.04 in the temporal lobes, basal 

ganglia, cerebellum, and high watershed region. No cerebral 

blood flow (CBF) ratio (lesion/baseline) could be determined 

as CBF was not detectable after occlusion. The left 

subclavian and bilateral carotid occlusion method produced 

DWI lesions in the high watershed regions and less severely 

the cerebellum in one pig and a global pattern that spared the 

cerebellum in the other. At 2 hours, the ADC ratio was 

reduced to 0.62 ± 0.13 and the CBF ratio was reduced to 0.38 

± 0.12. The left common carotid and left subclavian occlusion 

hypotension method produced DWI lesions in the left 

temporal lobe in one pig and lesions in the left temporal, 

cerebellum, and high watershed regions in the other. At 2 

hours, the ADC ratio and CBF ratios were reduced to 0.61 ± 

0.02 and 0.42 ± 0.11 respectively. Diffusion-weighted 

lesions corresponded well to the lesion area on histologic 

sections. 

CONCLUSION 

This highly malleable stroke model provides immediate serial 

information of cerebral perfusion and ADC changes after 

induction of cerebral ischemia and may be helpful in the elucidation 

of the evolution of stroke and interventions. 

KEY WORDS: Interventional MR imaging, stroke, model 

286 

Poster 3 

Quantitative Flow-Sensitive Alternating Inversion 

Recovery Imaging Using 3.0 T MR Scanner in Patients 

with Major Cerebral Artery Occlusive Disease: 

Comparison with Single Photon Emission CT 

Inoue, T. · Ogasawara, K. · Ogawa, A. 

Iwate Medical University 

Morioka, JAPAN 

PURPOSE 

To measure cerebral blood flow (CBF) is important to decide 

the strategy of treatment in major cerebral artery occlusive 

disease. Flow-sensitive alternating inversion recovery 

(FAIR) imaging has been developed recently for noninvasive 

MR perfusion technique. The aim of the present study was to 

clarify the usefulness of quantitative FAIR imaging using 3.0 

T MR scanner in patients with major cerebral artery occlusive 

disease. 


Twenty patients with unilateral major cerebral artery occlusive 

disease underwent FAIR and single photon emission CT 

(SPECT). Conventional MR imaging showed no obvious 

infarct lesion in the bilateral middle cerebral artery (MCA) 

territories. Both imagings were performed at least 1 month 

after the onset of symptoms in all patients. The study protocol 

was approved by the ethical committee. All subjects gave 

written informed consent prior to the study. All MR scans 

were performed using Signa VH/i 3.0 T MR imaging system 

(GE Medical Systems, Milwaukee, WI) and standard head 

coil. A spin-echo type echo-planar imaging sequence was 

used for FAIR imaging: repetition time 2000 ms, echo time 

20 ms, matrix 64 x 64, FOV 240 mm, inversion time 1200 

ms, 6 mm thickness and 2 mm gap. The regions of interest 

were set on the bilateral MCA territories and cerebral blood 

flow (CBF) was calculated for each patient. Correlations of 

between FAIR and SPECT imaging were determined by linear 

regression analysis and by computing regression equations 

and correlation coefficient. Statistical significant was 

set at the p < 0.05. 

RESULTS 

The CBF obtained from FAIR and SPECT imagings were 

varied from 15 to 45 (ml/100 g/min) and 25 to 52 (ml/100 

g/min) respectively. Representative FAIR and SPECT 

images were shown in Fig. 1. Cerebral blood flow from 

FAIR was significantly correlated with those from SPECT 

images (correlation coefficient was 0.67 and p < 0.05). 

Fig. 1. Sixty-four-year-old man with left internal carotid 

artery occlusion. FAIR image showing the reduction of CBF 

in left hemisphere the same as SPECT image.

CONCLUSION 

A significant reduction of CBF in patients with unilateral 

major cerebral artery occlusive disease could be observed 

using FAIR imaging. This technique can measure the CBF 

without the risk and expense of exogenous tracers. Because 

the risks of this technique are essentially those of routine MR 

imaging, one may perform repeated studies of the same individual 

patient. 

KEY WORDS: Major cerebral artery occlusive disease, MR 

imaging, cerebral blood flow 

Poster 4 

Analysis of Vertebral Arteries: Correlation between 

Color Doppler and Angiography 

Kizilkilic, O. 1 · Hurcan, C. 1 · Mihmanli, I. 2 · Oguzkurt, L. 1 · 

Yildirim, T. 1 · Tercan, F. 1 

1 Baskent University Adana Teaching and Medical Research 

Center, Adana, TURKEY, 2 Istanbul University Cerrahpasa 

Medical School, Istanbul, TURKEY 

PURPOSE 

Color Doppler ultrasonography is the most widespread diagnostic 

procedure in obstructive disease of the arteries supplying 

the brain. To our knowledge, there are only a few correlative 

color Doppler ultrasonographic and angiographic 

studies of the vertebral arteries, especially in patients who 

have flow-restrictive stenosis at the carotid bifurcation. The 

purpose of this study was to evaluate diameter, flow volume, 

and time-averaged mean velocities of angiographically verified 

normal vertebral arteries without collateral flow. 


One hundred fifty patients referred for carotid angiography 

with a normal vertebrobasilar system and with no patent posterior 

communicating arteries were investigated with color 

Doppler ultrasonography. Luminal diameter, time-averaged 

mean velocity, peak systolic velocity, and flow volume values 

were calculated for each patient. The parameters were 

compared between sexes, in different age groups, in patients 

with carotid stenosis of 50% or less and greater than 50%, 

and in patients who had clinical signs of vertebrobasilar 

insufficiency versus those who had not. 

RESULTS 

We have found no significant difference in evaluated parameters 

with the degree of associated carotid stenosis or vertebrobasilar 

insufficiency. Diameter and flow volume values 

on the left side were higher than on the right side. 

CONCLUSION 

We found similar flow volume values of vertebral arteries in 

both sexes and both groups of patients with carotid stenosis 

of 50% or less and greater than 50%. All parameters did not 

differ in patients with stenosis of 50% or less and greater 

than 50% and in patients with and without clinical signs of 

vertebrobasilar insufficiency. 

KEY WORDS: Vertebral artery, color Doppler ultrasound, 

angiography 

287 

Poster 5 

Perfusion CT and Carotid Surgery: A Prospective 

Hemodynamic Study in 12 Patients with Asymptomatic 

Stenosis 

Pasco, A. 1 · Papon, X. 1 · Darabi, D. 1 · Lemaire, L. 2 · Mas- 

Caradec, M. 1 · Marc, G. 1 · Ter Minassian, A. 1 · LeJeune, J. 2 

1Hopital Larrey - Centre Hospitalier Universitaire, Angers, 

FRANCE, 2INSERM U 646, Angers, FRANCE 

PURPOSE 

Surgery of asymptomatic carotid surgery (ACS) is still based 

on morphologic criteria (degree of stenosis). Perfusion CT 

(PCT) now allows an hemodynamic approach of these 

lesions but this technique is not currently used in chronic 

cerebrovascular disease. The purpose of this study was to 

describe the brain hemodynamic parameters with PCT 

before and after surgery of ACS. 


Twelve patients with unilateral ACS were prospectively 

studied. Perfusion CT (Philips Mx8000) were performed 

before (mean 10 days) and after surgery (mean 10 days). 

Arterial input function was determined by placing manually 

a region of interest (ROI) in contralateral to stenosis carotid 

artery or in the proximal segment of its main branches, perpendicular 

to transversal plane section. Venous outflow 

function was defined after manually placing a ROI in the distal 

segment of the superior sagittal sinus. Several ROI locations 

were examined to obtain the best arterial and venous 

time-attenuation curves. Regional ROIs were placed in main 

hemispheric arterial territories: anterior cerebral, superficial 

and deep middle cerebral, superficial and deep posterior 

cerebral territories. Contralateral mirror ROIs were placed 

automatically. For each ROI, mean transit time (MTT), time 

to peak (TTP) and cerebral blood volume (CBV) were registered 

with deconvolution-based algorithm of functional PCT 

Philips software. Two groups of patients were defined by the 

severity of stenosis; group 1: 70-85% stenosis (n = 7) and 

group 2: ≥ 85 % stenosis (n = 5). Hemodynamic parameters 

were compared between both affected and nonaffected side 

and pre and postoperative status. 

RESULTS 

Qualitative analysis of perfusion maps revealed no difference 

between each hemisphere before and after surgery. 

Semiquantitative analysis showed that before surgery, mean 

MTT and TPP in affected compared with nonaffected side, 

were significantly delayed (4.53 ± 1.66 versus 3.54 ± 1.33, p 

< 0.0001 / 12.73 ± 2.42 versus 11.74 ± 2.25, p < 0.0001 

respectively) in the superficial middle cerebral artery (MCA) 

territory in both groups without any correlation with severity 

of stenosis. Statistically significant difference was found 

between group 1 and 2 mean CBV values (2.60 ± 1.17 versus 

1.64 ± 1.02, p < 0.0001) in ipsilateral anterior cerebral 

artery (ACA) territory. After surgery, changes in mean MTT, 

TTP, and CBV were not homogeneous between patients and 

groups, but were symmetrical for each patient in both hemispheres. 

Comparing pre and postoperative findings, patients 

of group 2 had an ipsilateral increase of mean CBV prevailing 

in middle cerebral arterial (MCA) territory. On the contrary, 

all patients of group 1 had a decrease of these same 

values. 

Posters

Posters 

CONCLUSION 

Delay of preoperative mean TTP and MTT in superficial 

MCA territory is noticed in all patients with an ipsilateral 

≥70% ACS. Very low preoperative CBV mean value in ACA 

territory could identify patient with high-grade stenosis. 

Changes in CBV mean values after surgery could help to 

assess the efficiency of revascularization. More larger studies 

are necessary to confirm these results. 

KEY WORDS: Perfusion CT, carotid stenosis, asymptomatic 

Poster 6 

Dynamic Susceptibility Contrast MR Imaging in 

Unilateral Severe Internal Carotid Artery Stenosis before 

and after Endoarterectomy: Evaluation of Middle 

Cerebral Artery and Border Territories 

Gaudiello, F. · Garaci, F. G. · Marziali, S. · Ludovici, A. · 

Melis, M. · Floris, R. · Simonetti, G. 

University of Rome Tor Vergata 

Rome, ITALY 

PURPOSE 

To assess hemodynamic modifications in symptomatic 

patients with unilateral stenosis of the internal carotid artery 

(ICA) with perfusion-weighted MR imaging and compare 

these data with those obtained after endoarterectomy. 


Fifteen patients with unilateral 70-90% internal carotid 

artery stenosis were studied with perfusion MR imaging. 

Patients enrolled in this study had a normal contralateral 

internal carotid artery (stenosis < 60%). Fifteen healthy 

age/sex-matched subjects were enrolled as controls. 

Regional cerebral blood volume (rCBV) and mean transit 

time (MTT) values were calculated bilaterally in the middle 

cerebral artery (MCA) and borderzone (BZ) territories. All 

patients underwent endoarterectomy within 1 week after MR 

scans and were reexamined 1 month after surgery. 

RESULTS 

Statistically significant differences were noted in MTT values 

in the borderzones territories between patients (mean: 

0.95 ) and controls (mean: 1.13). No significant differences 

in rCBV and MTT values were observed between left and 

right hemisphere (patients and controls) both in patients and 

controls. After endoarterectomy a decrease of MTT (mean: 

1.03) in both hemispheres in BZ areas was observed while 

no differences were noted in rCBV and rCBF values. 

CONCLUSION 

This study demonstrates an adequate compensation of unilateral 

stenosis when the stenosis is less than 90% (with a 

normal contralateral carotid artery) and that endoarterectomy 

influences only MTT in borderzone. Other studies and 

long-term follow up are needed to confirm these data. 

KEY WORDS: Dynamic susceptibility contrast MR imaging 

288 

Poster 7 

Pial-Pial Collateral Vasculature: A Clinically 

Underestimated Phenomenon 

Kowal, D. J. · Sattenberg, R. J. · Brady, P. S. · Garfinkle, W. 

B. · Oleaga, J. · Htaik, T. · Polasani, R. S. 

Albert Einstein Medical Center 


PURPOSE 

To demonstrate and elaborate upon the nature and inherent 

importance of pial collateral blood supply. Pial-pial collaterals 

have the potential to provide a significant portion of cerebral 

hemispheric blood flow when required. These collaterals, 

and their ability to shunt blood to ischemic tissues, are 

very often underestimated. Their presence is well documented; 

however, their extent and capabilities cannot be appreciated 

until they are required to aid in cerebral perfusion. This 

exhibit demonstrates the imperative function these vessels 

serve, both clinically and angiographically. The extreme situation 

in which pial collateral vessels aid in supplying blood 

to an internal carotid artery occlusion will be demonstrated 

as well. It is hoped that through this exhibit, the abilities and 

importance of pial collateral vessels will become reinforced 

and brought to a conscious level. 


A review of the literature and digital subtraction angiographic 

images are utilized to demonstrate the importance of pialpial 

collateral vasculature during intracranial vascular occlusion 

or high-grade stenosis. 

RESULTS 

Intracranial pial-pial collateral vasculature provides a potential 

for an abundance of increased blood flow in times of 

decreased perfusion, such as in high-grade stenoses, or in 

large vessel occlusion. Furthermore, these collaterals can be 

demonstrated to have patency immediately status-post proximal 

occlusion. 

CONCLUSION 

Intracranial pial-pial collateral vasculature provides a potential 

for an abundance of increased blood flow in times of 

decreased perfusion, such as in high-grade stenoses, or in 

large vessel occlusion. Furthermore, these collaterals can be 

demonstrated to have patency immediately status-post proximal 

occlusion. This exhibit demonstrates the imperative 

function these vessels serve, both clinically and angiographically. 

The extreme situation in which pial collateral vessels 

aid in supplying blood to an internal carotid artery occlusion 

will be demonstrated as well. It is hoped that through this 

exhibit, the abilities and importance of pial collateral vessels 

will become reinforced and brought to a conscious level. 

KEY WORDS: Pial, collateral, vasculature

Poster 8 

Mineralization of the Basal Ganglia Correlates with the 

Extent of Subcortical Ischemic Disease 


J. B. 2 · Hurst, R. W. 2 

1University of California Los Angeles, Los Angeles, CA, 

2University of Pennsylvania, Philadelphia, PA 

PURPOSE 

Iron deposition and calcification of the basal ganglia have 

been associated with disorders that affect subcortical white 

matter. Mineralization of the basal ganglia evident on CT or 

MR imaging may confound acute stroke imaging. We 

hypothesized that mineralization of the basal ganglia may 

serve as a marker of subcortical disease burden due to small 

vessel ischemia and investigated the implications of mineralization 

on gradient-echo sequences in acute stroke imaging. 


Retrospective, blinded review of CT and MR imaging 

(including FLAIR, DWI and GRE) studies acquired contemporaneously 

in 85 cases of stroke admissions (median age 68 

years, range 39-92 years; 41 men, 44 women). Two readers 

scored basal ganglia calcifications on CT (4-point scale), 

GRE hypointensity (3-point scale), and FLAIR subcortical 

disease burden (3-point scale). 

RESULTS 

Basal ganglia calcifications on CT were observed in 32/85 

cases, most commonly situated in the globus pallidus. 

Symmetric signal loss or hypointensity of the basal ganglia 

was observed in 66/85 cases, obscuring the diagnosis of 

microhemorrhage in 11/85. Corresponding hypointensity on 

DWI also caused artifactual lesions of adjacent regions in 

8/85. Mineralization on CT or GRE was associated with 

increasing age (p < 0.010) and the extent of subcortical disease 

on FLAIR (p < 0.001). 

CONCLUSION 

Mineralization of the basal ganglia on CT or GRE may be a 

marker of lesion burden associated with small vessel 

ischemic disease. Multimodal MR sequences employing gradient-echo 

acquisition may be affected also by such mineralization. 

289 

REFERENCES 

1. Casanova MF, Araque JM. Mineralization of the basal ganglia: 

implications for neuropsychiatry, pathology and neuroimaging. 

Psychiatry Res 2003;121(1):59-87 

2. Bakshi R, Dmochowski J, Shaikh ZA, Jacobs L. Gray matter 

T2 hypointensity is related to plaques and atrophy in the 

brains of multiple sclerosis patients. J Neurol Sci 

2001;185(1):19-26 

3. Bartzokis G, Mintz J, Sultzer D, Marx P, Herzberg JS, Phelan 

CK, Marder SR. In vivo MR evaluation of age-related 

increases in brain iron. AJNR Am J Neuroradiol 

1994;15(6):1129-38 

KEY WORDS: Subcortical, basal ganglia, iron 

Poster 9 

Prevalence of Severe and Moderate Stenosis in Patients 

with Acute Carotid Territory Stroke 

Ackerman, R. H. · Speck, L. M. · McGraw, C. · Affel, M. E. 

Massachusetts General Hospital and Harvard Medical 

School 

Boston, MA 

PURPOSE 

In 1991, the results of the North American Symptomatic 

Carotid Endarterectomy Trial (NASCET) suggested benefit 

for carotid endarterectomy (CEA) in preventing stroke in 

patients with an internal carotid artery stenosis of 70% or 

more. Data from the same group and other trials have since 

suggested that even 50-60% (moderate) stenosis may justify 

CEA in symptomatic and asymptomatic patients. Our study 

examines the degree of internal carotid artery stenosis found 

with acute ipsilateral stroke, as documented by CT angiography 

(CTA). The purpose is to determine the respective 

prevalence of severe and moderate ipsilateral internal carotid 

artery stenosis in acute stroke patients with no other identifiable 

stroke cause. 


We reviewed the 2003 CTA findings on all emergency ward 

patients with a discharge diagnosis of stroke who were studied 

acutely on a high resolution (16-slice) CTA device. 

Severity of stenosis was graded on axial and reconstruction 

images. Stroke was documented by head CT or MR imaging. 

Medical charts were reviewed for potential noncarotid stroke 

causes. A 50-60% stenosis best approximates a 2.5-3 mm 

residual lumen diameter (RLD). 

RESULTS 

In 2003, 385 patients had head and neck CTA studies on the 

emergency ward high-resolution scanner. Of these, 110 had 

an acute stroke. CT angiography showed ipsilateral carotid 

artery stenosis of at least 50% in 23 of the 110, and showed 

a stenosis of less than 50% in 87. Of the 23 with 50% stenosis 

or more, 17 had no evident potential cause of stroke other 

than the internal carotid artery lesion. In 8 of these 17, CTA 

was consistent with occlusion, and in 9, with preocclusion 

(an estimated 1 mm RLD or less). Five of the 23 patients 

with stenosis of 50% or more had other identifiable potential 

causes of stroke: 3 had atrial fibrillation (AF), 1 had acute 

lacunar change (LAC), and 1 had grade 5 aortic arch atheromata 

(AAA) by transesophageal echocardiography. Of these 

Posters

Posters 

5, 2 (both in AF) had complete internal carotid artery occlusion 

by CTA, 1 (in AF) had an estimated 70% stenosis (about 

2 mm RLD), and 2 (with LAC or AAA) had an estimated 50- 

60% (moderate) stenosis. In the remaining patient with 

stenosis of 50% or more, the exact degree of stenosis and the 

evidence for noncarotid stroke were uncertain. Of the 87 

patients with less than 50% stenosis, the internal carotid 

arteries were stenosed normal or mildly. 

CONCLUSION 

The data suggest that, in patients with no identifiable cause 

other than extracranial carotid artery disease, stroke typically 

is associated with occlusion or severe narrowing of the 

internal carotid artery lumen. Moreover, if a comprehensive 

workup has been done, it is uncommon to find moderate 

internal carotid artery stenosis as the only identifiable potential 

etiology of stroke. The findings might warrant reconsideration 

of current guidelines that recommend CEA in 

patients with 50-60% stenosis. 

KEY WORDS: Stroke, Carotid, Stenosis 

Poster 10 

Lesional Fractional Anisotropy, Diffusivity and Fiber 

Tractography in Patients with Multiple Sclerosis with 

Diffusion Tensor Imaging at 3 T 

Soohoo, S. · Ge, Y. · Law, M. · Johnson, G. · Herbert, J. · 

Babb, J. · Mannon, L. · Grossman, R. I. 

New York University School of Medicine 

New York City, NY 

PURPOSE 

Diffusion tensor imaging provides in vivo visualization and 

quantitative assessment of white matter fiber tract integrity 

and directionality. This study was to obtain measurements of 

fractional anisotropy (FA), mean diffusivity (MD) and fiber 

tract (FT) number in different lesion types and normalappearing 

white matter (NAWM) in patients with relapsingremitting 

multiple sclerosis (MS) in order to investigate 

whether these parameters in combination with fiber tractography 

are useful in assessing white matter involvement. 


Ten patients with clinical relapsing-remitting MS disease 

were studied on a 3.0 T MR unit. After conventional MR 

imaging, which included T2- and enhanced T1-weighted 

images, axial diffusion tensor imaging (DTI) data were 

acquired with pulsed gradient, double spin-echo, echo-planar 

imaging (5300/74; 128 x 128 matrix; 220 x 220 mm 

FOV; forty-five 3 mm contiguous slices; b = 1000 s/mm2) in 

six directions. Fixed ROIs (4 voxels with a total of 32 tracts) 

were placed within lesions and the comparative contralateral 

NAWM. Measurements for FA, MD, FT and lesion diameter 

were made in 291 nonenhancing lesions (181 

hypointense and 110 isointense), 6 enhancing lesions, and 

the contralateral NAWM for each lesion. 

RESULTS 

A positive correlation between FT and FA (p < 0.0001) and 

a negative correlation between FT and MD (p < 0.0001) 

were found. We also observed a negative correlation 

between FT and lesion size (p = 0.0038) suggesting lesion 

290 

size may affect the number of fibers transected. Compared 

with NAWM (mean FT: 29.8, FA: 0.41, MD: 0.84), lesions 

were associated with significantly lower FT (as shown in 

Fig. 1) and FA, but significantly higher MD (mean FT: 21.4, 

FA: 0.37, MD: 1.19). Differences between lesion types were 

noted also: isointense lesions (mean FT: 25.2, FA: 0.39, MD: 

1.10), hypointense lesions (mean FT: 19.4, FA: 0.37, MD: 

1.24), and enhancing lesions (mean FT: 11.0, FA 0.19, MD: 

1.33). Fig. 1 shows the number of fibers in the left side is 

decreased significantly below a level with a lesion on the 

affected tracks when fixed-size seeds are placed symmetrically 

in both sides of frontal white matter regions. 

CONCLUSION 

Fiber tractography has a role in quantifying the degree of 

axonal loss and demyelination within different lesion types 

and NAWM. The differences in white matter tract disruption 

can be visualized directly and may help to better understand 

the association between lesion type and location with clinical 

symptomatology and in monitoring disease progression. 


fiber tractography 

Acknowledgment: This work was supported by grants R37 

NS29029-11 from NIH and NCRR M01 RR00096 (GCRC). 

Poster 11 

Diffusion Tensor Imaging in Amyotrophic Lateral 

Sclerosis: Predictor Role of Fractional Anisotropy 

Montanaro, D. 1 · Lombardo, F. 1 · Bongioanni, P. 2 · Frijia, F. 1 

· Minichilli, F. 1 · Bianchi, F. 1 · Licitra, R. 2 · Rossi, B. 2 · 

Canapicchi, R. 1 

1 Istituto di Fisiologia Clinica - Consiglio Nazionale Delle 

Richerche, Pisa, ITALY, 2 University of Neuroriabilitazione- 

Università di Pisa, Pisa, ITALY 

PURPOSE 

In amyotrophic lateral sclerosis (ALS), neuropathologic 

changes often begin in motor cortex with secondary degeneration 

of motor fibers and gliosis in the corticospinal tract. 

Upper motor neuron involvement at early stages of disease is

difficult to assess. Quite recently, diffusion tensor imaging 

(DTI), an imaging technique quantifying directionality of 

water diffusion by fractional anisotropy (FA), has been 

reported to detect structural changes in motor system. 

Previous DTI studies in ALS patients have shown reduced 

anisotropy in the pyramidal tracts. Few data are now available 

on relationships between patients’ DTI abnormalities 

and functional impairment. Our research work aimed to analyze 

FA along the pyramidal motor system in subgroups of 

ALS patients as compared to age- and sex-matched controls. 


Twenty-three ALS patients were studied. Diagnosis was 

made according to the El Escorial revisited criteria: 11 

patients were classified as having clinically probable-definite 

ALS; 12 possible-suspected ALS (9 males and 3 

females, mean age ± SD: 60 ± 12 years). Disease severity 

was assessed using the ALS functional rating scale (ALS 

FRS). Subsequently, patients were subgrouped into 3 classes 

according to ALS FRS scores. All patients had a spinal onset, 

and 7 developed bulbar symptoms over time. Patients and 

controls underwent conventional MR imaging (1.5 T, GE 

Signa Horizon System). Moreover, DTI, single voxel spectroscopy 

(MRS) and chemical shift imaging (CSI) were 

acquired in the same groups. Diffusion tensor images were 

acquired using 25 gradient directions to get FA maps. 

Fractional anistropy values of single regions of interest were 

analyzed in many different brain areas (central and precentral 

gray and white matter, internal capsule, cerebral peduncle 

and bulb pyramids). Statistical analysis was performed 

by means of nonparametric test of Mann-Whitney, Kruskal- 

Wallis, Dunn and Spearman. 

RESULTS 

Diffusion tensor imaging findings show significantly 

decreased mean FA values in definite ALS patients in motor 

and premotor areas as compared to those of age- and sexmatched 

healthy controls. Particularly, patients most severely 

impaired were found to have remarkably reduced FA values. 

No statistical differences were noted between males and 

females, and between younger (< 55 years old) and older 

subjects in both patients and controls, neither between suspected 

ALS patients and healthy subjects. 

CONCLUSION 

Our data are in accord with those of scientific literature and 

support at least in part the relationship between upper motor 

neuron involvement and clinical motor disability. In particular 

the statistical alterations of motor and premotor gray and 

white matter could be an index of pathology. We plan to integrate 

such data with those from MRS and CSI, and perform 

MR scans over time for monitoring disease progression in 

order to verify the real utility of MR imaging techniques in 

clinical trials. 

KEY WORDS: Amyotrophic lateral sclerosis (ALS), MR 

imaging, diffusion tensor imaging 

291 

Poster 12 

Normal Human Vermis: MR Imaging Evaluation of the 

Angular Relationships among the Vermian Lobules in 

Vivo 

Schwartz, M. B. · Naidich, T. P. 


New York, NY 

PURPOSE 

The human vermis exhibits 3 lobes with 9 lobules: the anterior 

lobe comprising the (1) lingula, (2) central lobule, and 

(3) culmen, the posterior lobe comprising the (5) declive, (6) 

folium, (7) tuber, (8) pyamis, and (8) uvula, and the flocculonodular 

lobe containing the (9) nodulus (see Fig.). 

Neuroimaging studies have provided qualitative data on 

changes in normal vermian morphology with age and gender, 

and on vermian pathology in autism, schizophrenia, 

ethanol abuse, and congenital malformations. Quantitative 

data on vermian areas and volumes are sparse, inconclusive, 

or contradictory. To our knowledge, no studies provide quantitative 

normative data on the orientation and angular interrelationships 

among the individual lobules of the vermis. 


IRB-approved retrospective analysis of midsagittal 1.5 T T2weighted 

MR images (GE Signa, Milwaukee, WI) was performed 

in 150 normal male and 150 normal female patients, 

distributed evenly across all decades of age. All angles and 

measurements were made directly on the digital image using 

the GE PACS workstation (see Fig.). The skull base was 

assessed by measuring the lengths of the nasion-tuberculum 

(N-T) and tuberculum-basion (T-B) lines, and the acute N-T- 

B angle they form. The cerebral hemisphere and brainstem 

were assessed by measuring the lengths of the anterior commissure-posterior 

commissure (AC-PC) and the PC-obex 

(Ob) lines, and the acute AC-PC-Ob angle they form. The 

angles formed by the major branches of the arbor vitae to 

each of the 9 vermian lobules were standardized to the skull 

base (N-T-B lines) and the cerebrum-brainstem (AC-PC-Ob 

lines). The resultant data for each decade and gender were 

analyzed statistically using the Student T test and Pearson’s 

correlation. 

RESULTS 

The lengths and angles of the skull base and the cerebrumbrainstem 

are highly reproducible. Males show a statistically 

significant larger size of the skull base, cerebrum, and 

brainstem. The angles formed by the arbor vitae are highly 

reproducible with small standard deviations of the means 

and no significant variation with age or gender. Preliminary 

findings indicate that the vermian angles of malformations 

and acquired atrophy differ substantially from the normative 

data above. 

Posters

Posters 

CONCLUSION 

Midsagittal T2-weighted MR imaging displays reproducible 

angular relationships among the skull base, cerebrum-brainstem, 

and the individual lobules of the vermis. These relationships 

are stable over the decades 2-9 in both males and 

females. Initial application of this method to vermian lesions 

reveals substantial differences in the measured angles. These 

abnormal data will be the subject of a separate report. 

KEY WORDS: Vermis, brain, anatomy—normal 

Poster 13 

Cerebrospinal Fluid Space In Hippocampal Formation 

And Value In Alzheimer’s Disease 

Li, Y. · Li, J. · Segal, S. · Santi, S. D. · Zhan, J. · Leon, M. J. D. 

Center for Brain Health 

New York, NY 

PURPOSE 

Small cerebrospinal fluid (CSF) -containing spaces in the 

hippocampal formation have been noted on MR studies of 

medial temporal lobe in Alzheimer’s disease (AD). Most are 

hippocampal sulcus residual cavity (HSC), or dilated uncal 

sulcus and perihippocampal fissure (PF). However, little is 

known about distinguishing among these spaces. This paper 

systematically studied their anatomy and imaging criteria 

and relationship to aging, hippocampal atrophy. Our working 

hypothesis was that the HSC does not like PF and is not 

influenced significantly by AD or atrophy. 


* Subjects: 130 normal control subjects, ranging in age from 

20 to 90 (there were 17 to 20 participants in each age 

decade), and 27 AD subjects, age from 51 to 87.** MR 

Studies: MR scans were performed on a 1.5 T GE scanner. 

Data were obtained using a T2 STIR and T1 coronal SPGR 

sequence and were reconstructed to axial plane parallel to 

the long axial of hippocampus with 1 mm thickness. *** 

HSC and PF Evaluations: each HSC measuring < 3 mm was 

rated 1, and a value of 3 for each cavity ≥ 3 mm in diameter 

292 

to yield a hippocampal sulcus cavity score (HSCS). PFs 

were rated on a 4-point scale on axial imaging: (0~none, 1— 

-question able, 2—mild/moderate, and 3—severe). 

RESULTS 

* Both PFs and HSCS were correlated with age; however 

PFs had significantly stronger association with age (for 

HSCS and age, R 2 = 0.067, P = 0.005; for PF and age R 2 = 

0.59, P = 0.005). ** HSCS were not correlated with hippocampal 

atrophy ( F (5,124) = 1.64, P > 0.05). *** There 

were significant differences in PFs between AD and normal 

control group (F (1.99) = 6.98, P > 0.01) but not in HSC (F 

(1.99) = 0.21, P > 0.5). 

CONCLUSION 

The key to differentiating HSC from PF is location (Fig.1). 

HSC are typically lateral to PF, and PF can be seen to communicate 

with ambient cistern if we track it in continual 

slices. MR imaging signs of brain atrophy consist in volumetric 

decrease of parenchyma and enlargement of CSF 

spaces. So distinguishing HSC and PF may avoid misreading 

hippocampal atrophy. Perihippocampal fissure is a good 

marker for atrophy of hippocampus and is useful to evaluate 

AD, where as HSC is not sensitive to atrophy or age or disease 

effect. 

KEY WORDS: Alzheimer’s disease, hippocampal sulcus 

residual cavity, perihippocampal fissure 

Poster 14 

In Vivo Mapping of Gray Matter Loss with Voxel-Based 

Morphometry in Alzheimer’s Disease and Mild Cognitive 

Impairment 

Moon, W. · Chung, E. 

Kangbuk Samsung Hospital, Sungkyunkwan University 



PURPOSE 

While traditional ROI volumetry cannot evaluate whole 

brain beside selective regions, voxel-based morphometry 

can provide whole brain analysis by voxel-by-voxel compar-

isons. To investigate and compare the regional change of 

gray matter volume using voxel-based morphometry in the 

whole brain with Alzheimer’s disease (AD) and mild cognitive 

impairment (MCI). 


Twenty-four patients with AD, 9 patients with MCI, and 21 

age-matched control subjects participated in this prospective 

study. T1-weighted 3D SPGR scan was acquired for each 

subject. These T1-weighted images were spatially normalized 

into standardized T1 template and segmented into gray 

matter, white matter, and CSF. The gray matter images were 

smoothed with a 12 mm Gaussian kernel. Using voxel-wise 

statistical parametric test (two sample t-test), each of 

smoothing gray matter images of patients with AD and MCI 

was compared with control images. 

RESULTS 

In patients with AD, significant gray matter loss was found 

in both hippocampi, parahippocampal gyri, insular, temporal 

poles, cingulate gyri, and medial thalami (p < 0.05, corrected 

for multiple comparisons). In patients with MCI, significant 

reduction in gray matter volume was detected in both 

cingulate gyri, right posterior temporal cortex, and left superior 

parietal lobule (p < 0.05, corrected multiple comparisons). 

CONCLUSION 

Alzheimer’s disease showed more widespread gray matter 

volume loss than MCI and different pattern of involvement. 

This voxel-based morphometry study suggests AD and MCI 

can be differentiated by using quantitative voxel-by-voxel 

analysis. 

KEY WORDS: Alzheimer’s disease, MR imaging, voxelbased 

morphometry 

Poster 15 

Evaluation of Multiple Sclerosis with Diffusion Tensor 

Imaging 

da Cruz, L. H. · Maia, R. · Ferreira, F. · Domingues, R. · 

Domingues, R. 

CDPI Barrashopping/Multi-Imagem Ressonancia 

Rio de Janeiro, BRAZIL 

PURPOSE 

To determine whether diffusion tensor imaging (DTI) can 

demonstrate differences in FA values of plaques, peri-plaque 

normal-appearing white matter, contralateral white matter in 

patients (CLP) and white matter in normal subjects (CLN). 

To determine whether there are subtle abnormalities in corpus 

callosum in patients that are not detected in conventional 

MR techniques (FLAIR and T2-weighted imaging). 


Thirty-five multiple sclerosis (MS) patients (21 female, 11 

male; mean age 32 ) and 35 healthy controls age- and sexmatched 

underwent diffusion tensor and conventional MR 

imaging exam in a 1.5 T clinical scanner. Fractional 

anisotropy (FA) maps were generated using the DTI data. 

Identical regions of interest (ROI) were placed by an experienced 

neuroradiologist on the plaques, in the normal-appear- 

293 

ing white matter (NAWM) surrounding the plaques, on CLP 

and on CLN, as well as on different locations in the corpus 

callosum in patients and in control subjects. 

RESULTS 

The mean FA of plaques (266.6 ± 84.5) was statically different 

(p < 0.001) from the NAWM peri-plaque FA values ( 

368.5 ± 81.7). There was also a significant difference ( p < 

0.001) between the FA values of plaque and peri-plaque 

regions with CLP (518.5 ± 86.3) and CLn (561.1 ± 93.7). 

Differences were observed between FA values of CLP and 

CLGC, but no statical differences were noted (p > 0.05). 

Significant difference (p < 0.01) was obtained when FA values 

of the jenu ( 776 ± 80.9) and corpus of the corpus callosum 

(698.2 ± 87.8) in patients were compared to normal subjects 

(819 ± 43.2). 

Posters

Posters 

CONCLUSION 

These results suggest that DTI may represent a more accurate 

MR imaging method to determine the extent of the 

demyelinating process in MS patients. DTI also contributes 

in the identification of subtle abnormalities in the corpus callosum 

in patients, maybe due to wallerian degeneration of 

affected axons. 

KEY WORDS: Multiple sclerosis, diffusion tensor imaging 

Poster 16 

CT and MR Imaging in Levamisole-Induced 

Leukoencephalopathy 

Liu, H. · Hsieh, W. · Wu, V. · Wu, K. 

National Taiwan University Hospital 

Taipei, TAIWAN REPUBLIC OF CHINA 

PURPOSE 

To report the imaging finding of levamisole-induced 

leukoencephalopathy. 


In the last 5 years, a total of 9 patients were diagnosed as levamisole-induced 

leukoencephalopathy in our institute. They 

were 5 females and 4 males. Their ages ranged from 26 to 66 

years old. 

RESULTS 

Most of them had clinical manifestation within 1 week after 

taking levamisole for the treatment of recurrent aphthus 

ulceration. Clinically it is very similar to acute demyelinating 

encephalomyelitis (ADEM) or multiple sclerosis (MS). 

The multiple enhancing white matter lesions scattered 

through the cerebral hemispheres, most were limited to 

supratentorial area (> 85%) while in few of them the cerebellum 

and brainstem and midbrian area were involved. The 

signals of these lesions were low on T1-weighted images and 

high on T2-weighted images. The number of lesions tended 

to be numerous in the periventricular area. The enhancement 

could be a nodular, rim or ring pattern. They were bright SI 

on diffusion-weighted imaging but no evidence of restriction 

of water motion on ADC mapping. Many of them appeared 

to be along the subependymal veins that mimicked multiple 

sclerosis. One case documented association with hemorrhage. 

Most lesions disappeared after appropriate management 

and treatment. 

CONCLUSION 

The imaging pictures of levamisole-induced leukoencephalopathy 

mimic ADEM and MS. The diagnosis is totally 

dependent on alertness of this disorder. The prognosis is 

usually good. 

KEY WORDS: Levamisole, demyelination, leukoencephalopathy 

294 

Poster 17 

Evaluation of Basal Nuclei Perfusion with Dynamic 

Susceptibility Contrast MR Imaging in Patients with 

Parkinson’s Disease 


Melis, M. · Floris, R. · Simonetti, G. 


Rome, ITALY 

PURPOSE 

The aim of this study was to confirm with dynamic susceptibility 

contrast (DSC) MR imaging the interhemispheric 

asymmetry of both basal ganglia and thalami observed by 

PET studies in patients with Parkinson’s disease (PD). 


Twenty subjects with idiopathic PD according to the Brain 

Bank Criteria were enrolled. Nineteen normal subjects 

matched for age and sex were included as controls. After 20day 

therapy withdrawal, PD patients underwent DSC MR 

imaging before and after subcutaneous apomorphine injection. 

Regional cerebral blood flow (rCBF) was calculated in 

both patients and normal subjects. Regions of interest of 50 

pixels were placed manually on rCBF maps over the caudate 

nucleus, putamen, external and internal globus pallidus, and 

on the ventrolateral nucleus of thalamus. Data were normalized 

evaluating the rCBF values of occipital white matter. 

Statistical analysis was performed with ANOVA test. 

RESULTS 

Parkinson’s disease patients showed a significant (p < 0.01) 

interhemispheric asymmetry. Regional CBF values were 

higher in the more clinically affected side. After apomorphine 

injection, no statistically significant rCBF changes 

were observed. Healthy subjects did not show any significant 

differences. 

CONCLUSION 

This study suggests that DSC MR imaging is a valuable tool 

in demonstrating the basal nuclei hemodynamic changes in 

PD previously reported by PET studies. The assessment of 

an hemodynamic MR imaging pattern may be of value in the 

diagnosis of Parkinson’s disease. 

KEY WORDS: Basal nuclei perfusion, Parkinson’s disease 

Poster 18 

Pictorial Essay of Corpus Callosum Lesions 

Shetty, R. S. · Hsu, L. 

Brigham and Women’s Hospital 

Boston, MA 

The corpus callosum serves as an important commissural 

pathway connecting the cerebral hemispheres of the human 

brain. Although the exact functional role of the corpus callosum 

is not clearly understood, various pathologic processes 

and new MR imaging techniques provide insight into the role 

of the corpus callosum. This pictorial essay will examine the 

myriad of the major disease processes that can affect the corpus 

callosum. Disease processes that will be displayed and

discussed include congenital malformations including callosal 

dysgenesis and agenesis, neoplasms including lymphomas 

and GBM, degenerative/demyelinating diseases such as multiple 

sclerosis, PML, and Marchiafava-Bignami, vascular diseases 

such as infarcts and callosal hemorrhage, infectious diseases 

such as toxoplasmosis, ADEM, and Lyme’s disease, and 

finally miscellaneous lesions that are less frequent but do 

require mentioning in the differential diagnosis such as 

Langerhan’s histocytosis, Susac’s disease, and ALS. Through 

further understanding of these disease processes and how they 

affect the patient clinically, the functional role of the corpus 

callosum can be understood more clearly. 

KEY WORDS: Corpus, callosum, pathology 

Poster 19 

Globus Pallidus Emodynamic Changes with Perfusion 

MR Imaging: Concausative for Selective Vulnerability to 

Hypoxia? 


Stanzione, P. · Floris, R. · Simonetti, G. 


Rome, ITALY 

PURPOSE 

Globus pallidus (GP) is known as one of the cerebral structures 

showing a high susceptibility to different hypoxic insults 

such as global ischemia, hypotension-induced ischemia, and 

anemic hypoxia due to carbon monoxide (CO) poisoning. The 

aim of this study was to evaluate in normal subjects the local 

perfusion of GP and its response to acetazolamide to assess the 

selective vulnerability of GP to anoxia. 


Thirty-one normal right-handed subjects (20 male and 11 

female; mean age 56 ± 4 years) were enrolled. Before the 

MR study each subject underwent extracranial vessels 

Doppler to exclude vascular defects. All subjects underwent 

perfusion MR imaging. In 10 out of 31 subjects MR imaging 

was repeated twice with at least 24 hours interval. In a further 

subgroup of 11 subjects a second MR examination was 

repeated 15 minutes after iv injection of 14 mg/kg of acetazolamide, 

at least after 24-hour interval. Administration of 

gadolinium and acetazolamide did not have any effect on 

blood pressure, periferal saturation, or heart rate. Regions of 

interest (ROIs) were drawn on the head of caudate nucleus 

(CN), on the putamen (PU), on the external and internal 

globus pallidus (GPe/GPi) separately and on the thalamus 

(TH). Regions of interest also were placed over the parietooccipital 

cortex (F1 F2 P PO). 

RESULTS 

Statistically significant lower rCBF and rCBV values were 

observed both in GPe and in GPi (p < 0.001) in comparison to 

the other three subcortical nuclei (CN, PU TH). Higher rCBF 

and rCBV values were found in GPe and GPi when compared 

with cortex. No differences in MTT values were noted. 

CONCLUSION 

Globus pallidus perfusion may take part, at least as a concausal 

mechanism, in the GP vulnerability to 

hypoxia/ischemia. Because GP has also a glutamatergic 

295 

input from subthalamic nucleus, a double mechanism (vascular 

and excitotoxic) may be hypothesized to explain GP 

susceptibility to hypoxia. Protective mechanisms for selective 

blood supply to GP may play an important role to prevent 

its vulnerability to hypoxia. 

KEY WORDS: Globus pallidus emodynamic changes 

Poster 20 

Demyelination of Brain White Matter on Patients 

Undergone Radiotherapy Treatment of Gliomas 

Evaluated by Magnetization Transfer Ratio 

Castro, J. D. V. 1,2 · Saraiva, L. A. L. 1 · Carlotti, C. G. 1 · Santos, 

M. B. M. 1 · Araujo, D. B. 1 · Santos, A. C. 1 

1 School of Medicine of Ribeirao Preto University of Sao 

Paulo, Ribeirao Preto - Sao Paulo, BRAZIL, 2 School of 

Medicine of Federal University of Ceara, Fortaleza-ce, 

BRAZIL 

PURPOSE 

To evaluate tissue alterations caused by radiotherapy for 

glioma treatment by means of magnetization transfer ratio 

(MTR). This study also aims to identify injuries caused by 

radiotherapy of the brain early, as well as to understand the 

physiopathologic process involved in such tissue changes, 

by using radiotherapy as a noninflammatory mixed (axonal 

and myelin) injury model. 


All patients with a possible diagnose of brain glioma were 

submitted prospectively to a specific MR imaging protocol. 

All exams were performed before surgery and 6 months after 

the onset of radiotherapy. The protocol consisted of two 

coregistered T1-weighted 3D spoiled-gradient recall (SPGR) 

sequence: the first with an “off resonance” pulse and a second 

without the pulse. One can them calculate a percentage 

difference in a voxel by voxel map, named magnetization 

transfer ratio (MTR), according to: MTR = [(Sequence without 

pulse - Sequence with pulse)/Sequence without pulse] x 

100. Using a ROI analysis, we calculated the MTR of the 

splenium of corpus callosum before and after radiotherapy. 

All patients had that region within the irradiated field. A total 

of seven patients completed the protocol. The results were 

analyzed using Wilcoxon T test. 

RESULTS 

The presence of a decreased MTR of the splenium of corpus 

callosum was observed on all patients who underwent radiotherapy 

for the treatment of brain glioma (p = 0.022). 

CONCLUSION 

Magnetization transfer ratio is able to show early alterations 

of the irradiated brain white matter, even before the onset of 

alterations seen on FLAIR and T2 sequences. Therefore, it 

presents a potential use as an alternative for the analysis of 

postradiotherapy injuries that may be used as an in vivo 

model of demyelination and remyelination. 

KEY WORDS: Magnetization transfer, radiotherapy, demyelination 

Posters

Posters 

Poster 21 

Neuroimaging Findings in Khat Intoxication 

Silbergleit, M.D., R. · Mallesh, A. · Kragha, K. K. · Hakim, 

M. · Wang, A. · Noujaim, S. E. 

William Beaumont Hospital 

Royal Oak, MI 

PURPOSE 

To demonstrate the MR and CT neuroimaging findings of 

leukoencephalopathy associated with Khat abuse. 


A 52-year-old male native of Yemen presented with tremor, 

fatigue, garbled speech, nausea, and vomiting. He had a history 

of Khat use every weekend for several years without 

similar episodes. 

RESULTS 

CT two days after onset of symptoms demonstrated low 

attenuation in the periventricular and subcortical white matter 

with sparing of the cortical and deep gray structures. MR 

imaging performed 3 days after onset of symptoms demonstrated 

high signal intensity in the white matter on T2weighted 

and FLAIR images. Diffusion-weighted images 

were unremarkable and there was no pathologic enhancement. 

CONCLUSION 

The white matter disease demonstrated on both CT and MR 

imaging in this patient who had abused Khat (from the tree 

Catha edulis, also known as Qat or Kat) is nonspecific. The 

pattern is identical to that in the one reported case with neuroimaging. 

Knowledge of this pattern may be useful in 

including the possibility of Khat abuse in the appropriate 

clinical setting. Khat abuse is increasing in Europe and 

North America as the immigrant population from Yemen and 

Somalia (where Khat chewing is a common social tradition) 

increases. 

KEY WORDS: Khat, white matter disease, substance abuse 

Poster 22 

Brain MR Imaging Findings in Liver Transplant 

Recipients 

Besenski, N. · Rumboldt, Z. · Nicholas, J. · Harkey, P. · 

Sachin, S. · Budisavljevic, M. 



PURPOSE 

To establish the prevalence and nature of brain lesions in 

liver transplant recipients (LTR) by MR imaging after the 

orthotopic liver transplantation (OLT) and to distinguish 

acute from chronic processes. 


Ninety-three adult LTR who presented with various neurologic 

symptoms underwent a total of 166 brain MR imaging 

exams. Their MR imaging data were divided into 2 groups, 

based on the time between the transplantation and MR study. 

296 

Liver transplant recipients in Group 1 (40/93) had MR imaging 

within 3 months, and in Group 2 (53/93) more than 3 

months after transplantation. The patients’ clinical status also 

was evaluated, including blood pressure, and laboratory 

parameters were collected, including levels of calcineurin 

inhibitors (CI, cyclosporine or tacrolimus), cholesterol and 

magnesium. Based on MR imaging findings the lesions were 

classified as acute or chronic. Acute changes included acute 

vascular lesion [acute and subacute infarction, posterior 

reversible encephalopathy syndrome(PRES), anoxic injury, 

hemorrhage], infection (meningitis, toxoplasma) and central 

pontine myelinolysis (CPM). Chronic changes included 

chronic vascular changes (chronic infarct, microangiopathy, 

atrophy), basal ganglia hyperintensity on T1-weighted 

images, neoplasms [primary CNS lymphoma (PCNSL)], and 

bony changes of calvaria. Differences between the two 

groups in the acute changes, chronic changes, and normal 

scans were compared on a pair-wise basis using the 2-sided 

Fisher’s exact test. 

RESULTS 

Results are summarized in Table 1. Normal MR imaging was 

found in 13% (12/93), and acute lesions were found in 23% 

(21/93) of all LTR patients. Patients with chronic vascular 

changes usually had more than one single lesion. 

Statistically significant differences between the groups in 

MR imaging findings were found for anoxic injury only, 

which was more common in the first 3 months after OLT. 

MRI findings in liver transplant recipients 

Group 1 Group 2 p value (2- 

(within 3 mos. (after 3 mos) sided Fisher's 

(40 patients) (53 patients) exact text) 

ACUTE LESIONS 

Acute Infarct 3 (8%) 5 (9%) 1.000 

Infection 0 2 (4%) .504 

Central Pontine Myelinolysis 2 (5%) 0 .182 

Posterior Reversible Encephalopathy 3 (8%) 0 0.076 

Anoxic Injury 

CHRONIC LESIONS 

5 (13%) 0 0.013* 

Chronic Vascular Lesions 36 (90%) 43 (81%) .380 

Lymphoma 0 1 (2%) 1.000 

High T1 Signal in Globus Pallidus 15 (38%) 11 (21%) .103 

Bony Calvarial Changes 0 1 (2%) 1.000 

NORMAL 3 (8%) 9 (17%) .222 

* statistically significant at p

297 

Poster 23 

Poster 24 

Hippocampal Volume Map in the Mesial Temporal Initial and Longitudinal Follow-up MR Imagings of 

Sclerosis Patient to Evaluate Postsurgical Outcome Progressive Multifocal Leukoencephalopathy 

Salamon, N. · Lin, J. · Lee, A. · Dutton, R. · Thompson, P. · 

Toga, A. · Engel, J. 



PURPOSE 

Using an anatomical surface modeling approach, 40 patients 

with mesial temporal lobe epilepsy and pathologically 

proven hippocampal sclerosis were studied to evaluate hippocampal 

atrophy profiles of patients who became seizure 

free (SF) after anteriomesial temporal resection, and those 

who did not attain complete seizure freedom (NSF). 


The patients were chosen according to the following criteria: 

1) Preoperative MR imaging showed unilateral hippocampal 

atrophy. 2) Seizure semiology compatible with mesial temporal 

lobe epilepsy. 3) Noninvasive video EEG monitoring 

showed unilateral anterior temporal ictal onset, concordant 

to the side of hippocampal atrophy on MR imaging. 4) 

Surgical pathology showed evidence of hippocampal sclerosis. 

The patients were followed for at least 2 years after the 

surgery. Thirty patients who became seizure-free after surgery 

and 10 patients who continue to have seizures after surgery 

were enrolled in the study.Using a surface-based 

anatomical mapping method, hippocampal structures were 

traced and volumes were analyzed, and compared to the contralateral 

side. Degree of the atrophy was compared between 

SF and NSF groups. 

RESULTS 

Mean hippocampal volumes showed greater atrophy in the 

NSF group. The hippocampal asymmetry map in the NSF 

group showed lesser degrees of asymmetry compared to the 

SF group along the entire hippocampal axis without regional 

specificity. Contralateral hippocampal atrophy showed 

significant atrophy in the anterior and lateral aspect in the 

NSF group. 

CONCLUSION 

MR imaging-based surface modeling revealed patients with 

NSF surgical outcome have significantly greater bilateral 

regional atrophy compared to the SF group. 

KEY WORDS: Hippocampal sclerosis, voxel-based morphometry, 

postsurgical outcome 

Sakai, M. 1 · Watanabe, Y. 2 · Tanaka, H. 2 · Mitomo, M. 1 · 

Hosoki, T. 1 · Shirasaka, T. 1,3 

1 2 Osaka National Hospital, Osaka, JAPAN, Osaka Universal, 

School of Medicine, Osaka, JAPAN, 3AIDS Medical Center, 

Osaka National Hospital, Osaka, JAPAN 

PURPOSE 

To evaluate MR images of multiple lesions in patients with 

progressive multifocal leukoencephalopathy (PML) on initial 

and longitudinal follow-up studies. 


We retrospectively reviewed the MR images of eight patients 

with a diagnosis of HIV-related PML who were hospitalized 

during 1995 to 2004. Diagnosis was established in five 

patients by PCR demonstration of JC virus DNA in the CSF, 

in one characteristic clinical presentations, in one by biopsy, 

and in one at autopsy. At the time of the initial MR imaging, 

durations of initial neurologic symptoms and signs attributable 

to PML had ranged from 10 days to 4 months with a 

mean of 2.7 months. Their ages ranged from 12 to 49 years 

with a mean age of 31.5. MR T1- and T2-weighted axial 

images were evaluated. Follow-up MR studies were available 

in six patients. 

RESULTS 

In total, we identified 62 lesions on the initial MR images 

(Table). All patients had multiple lesions. Among them, 48 

lesions of seven patients were supratentorial white matter 

(WM) lesions and five lesions of three patients were deep 

gray matter (GM) ones. White matter lesions were bilateral 

in six patients although asymmetrical in their severity. 

Frontal subcortical WM lesions were most common. Of five 

deep GM lesions, four were accompanied by surrounding 

WM lesions. Nine lesions (three patients) were infratentorial. 

One patient showed the only infratentorial lesions. All 

lesions were hyperintense relative to adjacent GM on T2weighted 

images and most were hypointense on T1-weighted 

images. Six of the eight patients underwent follow-up MR 

imaging. The duration of their illness varied between 2 and 

44 months after the initial PML signs and symptoms. We 

evaluated 53 lesions on the initial MR images and 39 lesions 

appearing during the follow up (total: 92 lesions). 

Progression of WM abnormalities occurred in all patients in 

a several months span. The lesions appear as patchy areas of 

T2-hyperintensity within the subcortical WM. Besides subcortical 

extension, the margins of the multifocal lesions coalesced 

and represented a diffuse pattern. Twenty of 92 

lesions manifested the spontaneous swelling during follow 

up. 

Posters

Posters 

Number and frequency of PML lesions on the initial MR imaging 

Location Number of lesions: total 62 Frequency (%) 

Supratentorial WM: total 48 77.4 

Frontal WM 13 21.0 

Parietal WM 7 11.3 

Temporal WM 9 14.5 

Posterior WM 6 9.7 

Subcortical WM 21 33.9 

Subcortical-deep WM 12 19.4 

Deep WM 2 3.2 

Internal capsule 3 4.8 

External capsule 3 4.8 

Corpus callosum 7 11.3 

Deep GM: total 5 8.1 

Thalamus 1 1.6 

Globus pallidus 2 3.2 

Putamen 2 3.2 

Caudate nucleus 0 0.0 

Infratentorial: total: total 9 14.5 

Brain stem 2 3.2 

Cerebellum 7 11.3 

CONCLUSION 

Progressive multifocal leukoencephalopathy lesions tends to 

extend from the supratentrial subcortical WM to deep WM. 

Although lesions show atrophy with progression of PML, 

not a few lesions swell up during follow up before becoming 

atrophic. Radiologists are not able to exclude PML from the 

differential diagnosis even when the lesions show swelling. 

KEY WORDS: Progressive multifocal leukoencephalopathy, 

HIV, MR imaging 

Poster 25 

MR Imaging Features of Cerebral Cryptococcoma with 

Diffusion-Weighted Images 

Ho, T. 1 · Lee, H. 2 · Chen, W. 1 · Lee, K. 1 

1Changhua Christian Hospital, Changhua, TAIWAN 

2 REPUBLIC OF CHINA, New Jersey Medical 

School/University of Medicine & Dentistry of New Jersey, 

Newark, NJ 

PURPOSE 

Diffusion-weighted imaging and apparent diffusion coefficient 

(ADC) maps provide a way to evaluate the diffusion 

properties of water molecules in tissue and have been said to 

be valuable in the differentiation between pyogenic brain 

abscesses and cystic necrotic tumors. We report a case of 

cerebral cryptococcoma to show its signal intensity features 

on the diffusion-weighted images and ADC maps, compared 

to that of a pyogenic brain abscess and a brain tumor. 


The patient was a 55-year-old woman, a factory worker, who 

denied any past history of illness. She suffered from 

headache for about 1 year and left face weakness for 1 day. 

Physical examination showed coarse breathing sound bilaterally. 

Neurologic examination showed leftward nystagmus, 

decreased pain sensation in left face, central type facial palsy 

on the left face, and decreased hearing on the left. She has no 

history of pigeon contact. Contrast-enhanced CT scan of the 

chest showed a shaggy nodular lesion in the superior segment 

of right lower lobe, and lymph node enlargement in the 

right pulmonary hilar region. Contrast-enhanced CT scan of 

the head showed a lobulated, ring-enhancing mass lesion 

with a small central cavity, associated with significant peri- 

298 

focal white matter edema, located in right frontal lobe of the 

brain. Contrast-enhanced MR imaging with diffusionweighted 

imaging and ADC maps of the head was performed. 

RESULTS 

T1-weighted image showed a hypointensity mass lesion with 

associated perifocal white matter edema. FLAIR image 

showed hypointensity in the central cavity of the mass 

lesion. T2-weighted image showed hyperintensity in the central 

cavity, as well as prominent perifocal white matter 

edema. Diffusion-weighted imaging (diffusion sensitivity of 

b = 1000 s/mm 2 ) showed hypointensity in the central cavity 

of the mass lesion, while ADC maps showed hyperintensity 

in that central cavity. Contrast-enhanced T1-weighted image 

showed a ring-pattern enhancing-mass lesion. Under the 

impression of brain tumor, craniotomy was performed and 

the mass lesion was removed. Histopathology showed cryptococcosis, 

manifested as the presence of numerous round 

microorganisms, proved by mucicarmine, periodic acid- 

Schiff, and methenamine silver stains. Serum cryptococcal 

antigen was positive with a titer of 1:512. Follow-p brain tissue 

culture with Sabouraud’s media at 20-37° C showed 

Cryptococcus neoformans. 

CONCLUSION 

Diffusion-weighted imaging to distinguish between pyogenic 

brain abscesses and cystic or necrotic brain masses has 

been reported to be useful in several publications. The viscous 

consistency of the pus could account for the restricted 

diffusion and therefore high diffusion-weighted imaging signal 

intensity and hypointensity on the ADC maps. By contrast, 

the diffusion signal features (hypointensity on diffusion-weighted 

imaging and hyperintensity on ADC maps) of 

the central cavity of a cryptococcoma, as shown by the present 

case, are opposite to that of a pyogenic abscess. It may 

be reasoned that the central cavity content of a cryptococcoma, 

with signal intensity reflecting marked unrestricted diffusion, 

is much less viscous than pus, and more like free 

water. Brain tumors usually present with imaging features 

similar to a cryptococcoma (i.e., ring-enhancing masses, 

with a central cavity showing hypointensity on diffusionweighted 

imaging and hyperintensity on ADC maps). 

KEY WORDS: Apparent diffusion coefficient maps, cryptococcoma, 

diffusion-weighted imaging 

Poster 26 

Diffusion-Weighted Images Of Progressive Multifocal 

Leukoencephalopathy 

Sakai, M. 1 · Watanabe, Y. 2 · Mitomo, M. 1 · Hosoki, T. 1 · 

Shirasaka, T. 1 

1 Osaka National Hospital, Osaka, JAPAN, 2 Osaka 

University, School of Medicine, Osaka, JAPAN 

PURPOSE 

The purpose of this presentation is to illustrate the characteristics 

of diffusion-weighted imaging findings of progressive 

multifocal leukoencephalopathy (PML).


Two patients with PML diagnosed by polymerase chain reaction-based 

demonstration of JC virus DNA. They underwent 

MR imaging including diffusion-weighted imaging (b = 0, 

1000 s/mm 2 ). One of them (patient 2) underwent follow-up 

studies at 5 and 10 weeks after the initial MR imaging. The 

duration of their illness were 2 (patient 1) and 4 (patient 2) 

months. 

RESULTS 

Both patients had multiple lesions. They were bilateral 

although asymmetrical in their severity. Diffusion-weighted 

intensities in the lesions varied from marked hypo to hyper. 

Newer lesions and the advancing edge of large lesions 

showed mild decrease to slight elevation of apparent diffusion 

coefficient (ADC) values and gave hyperintensity on 

diffusion-weighted images, while older lesions and the center 

of large lesions showed markedly increased ADC values 

and gave hypointensity. In patient 2, diffusion-weighted 

intensities decreased during follow up (Fig.) with an increase 

in signal on T2-weighted images and atrophy. The differences 

in MR images with diffusion-weighted findings were 

considered to corresponds to the degree of demyelination. 

CONCLUSION 

Our observation indicates characteristic diffusion-weighted 

imaging findings of the PML lesion depending on stage of 

the lesion corresponding degree of demyelination. These 

findings might help imaging diagnosis of PML and clinical 

treatment. 


difffusion-weighted images, MR spectroscopy 

Poster 27 

MR Spectroscopy and MR Perfusion Characterization of 

Intracranial Tuberculoma 

Datir, A. P. 

Sir Jamsetjee Jeejebhoy Group of Hospitals 

Mumbai, INDIA 

PURPOSE 

Characterizing tuberculomas on the basis of 1) Ratios of different 

spectroscopic parameters between the tuberculoma 

and the normal brain parenchyma, 2) Normalized ratios of 

rCBV and rCBF measured at the center, periphery of tuberculoma, 

and the surrounding white matter, on the MR perfusion 

study. 

299 


Twenty patients diagnosed of CNS tuberculoma on the basis 

of CT and MR findings, CSF picture and positive follow-up 

response to antituberculous treament were subjected to MR 

spectroscopy and perfusion studies. 

RESULTS 

In MRS for NAA/Cr, the ratios were found to be 3.7 on SVS 

and 4.0 on MVS. For Cho/Cr, the ratios were found to be 3.5 

on SVS and 2.6. Ratio of NAA between tuberculoma and 

normal brain parenchyma was found to be 0.72. Similarly 

the ratio of Cho between tuberculoma and normal brain 

parenchyma was 0.80. Observations on MRP: 1) center of 

tuberculoma, the mean value of rCBV and rCBF are 0.8915 

and 0.845 respectively, 2) periphery of tuberculomas, the 

mean value of rCBV and rCBF are 1.265 and 1.524 respectively, 

and 3) perilesional white matter showed rCBV and 

rCBF to be 1.043 and 1.194 respectively. 

CONCLUSION 

We conclude that the mean value of 0.72 for NAA and 0.8 

for choline was found quite consistently. With increasing 

availability of MR spectroscopy and perfusion techniques, 

these may be used more frequently in near future in the diagnosis 

of tuberculoma. 

KEY WORDS: Tuberculoma, spectroscopy, perfusion imaging 

Poster 28 

MR Imaging of Cerebral Cryptococcal Infection in the 

Immunocompetent Patients 

Kan, S. 1 · Woodhams, R. 1 · Matsunaga, K. 1 · Hayakawa, K. 1 · 

Ootaka, H. 2 

1 2 Kitasato University, Sagamihara, JAPAN, Higashiyamato 

Hospital, Higashiyamato, JAPAN 

PURPOSE 

Cryptococcal infection is often seen in patients with compromised 

cellular immune response, particularly in those 

with AIDS. However, up to 30% of cases are reported in 

patients with no predisposing condition. The purpose of this 

study is to review MR imaging of cerebral cryptococcal 

infection in the immunocompetent patients. 


MR imagings of five cases of cerebral cryptococcal infection 

were reviewed. Age ranged from 20 to 55 years old. In four 

cases, cryptococcal infection was diagnosed by positive 

cryptococcal antigen titer and India ink stain from cerebral 

spinal fluid. The diagnosis was made with pathologic study 

in one operated case. 

RESULTS 

MR findings were: ventricular enlargement 3/5 (Fig.1), 

Virchow Robin space enlargement 3/5, small infarction 2/5, 

hight intensity on T1-weighted image. Contrast study was performed 

in four cases: leptomeningeal enhancement 2/4, 

ependymal enhancement 1/4, multicystic enhancement 1/4 

(Fig. 2), nodular enhancement 2/4. Diffusion-weighted image 

was done in three cases. High intensity was noted in two 

cases, one is due to infarction and the other is due to ventri- 

Posters

Posters 

culitis. Low intensity was noted on diffusion-weighted imaging 

at cryptococcoma; FLAIR image was done in three cases. 

Additional information was not obtained with these studies. 

CONCLUSION 

The most common pattern of cryptococcal infections in central 

nervous system is ventricular enlargement and Virchow 

Robin space enlargement. 

Several patterns of contrast enhancement were noted. Initial 

symptom might be due to infarction with cryptococcal 

meningitis. Diffusion-weighted image showed low intensity 

at cryptococcoma. 

REFERENCES 

1. Yu-Chen C, Jing-Feng L, Feng-Chi C, et al. Radiological manifestations 

of Cryptococcal infection in central nervous system. 

J Chin Med Assoc 2003;66:19-26 

2. Aharon-Peretz J, Kliot D, Finkelstein R, et al: Cryptococcal 

meningitis mimicking vascular dementia. Neurology 

2004;62:2135 

KEY WORDS: MR imaging, cryptococcus, brain 

Poster 29 

Laminar Necrosis: Hemorrhage or Inflammation? 


J. B. 2 · Hurst, R. W. 2 

1 University of California Los Angeles, Los Angeles, CA, 

2 University of Pennsylvania, Philadelphia, PA 

PURPOSE 

Laminar necrosis may be evident in cortical infarcts as T1weighted 

hyperintensity due to suspected microhemorrhage, 

although histopathology has failed to reveal the presence of 

heme. We hypothesized that laminar necrosis evident on 

FLAIR and GRE may represent sequelae of inflammatory 

pathology including lipid-laden macrophage deposition and 

iron. 


Retrospective review of clinical data and MR imaging 

(including FLAIR and GRE) in 95 cases of stroke with laminar 

necrosis (median age 61 years, range 30-89 years; 40 

men:55 women). Inflammatory stroke etiologies were noted 

on chart review. MR imaging lesion extent, location, sulcal 

or gyral predominance, adjacent subcortical lesions, and parent 

vessel patency were reviewed on pre and postcontrast 

T1-weighted sequences, FLAIR, GRE and diffusion-weighted 

imaging. Parent vessel status was documented on MRA. 

300 

RESULTS 

Laminar necrosis was more pronounced on postcontrast T1weighted 

sequences with frequent enhancement of adjacent 

cortical tissue (62%), into the chronic phase. Prominent 

FLAIR hyperintensity noted in all cases suggested T1weighted 

effects, possibly due to lipid. Isolated, serpentine 

GRE hypointensity in these lesions without associated 

microhemorrhages suggested possible iron. Diffusionweighted 

imaging hyperintensity was noted in chronic cortical 

lesions. Underlying subcortical tissue frequently was 

spared with parent vessel typically patent. Inflammatory 

stroke etiologies and seizures were noted in a disproportionately 

high number of cases. 

CONCLUSION 

Laminar necrosis may be associated with inflammation at 

the cortical level, possibly increasing the occurrence of 

seizures due to gliosis and relative sparing of underlying 

subcortical tracts. 

REFERENCES 

1. Siskas N, Lefkopoulos A, Ioannidis I, Charitandi A, Dimitriadis 

AS. Cortical laminar necrosis in brain infarcts: serial MRI. 

Neuroradiology 2003;45(5):283-288 

2. Komiyama M, Nakajima H, Nishikawa M, Yasui T. Serial MR 

observation of cortical laminar necrosis caused by brain 

infarction. Neuroradiology 1998;40(12):771-777 

3. Saleh A, Schroeter M, Jonkmanns C, Hartung HP, Modder U, 

Jander S. In vivo MRI of brain inflammation in human 

ischaemic stroke. Brain 2004;127(Pt 7):1670-1677 

KEY WORDS: Laminar necrosis, stroke, inflammation 

Poster 30 

Demyelinating Course of HIV-Related Progressive 

Multifocal Leukoencephalopathy: Validated by Proton 

MR Spectroscopy and Diffusion-Weighted Imaging 

Sakai, M. 1 · Watanabe, Y. 2 · Mitomo, M. 1 · Hosoki, T. 1 · 

Takahama, S. 1 · Shirasaka, T. 1 

1 Osaka National Hospital, Osaka, JAPAN, 2 Osaka 

University, School of Medicine, Osaka, JAPAN 

PURPOSE 

To illustrate the characteristics of progressive multifocal 

leukoencephalopathy (PML) lesions on proton MR spectroscopy(1H-MRS) 

and diffusion-weighted images


We retrospectively reviewed the 1H-MRS and diffusionweighted 

images of 4 patients (3 male) infected with HIV-1, 

with PML established by PCR demonstration of JC virus 

DNA in the CSF who were hospitalized during 1998 to 2004. 

At the time of the initial MR imaging, durations of initial 

neurologic symptoms and signs attributable to PML had 

ranged from 2 to 23 months with a mean of 8.8 months 

(Patient 1: 23 months, 2: 2 months, 3: 4 months, 4: 6 

months). Standard T1-, T2-, and diffusion-weighted 

sequences (b = 0, 1000 s/mm2) were performed in all 

patients; 1H-MRS was performed in two locations in three of 

the patients (patients 1, 3, 4). From a transaxial localizer, 

spectroscopy voxel locations were defined and placed within 

the largest lesion in each patient. A contralateral voxel, 

contralateral from the lesion, also was obtained in each 

patient. On 1H-MRS, N-acetyl aspartate (NAA), creatinine 

(Cr), choline-containing compounds (Cho), and lactate were 

estimated. Patient 3 underwent repeated MR studies at 5, 10, 

and 14 weeks after the initial MR study. Patient 4 underwent 

repeated MR studies with 1H-MRS at 3 weeks after the initial 

MR study. 

RESULTS 

All PML lesions had a distinct pattern in the combination of 

conventional MR images, diffusion-weighted images, and 

1H-MRS spectra. The lesions were asymmetrical confluent 

white matter lesions hyperintense relative to adjacent gray 

matter on T2-weighted images and most were hypointense 

on T1-weighted images. Hyperintense lesions on diffusionweighted 

images showed decreased or slightly increased 

apparent diffusion coefficient (ADC) values, while iso or 

hypointense lesions on diffusion-weighted images showed 

markedly increased ADC values. In patient 3 and 4, diffusion-weighted 

intensities decreased during follow up with an 

inverse increase in signal on T2-weighted images and with 

corresponding increase of ADC values. 

Correlating with 1H-MRS, in the lesions of diffusionweighted 

hyperintense 1H-MRS revealed mild decrease in 

NAA/Cr and increased Cho/Cr. Excess lactate was not noted 

(lactate negative pattern). On the other hand, in the lesions of 

diffusion-weighted iso or hypointensity, excess lactate was 

present and marked decrease in NAA/Cr and mild increased 

Cho/Cr were shown (lactate positive pattern). Patient 1, 

whose lesion showed marked hyperintensity near CSF with 

severe atrophy on T1- and T2-weighted images and 

hypointensity on diffusion-weighted images showed 

decreases in all metabolites. In patient 4, findings of diffusion-weighted 

imaging and 1H-MRS pattern showed initially 

lactate negative pattern and then lactate positive pattern 3 

weeks after. 

CONCLUSION 

The combination of conventional MR imaging with some of 

its new developments, such as diffusion-weighted imaging 

and 1H-MR spectroscopy, seems to constitute a powerful 

diagnostic and follow-up tool with potentially prognostic 

value. Because these techniques are noninvasive, they can be 

used longitudinally to determine the progressive biochemical 

and pathologic changes in the lesions and to monitor the 

response to therapy. 


MR spectroscopy, diffusion 

301 

Poster 31 

Evaluation of Perfusion-Weighted MR Imaging in 

Anaplastic Gliomas with or without Oligodendroglial 

Components 

Hirai, T. 1 · Kitajima, M. 1 · Fukuoka, H. 1 · Yamura, M. 1 · 

Hayashida, Y. 1 · Korogi, Y. 2 · Yamashita, Y. 1 · Nakamura, H. 1 

· Kuratsu, J. 1 

1Kumamoto University Graduate School of Medical 

Sciences, Kumamoto, JAPAN, 2University of Occupational 

and Environmental Health, School of Medicine, Kitakyushu, 

JAPAN 

PURPOSE 

Prognosis of anaplastic gliomas is usually poor; however, it 

has been noted that existence of the oligodendroglial components 

leads to relatively better response to chemotherapy, 

and anaplastic gliomas with oligodendroglial components 

have a better prognosis than anaplastic astrocytomas. 

Although characteristic imaging findings of oligodendroglial 

tumors have been reported, the differentiation of anaplastic 

astrocytomas from anaplstic gliomas with oligodendroglial 

components is often difficult with conventional MR imaging. 

The purpose of this study was to evaluate the quantitative 

perfusion-weighted MR imaging of anaplastic gliomas 

with oligodendroglial components including anaplastic 

oligodendrogliomas (AOD) and anaplastic oligoastrocytomas 

(AOA) comparing with that of anaplastic astrocytomas 

(AA). 


Thirty consecutive patients (20 men and 10 women, ranging 

in age from 19 to 77 years) with histologically proven highgrade 

gliomas (AOD, n = 8; AOA, n = 14; AA, n = 8) underwent 

conventional and spin-echo echo-planar imaging (EPI) 

perfusion-weighted MR imaging. Representative maximal 

rCBV regions of interest were chosen from each lesion. 

Resultant values were normalized to those of corresponding, 

contralateral, uninvolved white matter regions. Mean normalized 

CBV (nCBV) value of each histologic group was 

compared among the three glioma groups. Statistically 

analyses were performed using one factor ANOVA and 

Scheff’s F test. 

RESULTS 

The mean nCBV values in AOD, AOA, and AA were 1.1 ± 

0.05, 1.02 ± 0.05, and 1.02 ± 0.05, respectively (Fig.). The 

mean nCBV value of AOD was significantly higher than that 

of AA or AOA (p < 0.05). There was no significant difference 

in mean nCBV between AA and AOA. 

Posters

Posters 

CONCLUSION 

In the evaluation of nCBV of anaplastic gliomas with perfusion-weighted 

MR imaging, AOD may have higher nCBV 

than AA or AOA. This information should be taken into 

account for the evaluation of perfusion-weighted MR imaging 

in anaplastic gliomas. 

KEY WORDS: Brain tumors, perfusion-weighted MR imaging, 

anaplastic gliomas 

Poster 32 

Role of Diffusion- and Perfusion-Weighted MR Imaging 

in Differentiating Meningioma and Dural Metastasis 

Hakyemez, B. · Erdogan, C. · Yildirim, N. · Parlak, M. 

Uludag University Medical School 

Bursa, TURKEY 

PURPOSE 

Meningiomas are the most common primary nonglial tumors 

of the brain and can be recognized easily with conventional 

MR sequences (1). Isolated dural metastases, on the other 

hand, comprise less than 1% of all intracranial metastasis. 

Breast, lung, and renal malignancies are usual causes of 

dural metastasis (2). In some cases it is not possible to differentiate 

meningiomas from isolated dural metastasis with 

conventional imaging techniques (3). In this study our aim is 

to evaluate the role of diffusion and dynamic contrastenhanced 

perfusion-weighted MR imaging in differentiating 

meningiomas and dural metastasis. 


In this study, 14 patients with solitary dural metastasis (4 

patients with breast carcinoma, 5 patients with lung cancer, 

1 patient with lymphoma, 1 patient with malignant 

melanoma and 1 patient with pancreas carcinoma) and 26 

patients with meningiomas were involved. All of the meningiomas 

were operated and histologically proved. All metastasis 

except 5 also were histologically proven after surgical 

excision. The diagnosis was established by clinical follow up 

in patients who did not undergo surgery. In the present study, 

the imaging characteristics were analyzed using conventional 

MR imaging (T1-weighted, T2-weighted, and contrastenhanced 

T1-weighted sequences). In addition to conventional 

MR sequences, diffusion-weighted imaging was performed 

from the solid components of the lesions. Perfusionweighted 

MR imaging was performed by using a first-pass 

gadopentetate dimeglumine T2*-weighted spin-echo singleshot 

echo-planar sequence. Relative cerebral blood volume 

(rCBV) was calculated using the ratio between the CBV in 

high color areas of the lesions and in the normal contralateral 

white matter. Mann-Whitney test was used for the statistical 

analyze of diffusion- and perfusion-weighted MR imaging 

findings of meningiomas and dural metastasis. 

RESULTS 

Dural metastasis and meningiomas could not be differentiated 

from each other with qualitative assessment of conventional 

and diffusion-weighted trace MR images. Quantitative 

evaluation of dural metastasis revealed that mean ADC value 

was 1.17 ± 0.25 while it was 1.11 ± 0.17 for meningiomas. 

The difference between these values was not statistically significant 

(p > 0.05). On the other hand, rCBV ratios for dural 

302 

metastasis and meningiomas were 4.13 ± 2.32 and 7.32 ± 

4.10 respectively and the difference between two was statistically 

significant (p = 0.003). 

CONCLUSION 

Conventional MR findings of dural metastasis and meningiomas 

may overlap in some of the lesions. In differentiation 

of these lesions diffusion-weighted MR imaging yields no 

further information additional to conventional sequences. 

Perfusion-weighted MR imaging can be useful in differentiating 

these lesions by demonstrating high rCBV ratios for 

meningiomas and low rCBV ratios for metastasis. 

REFERENCES 

1. Yang S, Law M, Zagzag D, et al. Dynamic contrast-enhanced 

perfusion MR imaging measurements of endothelial permeability: 

differentiation between atypical and typical meningiomas. 


2. Johnson MD, Powell S, Boyer P, Weil R, Moots P. Dural lesions 

mimicking meningiomas. Hum Pathol 2002;33:1211-1226 

3. Tagle P, Villanueva P, Torrealba G, Huete I. Intracranial metastasis 

or meningioma? An uncommon clinical diagnostic 

dilemma. Surg Neurol 2002;58:241-245 

KEY WORDS: MR perfusion imaging, metastases, meningiomas 

Poster 33 

Gadobenate Dimeglumine Contrast Enhancement of the 

Central Nervous System: Does Improved Relaxivity 

Matter? 

Tanenbaum, L. N. 

Edison Radiological Group/NJ Neuroscience 

Edison, NJ 

PURPOSE 

Gadobenate dimeglumine is a novel gadolinium-based contrast 

agent with T1 relaxivity in blood (9.7 mM -1 s -1 ) greater 

than that of other available gadolinium agents due to weak 

and transient interaction with serum albumin. This property 

might improve the visualization of small or poorly-enhancing 

lesions. 


The enhancement mechanism of gadobenate dimeglumine 

and the principal results from studies in patients with CNS 

disease were reviewed. The published studies with this agent 

include a multicenter trial in patients with CNS metastases in 

which gadobenate dimeglumine was administered at cumulative 

doses up to 0.3 mmol/kg, two multicenter controlled 

trials in which gadobenate dimeglumine at cumulative doses 

of 0.15 and 0.2 mmol/kg was compared with gadodiamide at 

a cumulative dose of 0.3 mmol/kg, and a crossover study in 

which gadobenate and gadopentetate were given to the same 

patients at a dose of 0.1 mmol/kg. The effect of gadobenate 

dimeglumine on signal intensity, lesion enhancement, and 

patient management is discussed. 

RESULTS 

The principal finding from the study in metastatic disease is 

that a dose of 0.1 mmol/kg gadobenate dimeglumine is sufficient 

for most clinical situations but that an increased dose

of 0.2 mmol/kg provides additional information in certain 

cases (e.g., for the detection of additional metastases). A 

cumulative dose of 0.3 mmol/kg gadobenate dimeglumine 

did not provide information beyond that available with a 

cumulative dose of 0.2 mmol/kg. The controlled trials further 

showed that a cumulative dose of 0.2 mmol/kg gadobenate 

dimeglumine was diagnostically equivalent to a cumulative 

dose of 0.3 mmol/kg gadodiamide. In these trials, 0.1 

mmol/kg gadobenate dimeglumine was shown to provide 

higher S/N than an equivalent dose of gadodiamide in tumor 

patients. These findings were confirmed in a recent 

crossover study in which enhancement after 0.1 mmol/kg 

gadobenate dimeglumine was shown to be superior to 

enhancement after the same dose of gadopentetate dimeglumine. 

Blinded readers preferred the gadobenate images in 21 

of 24 patients, finding the enhancement equivalent in the 

remaining cases. Preliminary evidence suggests that 

improved relaxivity agent also may be of benefit for the 

imaging of specific vascular pathologies such as AVMs. 

CONCLUSION 

Gadobenate dimeglumine provides superior enhancement to 

other gadolinium agents in enhancing lesions at a dose of 0.1 

mmol/kg. The increased relaxivity of this agent compared to 

other agents may be of significant benefit when a greater 

contrast effect is desired. 

KEY WORDS: Contrast agents, relaxivity, metastatic disease 

Poster 34 

Diffusion-Weighted Images of Brain Metastasis: 

Comparison with Histologic Type and Cellularity 

Morishita, S. 1 · Hirai, T. 1 · Kitajima, M. 1 · Hayashida, Y. 1 · 

Yamura, M. 1 · Korogi, Y. 2 · Yamashita, Y. 1 · Kuratsu, J. 1 

1 Kumamoto University Graduate School of Medical 

Sciences, Kumamoto, JAPAN, 2 University of Occupational 


JAPAN 

PURPOSE 

To assess the relationship between signal intensity on diffusion-weighted 

images and histologic types in brain metastasis, 

and to determine whether the apparent diffusion coefficient 

(ADC) values correlate with tumor cellularity. 


We retrospectively assessed diffusion-weighted images and 

ADC maps in 20 brain metastatic lesions of 20 consecutive 

patients, 10 of whom underwent surgical resection of the 

lesion. The malignant lesions included lung (n = 14), ovarian 

(n = 1), esophageal (n = 1), breast (n = 1), and uterine corpus 

cancer (n = 1), and 2 soft tissue sarcomas. The histologic 

types were adenocarcinoma (well, n = 5; moderately, n = 

3; poorly differentiated, n = 3), large cell carcimomas (n = 2) 

and small cell carcinomas (n = 2), squamous cell carcinoma 

(n = 2), sarcoma (n = 2), and 1 mucoepitheloid carcinoma. 

Two radiologists qualitatively compared the signal intensity 

of the lesion with surrounding normal-appearing white matter 

on diffusion-weighted imaging. Regions of interest were 

drawn on areas of enhancing lesions on each image and ADC 

values were measured on ADC maps. Tumor cellularity was 

defined as the total area of tumor cell nuclei divided by the 

303 

area of the histologic section. In 10 lesions of 10 patients 

who underwent surgery, the ADC values were compared 

with tumor cellularity using the Spearman rank correlation. 

RESULTS 

According to the qualitative assessment, all 5 well differentiated 

adenocarcinomas had low signal intensity, and both 

small cell carcinomas and both sarcomas had high signal 

intensity. The lesions of the other types showed slightly high 

or iso signal intensity. The ADC values tended to correlate 

with tumor cellularity, although there were no significant 

differences between them. 

CONCLUSION 

In enhanced areas of brain metastatic lesions, diffusionweighted 

imaging reveals different signal intensities according 

to the histologic type. Tumor cellularity in the brain 

metastatic lesions may contribute to the ADC values. 

KEY WORDS: Metastatic brain tumors, diffusion-weighted 

imaging 

Poster 35 

Evaluation of Primary and Metastatic Solitary Intraaxial 

Malignant Brain Tumors by Dynamic Susceptibility 

Contrast-Enhanced Perfusion MR Imaging: Results of a 

Clinical Study 

Bulakbasi, N. · Kocaoglu, M. · Guvenc, I. · Ucoz, T. · Tayfun, 

C. · Somuncu, I. 

Gulhane Military Medical Academy 


PURPOSE 

Relative cerebral blood volume (rCBV) has long been used 

for grading of glial tumors (1, 2). We aimed to evaluate efficacy 

of relative rCBV measurement in the preoperative grading 

and differentiation of the solitary intraaxial malignant 

brain tumors. 


Twenty one low-grade (6 pilosytic, 8 diffuse astrocytoma, 

and 7 oligodendroglioma), 21 high-grade (7 anaplastic astrocytoma, 

7 anaplastic oligodendroglioma, 7 glioblastoma 

multiforme) glial tumors and 16 metastases in 58 patients 

were evaluated prospectively by conventional contrastenhanced 

MR imaging and dynamic susceptibility contrastenhanced 

T2* weighted echo-planar imaging during first 

pass of a bolus injection of 0.2 mmol/kg contrast material. A 

small dose (5-7 mL) of contrast material was given intravenously 

to presaturate the tumoral interstitium (2). Ten sections 

were obtained without intersection gap to cover the 

entire lesion volume. A series of 50 multisection acquisitions 

was acquired at 1.9 second intervals. The highest CBV value 

of five uniform ROIs (maximum 5 pixels), which had been 

measured from different areas showing the greatest visually 

identifiable CBV values on color maps, was selected. 

Normalized tumoral rCBV values were calculated by standard 

software. All rCBV values were compared with independent 

samples T test assuming unequal variances and 

Pearson statistical analysis was used to correlate degree of 

malignancy (as low- or high-grade) or tumor types with 

rCBV values. The mean difference was significant at the 

Posters

Posters 

p < 0.05 level. The binormal receiver operating characteristic 

(ROC) curves for rCBV values were constructed to distinguish 

cutoff values. 

RESULTS 

The mean differences of rCBV values between low- (1.71 ± 

0.59) and high-grade tumors (5.01 ± 1.03) (p < 0.000); highgrade 

and metastases (2.45 ± 0.70) (p < 0.000); and lowgrade 

and metastases (p < 0.002) were significant. No clear 

cutoff value was present. The clear rCBV cutoff value of 1.9 

for differentiation low- (1.44 ± 0.23) vs high-grade (4.88 ± 

1.21) astrocytomas, 3.4 for differentiation low- (2.25 ± 0.73) 

vs high-grade (5.28 ± 0.48) oligodendrogliomas, and 3.8 for 

differentiation glioblastoma multiforme (5.72 ± 0.90) vs 

metastases was detected with the sensitivity of 100%, the 

specificity of 100%, and the accuracy of 100%. Although the 

mean differences between anaplastic astrocytoma (4.04 ± 

0.83) and anaplastic oligodendroglioma (p < 0.000); 

anaplastic astrocytoma and metastases (p < 0.000); and lowgrade 

astrocytoma and metastases (p < 0.05) were significant, 

no clear cutoff value was detected for exact differentiation. 

The rCBV values were correlated with degree of 

malignancy (r = 0.869, P < 0.000) and with tumor type (r = 

0.823, P < 0.000). The overall efficacy of rCBV calculation 

was higher in grading of glial tumors than in differentiation 

of high-grade glial tumors from metastases. 

CONCLUSION 

DSC perfusion imaging with rCBV calculation seemed quite 

effective in both the grading of malignant glial brain tumors 

and in the differentiation of high-grade glial tumors from 

solitary intraaxial metastasis. The astrocytomas and oligodendrogliomas 

have to be evaluated separately for exact 

grading. 

REFERENCES 

1. Aronen HJ, Gazit IE, Louis DN, et al. Cerebral blood volume 

maps of gliomas: comparison with tumor grade and histologic 

findings. Radiology 1994;191:41-51 

2. Lev MH, Ozsunar Y, Henson JW, et al. Glial tumor grading 

and outcome prediction using dynamic spin-echo MR susceptibility 

mapping compared with conventional contrastenhanced 

MR: confounding effect of elevated rCBV of oligodendrogliomas 

[corrected]. AJNR Am J Neuroradiol 

2004;25:214-221 

KEY WORDS: Brain-neoplasm, perfusion imaging, relative 

cerebral blood volume 

Poster 36 

Correlation of MR Perfusion Imaging and Vessel 

Tortuosity Parameters in Assessment of Intracranial 

Neoplasms 

Parikh, A. H. · Smith, J. K. · Ewend, M. · Bullitt, E. 

University of North Carolina at Chapel Hill 


PURPOSE 

To determine the correlation between tumor vessel tortuosity 

as measured from vessels extracted from MR angiograms 

(MRA) and perfusion parameters of cerebral blood flow 

(CBF) and cerebral blood volume (CBV) in intracranial neo- 

304 

plasms. We hypothesized that tumor blood vessel tortuosity 

measures and perfusion measures would be correlated, since 

both are increased by tumor angiogenesis. 


Eighteen patients with 19 cerebral neoplasms were evaluated 

with conventional MR imaging and dynamic contrastenhanced 

T2-weighted perfusion MR imaging. Both benign 

and malignant lesions were included, as were hyper and 

hypovascular tumors. Regions of interest were plotted within 

the tumor area to locate foci of maximum CBV and CBF. 

Cerebral blood volume and CBF measurements were recorded 

also in contralateral normal appearing white matter to calculate 

relative CBV (rCBV) and relative CBF (rCBF). Vessel 

tortuosity analyses were conducted upon vessels segmented 

from MRA images of the same patients using two tortuosity 

descriptors (SOAM and ICM) which have been demonstrated 

previously to have efficacy in separating benign from 

malignant disease. Linear regression analyses were conducted 

to determine if correlations exist between CBV or CBF 

and the two tortuosity measurements. 

RESULTS 

For the overall set of tumors, no significant correlations were 

found between flow or volume measures and the tortuosity 

measures. However, when the 7 glioblastoma multiforme 

tumors were examined as a subgroup, the following significant 

correlations were found: rCBV and SOAM (R 2 = 0.799), 

rCBV and ICM (R 2 = 0.214). 

CONCLUSION 

Our results demonstrate that MR perfusion imaging data do 

not correlate significantly with vessel tortuosity parameters 

as determined from the larger vessels seen by MRA. 

However, for subgroups of a particular tumor type such as 

GBM, there may be significant correlations. It appears that 

perfusion and tortuosity data may provide independently 

useful data in the assessment of cerebral neoplasms. 

KEY WORDS: Intracranial neoplasm, MR perfusion imaging, 

vessel tortuosity 

Poster 37 

Relative Cerebral Blood Volume Ratios in the following 

of Posttherapeutic Glial Tumors 

Bulakbasi, N. · Ilkbahar, S. · Guvenc, I. · Kocaoglu, M. · 

Tayfun, C. · Ucoz, T. 

Gulhane Military Medical Academy 


PURPOSE 

We aimed to determine the role of dynamic contrastenhanced 

perfusion MR imaging, using normalized relative 

blood volume (rCBV) in the evaluation of postoperative and 

radiotherapy-related changes of patients with glial tumors. 


Twenty-four patients (13 male, 11 female, age range from 21 

to 67 years, mean 36.3 years) treated by surgery and radiotherapy 

with the diagnosis of glial tumors were included in 

this study. MR imaging was performed using a 1.5 T superconductive 

magnet with standard head coils. Conventional

MR imaging and dynamic contrast-enhanced MR imaging 

were performed at same period. Perfusion MR imaging was 

performed with standard dynamic susceptibility contrastenhanced 

T2*-weighted gradient-echo echo-planar imaging 

during first pass of a bolus injection of 0.2 mmol/kg contrast 

material, 3-30 months (mean 12 months) after the therapy. 

Ten sections were obtained without intersection gap to cover 

the entire lesion volume. A series of 50 multisection acquisitions 

was acquired at 1.9 second intervals, the total acquisition 

time being 1 min 37 s. The highest CBV value of five 

uniform ROIs (max 5 pixels), which had been measured 

from different areas showing the greatest visually identifiable 

CBV values on color maps, was selected. Normalized 

tumoral rCBV values were calculated by standard software 

and statistically tested independently. Presence of recurrence/residual 

tumor or postradiotherapy changes were 

described based on the histopathology or clinical and radiological 

follow up. All rCBV values were compared with 

independent samples T test assuming unequal variances. The 

mean difference was significant at the 0.05 level. The binormal 

receiver operating characteristic (ROC) curves for rCBV 

values were constructed to distinguish cutoff values. 

RESULTS 

Fourteen of 24 (58%) patients had a diagnosis of radiation 

necrosis. One of these patients was diagnosed by stereotactic 

biopsy. No surgical treatment was applied to 13 patients and 

they were diagnosed by clinical and radiologic follow up. Ten 

of 24 (42%) patients had residual/recurrent tumors. Four of 

them were diagnosed histopathologically, while 6 cases were 

diagnosed with clinical and radiologic data. The mean rCBV 

value of radiation necrosis was 0.90 ± 0.34. The mean rCBV 

value of recurrent tumor was 3.45 ± 1.95. The difference 

between rCBV ratios of residual-recurrent tumor and radiotherapy 

necrosis was statistically significant (p < 0.001). 

When the threshold value of rCBV was 1.35, sensitivity was 

86% and specificity was 90% (≥ 1.35 residual/recurrent 

tumor, < 1.35 radiotherapy-induced changes). 

CONCLUSION 

We suggest that rCBV maps are complementary technique to 

conventional MR imaging in patients with treated gliomas 

and allow differentiating recurrence/residual tumor from 

radiation necrosis. 

KEY WORDS: Tumor recurrence, radiotherapy-related 

changes, perfusion imaging 

305 

Poster 38 

Differential Cortical Thickness of the Motor and Sensory 

Cortex on T2-Weighted MR Imaging in the Presence of 

Vasogenic Edema: A Confirmatory Method for 

Identifying the Central Sulcus 

Biega, T. J. 1 · Lonser, R. L. 2 · Butman, J. A. 3 

1 George Washington University, Washington, DC, 2 National 

Institute of Neurological Disorders and Stroke, National 

Institutes of Health, Bethesda, MD, 3 Warren G. Magnuson 

Clinical Center, National Institutes of Health, Bethesda, MD 

PURPOSE 

Identification of the motor strip may be compromised by the 

presence of mass effect and edema, as sulcal landmarks are 

often obscured. We demonstrate that the central sulcus may 

be identified by the differential thickness between motor and 

sensory cortices which is readily apparent on T2-weighted 

images in the presence of vasogenic edema. 


T2-weighted 5 mm axial sections aligned along the AC-PC 

plane were obtained at 1.5 or 3.0 T using 22 cm FOV interpolated 

to a 512 2 matrix (0.43 mm resolution) from 14 patients 

with vasogenic edema due to lesions in the frontal and parietal 

lobes. Linear cortical thickness measurements were made perpendicular 

to and on opposing banks of sulci to measure differential 

cortical thickness across the central sulcus (210 

measurements) and across other frontal and parietal sulci (122 

measurements) arbitrarily selected based on the pattern of 

vasogenic edema. Location of the central sulcus was confirmed 

utilizing standard anatomical landmarks in all cases, 

and with intraoperative cortical mapping in two cases. 

RESULTS 

Differences in cortical thickness between the pre and postcentral 

sulci were readily apparent on T2-weighted images in 

the presence of vasogenic edema, even when there was 

marked distortion of gross anatomy due to mass effect (Fig. 

1). The motor strip (precentral) was much thicker [3.3 ± 0.8 

mm (mean ± SD)] than the sensory strip (postcentral) (1.7 ± 

0.6 mm). The anterior cortical bank of “noncentral” sulci 

was 2.4 ± 0.4 mm, comparable in thickness to the posterior 

bank (2.2 ± 0.5 mm). The ratio of the thickness of the two 

opposing cortical banks adjacent to the central sulcus (2.0 ± 

0.6) was significantly larger than the ratio of cortical thickness 

adjacent to “noncentral” sulci (1.1 ± 0.2) (p < 0.001 

paired t-test). 

Posters

Posters 

CONCLUSION 

Because of cytoarchitectonic differences in the motor and 

sensory cortices, the anterior bank of the central sulcus 

(motor strip) is significantly thicker than the posterior bank 

of the central sulcus (sensory strip) (1). The presence of 

vasogenic edema in pathologic states makes this phenomenon 

readily demonstrable on routine T2-weighted imaging, 

precisely when mass effect and edema make identification of 

the motor strip most difficult. Neuroradiologists should be 

aware of this method to confirm localization of the motor 

strip, in addition to well known methods based on sulcal 

anatomy (2-3). 

REFERENCES 

1. Meyer JR, et al. AJNR Am J Neuroradiol 1996;17(9):1699-1706 

2. Kido DK, et al. Radiology 1980;135(2):373-377 

3. Naidich TP. Neuroradiology 1991;33(Suppl):S95-S99 

KEY WORDS: Tumor 

Poster 39 

Do Cerebral Blood Volume and Contrast Transfer 

Coefficient Predict Prognosis in Human Glioma? 

Jackson, A. 1 · Patankar, T. 1 · Haroon, H. 1 · Baleriaux, D. 2 · 

Mills, S. 1 

1University of Manchester, Manchester, UNITED KING- 

2 DOM, Hôpital Erasme, Cliniques Universitaires de 

Bruxelles, Bruxelles, BELGIUM 

PURPOSE 

Advances in dynamic contrast-enhanced MR imaging allow 

generation of parametric maps of physiologicl variables 

including cerebral blood volume (CBV) and transfer coefficient 

(K trans ). In gliomas CBV values correlate to grade 

although the relationship with K trans is less clear. This study 

explores the relationship between CBV and K trans and prognosis 

is related to grade. 


Twenty-seven patients with clinical gliomas were recruited. 

MR imaging including dynamic contrast-enhanced MR 

imaging was performed at diagnosis and tumor grade was 

determine using criteria set out by the World Health 

Organization (WHO) (1). Dynamic images were obtained 

and analyzed using a previously published and validated 

technique (2). Parametric variables were expressed as median 

values from regions of interest including all enhancing 

tumors. Differences between parametric values tested using 

an analysis of variance (ANOVA) with a-posteriori pairwise 

testing. Values of Kfp and CBV each were used to categorize 

patients into four groups and survival compared using 

Kaplan-Meier plots and Log- rank analysis. The predictive 

value of individual parameters was assessed using a Cox 

regression model. 

RESULTS 

Cerebral blood volume and Kfp showed a clear trend to 

increase with increasing grade and there were significant 

group differences for both Kfp (p < 0.01) and CBV (p < 0.01, 

ANOVA). Post-hoc tests (Tamhane) showed significant differences 

in Kfp between grade 2 and grade 3 (p < 0.01) and 

between grade 2 and grade 4 (p < 0.001) but not between 

306 

grade 3 and grade 4. There were also significant differences 

in CBV between grade 2 and grade 4 (p < 0.001) but not 

between grade 2 and grade 3 or between grade 3 and grade 

4. Data are summarized in Fig 1 (left). Log rank tests showed 

significant differences in survival for histologic grade (p < 

0.0001), Kfp (p = 0.005) and CBV (p = 0.006; Fig 2). Log 

rank tests for grade 4 tumors showed no residual significant 

differences between groups defined by CBV. However, there 

was a significant difference in survival in patients with grade 

4 tumors between those in Kfp groups 2, 3, and 4 (p < 0.01). 

Fig 1 (right) shows the Kaplan Meir plot for these three 

patient groups in the grade 4 tumor population. Cox regression 

analysis of individual variables demonstrated significant 

relationships between overall survival and histologic 

grade (p < 0.001), CBV (p = 0.004), and Kfp (p = 0.008). The 

full forward stepwise regression analysis showed independent 

relationships between survival and histological grade (p 

= 0.002) and Kfp (p = 0.03). 

CONCLUSION 

Histologic grade remains the most significant prognosticator 

of survival and its effect is predicted by measurement of 

CBV. K trans relates to grade but contains significant additional 

prognostic information to grade, particularly in high grade 

tumours. 

KEY WORDS: Neoplasms, adult brain 

Poster 40 

Dose-Effectiveness of Gadobenate Dimeglumine in MR 

Imaging of Brain and Spine Metastatic Disease 

Runge, V. M. 1 · Ross, J. S. 2 · Nelson, K. L. 3 

1Scott and White Clinic and Hospital, Temple, TX, 

2 3 Cleveland Clinic Foundation, Cleveland, OH, Methodist 

Hospital, Omaha, NE 

PURPOSE 

Gadobenate dimeglumine (MultiHance) is a gadoliniumbased 

contrast agent with approximately twice the T1 relaxivity 

in plasma of conventional gadolinium chelates. This 

study was conducted to determine whether the increased 

relaxivity permits a lower overall dose to be used for detection 

of CNS metastases. 


One hundred and forty-nine adult patients with CNS metastases 

received gadobenate dimeglumine at a dose of either 

0.05 (n = 74) or 0.1 mmol/kg (n = 75). T1- and T2-weighted 

images were acquired predose followed by T1-weighted 

images at 10 min postdose. Quantitative evaluations of 

lesion-to-background ratio (L/B), contrast-to-noise ratio 

(C/N), and percent enhancement ( %EN) were performed on

pre- and postdose T1-weighted images. Three fully blinded 

neuroradiologists qualitatively evaluated in randomized 

order predose images alone and pre- + postdose images combined. 

Five-point scales from 0 = none to 4 = excellent were 

used to compare the two doses in terms of changes from predose 

in lesion border delineation, visualization of internal 

lesion morphology, and contrast enhancement of lesions. 

Individual predose to predose + postdose changes and differences 

between groups were evaluated using paired t-tests 

and analysis of covariance (ANCOVA), respectively. 

Interreader agreement was evaluated using intraclass correlation 

coefficients. The Wilcoxon Signed Rank test was used 

to test the difference from predose to predose + postdose in 

the numbers of lesions detected in each dose group while the 

Wilcoxon Rank Sum statistic was used to test for differences 

between groups. 

RESULTS 

Quantitative evaluation revealed significant (p < 0.001) 

increases postdose for each quantitative parameter (L/B, 

C/N, and %EN) and consistently higher values for the 0.1 

mmol/kg group than for the 0.05 mmol/kg group. Pre- to pre- 

+ postdose increases in score were significant (p < 0.001) for 

each reader for all qualitative parameters. All three blinded 

readers noted greater increases for the 0.1 mmol/kg group 

compared to the 0.05 mmol/kg group. Two readers detected 

significantly (p < 0.05) more lesions postdose in the 0.1 

group compared to the 0.05 group for all patients considered 

together and for patients with 0-2 lesions detected on predose 

images. The increases from predose in the number of 

detected lesions were 53.8%, 53.8%, and 27.7% (readers 1, 

2 and 3, respectively) for the 0.05 mmol/kg dose, and 95.8%, 

78.6%, and 92.1% for the 0.1 mmol/kg dose. The proportions 

of patients with increases in lesion counts were consistently 

higher with the 0.1 mmol/kg dose, a finding that was 

significant (p < 0.005) for readers 1 and 3. 

CONCLUSION 

Reducing contrast dose below 0.1 mmol/kg with gadobenate 

dimeglumine, despite its two-fold improved relaxivity, is not 

recommended in the evaluation of CNS metastatic disease. 

The 0.1 mmol/kg dose provides significantly higher L/B 

ratio and %EN compared to the 0.05 mol/kg dose and permits 

improved lesion delineation and visualization of internal 

lesion morphology. While the higher relaxivity of Gd- 

BOPTA allows use of a 0.05 mmol/kg dose for the detection 

of CNS metastases, a 0.1 mmol/kg dose improves further 

lesion detection. 

KEY WORDS: Metastatic disease, MR imaging, brain 

307 

Poster 41 

Discriminating Brain Tumor Recurrence from Radiation 

Injury Using 2D CSI MR Spectroscopy 

Petrou, M. · Sundgren, P. C. · Weybreight, P. · Nan, B. · 

Foerster, B. · Maly, P. 


Ann Arbor, MI 

PURPOSE 

To explore the ability of 2D CSI MR spectroscopy in differentiating 

radiation injury versus tumor recurrence in patients 

previously treated for brain neoplasm who presents with new 

contrast-enhancing lesion(s) at the site or vicinity of their 

primary tumor. 


Two-diemnsional CSI MR spectroscopy (PRESS, TE 

144/TR 1000) was performed in 21 patients (11 male, 8 

female aged 4-54 years, mean 33.4 years). Each patient had 

newly found contrast-enhancing lesions in the pons, posterior 

fossa, or supratentorial region at the site of a previously 

diagnosed and treated primary brain neoplasm. Histologic 

types included glioma, astrocytoma, ependymoma, and 

medulloblastoma. Patients were selected based on request by 

the referring clinician for spectroscopy to specifically evaluate 

for tumor recurrence versus radiation injury. 

Determination of tumor recurrence (11 patients) or necrosis 

(8 patients) was based on clinical course, subsequent imaging 

characteristics, and stereotactic biopsy/surgery. Within 

the VOI, smaller voxels (1 x 1 x 1 cm) were manually placed 

in the contrast-enhancing lesion and in normal-appearing 

white matter on the contralateral hemisphere. Normalization 

was achieved by placing the contralateral white matter value 

for a given metabolite in the denominator, which led to the 

calculations of the following ratios: Cho/normal Cr(n) , 

Cr/Cr(n), NAA/Cr(n), Cho/Cho(n), NAA/NAA(n). In addition 

Cho/NAA and NAA/Cho ratios in the contrast-enhancing 

lesion were calculated. 

RESULTS 

The mean Cho/Cr(n) and Cho/Ch(n) ratios were significantly 

higher in patients with tumor recurrence compared to 

those with radiation injury ( 1.91 versus 1.12, p = 0.002) 

(1.57 versus 0.99, p = 0.005), respectively. The mean 

NAA/NAA(n) ratio was significantly lower in patients with 

tumor recurrence compared to those with radiation injury, 

0.41 versus 0.6, p = 0.07. No significant differences between 

the two groups were seen in the mean NAA/Cr(n) or 

LL/Cr(n) ratios. The mean Cho/NAA ratio was significantly 

higher in patients with tumor recurrence compared to those 

with radiation (3.92 versus 1.44, p = 0.004) and the mean 

NAA/Cho ratio was significantly lower in patients with 

tumor recurrence compared to those with radiation injury 

(0.33 versus 0.78, p < 0.0001). 

CONCLUSION 

The significant differences in the mean Cho/Cr(n), 

Cho/Cho(n), NAA/NAA(n), Cho/NAA and NAA/Cho ratios 

indicate the potential for spectroscopy to aid in differentiation 

of recurrent neoplasm from radiation injury in the evaluation 

of recurrent contrast-enhancing lesions. While overlap 

between ratios does exist, part of the overlap may be due 

to histologic heterogeneity and volume averaging within the 

Posters

Posters 

voxel of normal and neoplastic tissue or of mixed neoplastic 

and radiation injured tissue. Nonetheless, the significant differences 

in the trends of these ratios between the two groups 

in conjunction with conventional imaging may enable discrimination 

among recurrent neoplasm from radiation injury. 

KEY WORDS: Spectroscopy, radiation, neoplasm 

Poster 42 

Arteriovenous Shunting and Steal Effects in High-Grade 

Glioma Displayed by Dynamic Susceptibility-Enhanced 

MR Imaging 

Ulmer, S. · Liess, C. · Otto, N. · Engellandt, K. · Jansen, O. 

University Hospital of Schleswig-Holstein 

Kiel, GERMANY 

PURPOSE 

Pathomechanisms of the genesis of glioblastoma, which are 

characterized by nuclear atypia, mitosis, necrosis and 

microvascular proliferation, are not well understood. 

Conventional MR imaging can only display areas of a breakdown 

of the blood-brain barrier with a characteristic 

enhancement after intravenous contrast agent application. 

Dynamic susceptibility (DSC)-enhanced MR imaging has 

been used to map hemodynamic changes in both stroke and 

tumor enabling a differentiation between high- and lowgrade 

gliomas. In this study we investigated whether characteristic 

hemodynamic changes in glioblastoma can be found 

using DSC MR imaging using maps for relative regional 

blood flow (rCBF) and volume (rCBV) compared to the 

unaffected surrounding tissue. We hypothetize, that high 

malignant brain tumors produce a steal effect from the surrounding 

tissue to support its increased demand, which can 

be demonstrated by one or a combination of the previously 

mentioned parameters. 


Dynamic susceptibility-enhanced MR imaging was performed 

on a 1.5 T scanner (Siemens, Magnetom Vision, 

Erlangen, Germany) using a standard head coil in 15 patients 

with glioblastoma after written informed consent was 

obtained. A multislice T2*-weighted EPI-sequence (TR / TE 

= 2000 / 62 ms, FOV 240 m matrix 128 x 128, slice thickness 

6 mm, 40 images per slice) was used for hemodynamic 

mapping. After an initial baseline period of 8 images per 

slice 40 cc of contrast agent (Ga-DTPA, Schering, Berlin, 

Germany) was injected at a flow rate of 6 ml/sec. Before and 

after the PWI measurement, spin-echo T1-weighted images 

with identical slice positions were acquired. After determination 

of the arterial input function from the MCA, maps of 

relative rCBF and rCBV were created using a software program 

(Sorensen). 

RESULTS 

Highest values for rCBV and rCBF were found in the tumor 

margin. Values for rCBV and rCBF in the gray matter did not 

differ significantly between the affected (rCBVi; rCBFi)and 

the nonaffected hemisphere (rCBVc; rCBFc; paired t-test). 

There was a highly significant correlation between rCBVi 

and rCBVc (R2 = 0.94; p < 0.0001) and between rCBFi and 

rCBFc (R2 = 0.97; p < 0.001). Maximum values of both 

rCBVt and rCBFt were significantly increased compared to 

308 

the adjacent gray matter (p < 0.001) with a highly significant 

correlation between rCBVi and rCBVt in favor of rCBVt (R2 

= 0.64; p < 0.0005) and between rCBFi and rCBFt (R2 = 

0.58; p < 0,001) in favor of the tumor respectively. Relative 

rCBF and rCBV were elevated in the tumor. No significant 

difference was found between the values of the affected and 

nonaffected hemisphere. We always found a close correlation 

between the gray matter of the affected and nonaffected 

hemisphere as well as the affected hemisphere and the 

tumor, respectively. 

CONCLUSION 

Due to its increased demands, glioblastoma “steal” blood 

from the surrounding brain tissue with elevated rCBV that 

passes the tisue at a higher flow (rCBF). This is enabled by 

AV shunting of the neoangiogenetic vessels in the tumor 

with leakage of contrast agent through the destroyed bloodbrain 

barrier leading to the specific contrast enhancement 

found on T1-weighted images. As a tendency rCBF was 

slightly elevated in the affected hemisphere to compensate at 

a baseline rest condition. 

KEY WORDS: DSC MR imaging, glioblastoma, AV shunting 

Poster 43 

Role of Diffusion-Weighted MR Imaging and MR 

Spectroscopy in Differentiating Recurrent Gliomas from 

Radiation Necrosis 

Abdel Razek, A. A. · Elmogy, S. 



PURPOSE 

To evaluate the role of diffusion-weighted MR imaging and 

proton MR spectroscopy in differentiating recurrent gliomas 

from radiation necrosis. 


Diffusion-weighted MR imaging and proton multivoxel MR 

spectroscopy were done for 46 patients after complete course 

of radiotherapy for histologically proven glioma. Diffusion 

MR imaging was done using a single-shot echo-planar imaging 

(EPI) with a diffusion-weighted factor, factor b of 0,500 

and 1000 sec/mm2. The apparent diffusion coefficient 

(ADC) map was reconstructed with calculation of ADC values. 

Proton MR spectroscopy was performed using multivoxel 

chemical shift imaging (CSI). The acquisition parameters 

were: TR = 3000 msec, TE = 135 msec, 160 mm FOV, 

slice thickness of 10 mm and 2 averages. Final diagnosis was 

established by histopathology (n = 34) or clinical course (n 

= 12). The ADC values and metabolite signals correlated 

with histopathologic results and statistical analysis was 

done. 

RESULTS 

The mean ADC value of the recurrent gliomas (1.13 ± 0.17 

X 10-3 mm2/sec) was statistically different (p < 0.001) than 

that of radiation necrosis (1.89 ± 0.11 X 10-3 mm2/sec). 

Good MRS spectra were obtained in 42 patients. Proton MR 

spectra of recurrent gliomas showed significantly high 

amount of choline compared with that of radiation necrosis. 

Broad band of lipids, lactate and amino-acids were seen in

patients with radiation necrosis. The mean Choline/creatine 

ratio was significantly higher than in recurrent gliomas (P < 

0.003) with overlap in N-acetyl-aspartate/creatine ratio (P 

4 separate lesions. 

Thirty-three patients did not show Gd enhancement. In 2 

patients, perfusion-weighted imaging showed increased 

CBV values prior to Gd enhancement and anaplastic evolution. 

Twenty-four patients developed a glioblastoma (GBM): 

3 are still alive (follow-up range, 18-36 months after first 

MR imaging). Twenty-one had a final diagnosis of anaplastic 

astrocytoma or oligodendroglioma: 5 are still alive (15-49 

months). Eight had a final diagnosis of low-grade astrocytomas, 

oligodendrogliomas, or mixed gliomas: 6 are alive, 2 

remains unchanged (6-13 years). 

Posters 

Posters

Posters 

CONCLUSION 

True MG and GC are more frequent than generally considered. 

In most cases, a differential diagnosis is not possible by 

MR imaging. Despite poor prognosis in most patients, there 

is a considerable percentage of low-grade tumors. Early 

recognition of anaplastic evolution and/or evolution toward 

GBM by MR imaging or perfusion-weighted imaging may 

allow prognosis improvement. 

KEY WORDS: Multicentric brain gliomas, gliomatosis cerebri, 

MR imaging 

Poster 46 

Grading of Cerebral Gliomas with MR Spectroscopy and 

Perfusion Imaging 

Incesu, L. · Diren, B. · Gelmez, S. · Ture, U. 

Ondokuz Mayis University, Samsun, TURKEY 

PURPOSE 

Thirty percent to 40% of all intracranial tumors are gliomas. 

Of these about 50% are glioblastoma multiforme (GBM). It 

is important that gliomas especially high-grade ones 

(anaplastic astrocytoma and GBM) are heterogeneous characteristics. 

Some parts of the tumor can be of high grade 

whereas some, the other parts may be of low grade based on 

the histopathologic characteristics. Conventional MR imaging 

is a valuable and widely used tool in the diagnosis of 

intracranial gliomatous tumors but it is insufficient in predicting 

glioma grade by itself. Advanced MR imaging techniques 

such as MR spectroscopy and MR perfusion, show 

different aspects of tumoral characteristics, have been shown 

to correlate reliably with tumour grade determined by 

histopathologic methods. We evaluated the use of both of 

these advanced techniques in glial tumor grading and compared 

them in voxe- to-voxel basis. 


Twenty-one GBM patients were evaluated by both of these 

techniques. Imaging was performed with a 1.5 T MR unit 

(Symphony; Siemens AG, Erlangen, Germany). Multivoxel 

2D proton chemical shift imaging (1500/135 TR/TE) which 

uses a point-resolved spectroscopy sequence with a volume 

of interest (VOI) of 8 x 8 region placed within 16 x 16 field 

of view on a 1 cm transverse section was performed for each 

patient. After that dynamic contrast-enhanced T2*-weighted 

echo-planar images were acquired during the first pass of a 

standart dose ( 0.1 mmol/kg) bolus of Gd-DTPA. To perform 

a comparative analysis with spectroscopy, slice positioning 

was performed carefully to include the slice on which spectroscopy 

examination was performed. Because it is a well 

known fact that glioma cells are found in both tumoral and 

peritumoral edematous regions, the voxels that comprise the 

tumor and the peritumoral edema seen hyperintense in T2weighted 

and FLAIR images were analyzed with both techniques. 

Grading was performed based on relative cerebral 

blood volume (rCBV) and metabolite ratios derived from 

MR spectroscopy and perfusion data, respectively. Choline 

(Cho)/Creatinine (Cre and Cho/N-acetyl-aspartate(NAA) 

ratios were used for spectroscopic analysis. Grading was 

performed on two-point scale (low grade and high grade) for 

each voxel by the established criteria for rCBV or metabolite 

ratios. In determination of tumor grade both these techniques 

310 

were analyzed comparatively on voxel-to-voxel basis. 

Tumoral mapping was performed for the slice of interest by 

combining the data derived from both techniques. 

RESULTS 

A total of 1324 voxels were evaluated. Tumoral and peritumoral 

edema comprised 280 voxels and comparative analysis 

was performed on that 280 voxels. Cho/Cre and 

Cho/NAA elevation was found in 182 and 186 voxels 

respectively. In 160 of them both of these ratios were elevated. 

In 214 voxels rCBV was elevated. Kappa analyses were 

performed to evaluate the voxel grading results of these two 

tests. 

CONCLUSION 

This study suggests that rCBV maps and MR spectroscopy 

are complimentary techniques that may help the determination 

of grade of gliomatous tumors preoperatively. 

Combination of data derived from both of these techniques 

can be used to generate tumoral grading maps which can be 

used to guide biopsies and be helpful for the surgeon to 

determine which parts of the tumor should be preferentially 

removed in the operations that primarily aim debulking. 

KEY WORDS: Glioma, MR spectroscopy, perfusion imaging 

Poster 47 

Absolute Quantification of Cerebral MR Perfusion Using 

Gd-BOPTA: Serial Volunteer and Patient Studies 

Essig, M. 1 · Le-Huu, M. 1 · Kirchin, M. 2 · Schoenberg, S. 3 · 

Reith, W. 4 · Lodemann, K. 5 

1German Cancer Research Center, Heidelberg, GERMANY, 

2 3 Bracco Diagnostics, Milano, ITALY, University of Munich, 

4 Munich, GERMANY, University of Homburg, 

5 Homburg/Saar, GERMANY, Bracco-Altanapharma, 

Konstanz, GERMANY 

PURPOSE 

Perfusion MR imaging has proven to be effective for the 

assessment of cerebrovascular diseases, brain tumors, and 

dementia. For therapy monitoring an absolute or “semiabsolute” 

quantification is mandatory. A crucial aspect for the 

quality the quantification of MR perfusion is the arterial 

input function which has to be robust and reliable and the 

amount of the relative signal reduction, which is the basis for 

the calculation of CBV and CBF. Both requirements depend 

on the concentration and the T2- relaxivity of the contrast 

agents used. The present study was conducted to evaluate the 

usefulness of Gd-BOPTA (MultiHance®, Bracco 

Diagnostics, Princeton, NJ) for quantitative cerebral perfusion 

MR imaging. 


A blinded randomized intraindividual comparative study 

was conducted in 12 healthy male volunteers and 30 patients 

with cerebral gliomas. The imaging parameters, slice positioning, 

and contrast media application were standardized. 

Volunteers were examined with single and double dose, 

patients with a single dose of Gd-BOPTA. For a quantitative 

assessment of the signal time curve, the percentage signal 

drop and the full width half maximum were assessed in a 

ROI analysis with ROIs in normal gray and white matter.

Beside that, the rCBV and rCBF values of gray and white 

matter were calculated and compared in-between the different 

imaging times. For the quantitative analysis the image 

quality of the rCBV and rCBF maps in respect of delineation 

of clinical utility, gray and white matter, and basal ganglia 

delineation were evaluated in an independent off-site assessment. 

RESULTS 

The studies revealed that a single dose of Gd-BOPTA was 

sufficient to achieve high quality MR perfusion data. The 

mean signal drop of approximately 32% was sufficient to 

achieve a high quality input function. With the used software 

package, the calculated rCBV and rCBF absolute values of 

the different ROIs were constant for both dosages within one 

subject. There were marginal differences in-between the subjects. 

At the qualitative assessment both readers found the 

single-dose images well suited for gray-white matter differentiation 

and the delineation of the basal ganglia. In the 

patient study performed with single dose of Gd-BOPTA all 

perfusion scans were of excellent diagnostic quality. The 

qualitative assessment presented a substantial clinical benefit 

from the perfusion scans in respect of lesion detection and 

lesion delineation. In the glioma patients a differentiation 

between low-grade and high-grade tumors was possible. 

CONCLUSION 

Our results proved that with an adequate software package 

and the use of the new generation contrast media 

MultiHance®, an absolute quantification and high quality 

paps of MR perfusion data is possible. The intraindividual 

values proved to be constant over time whereas the interindividual 

differences are objective. The achieved values are 

close to the expected physiologic data. The interindividual 

differences may arise from different input function quality; 

however, the intraindividual constancy is the key point for 

using this method for treatment monitoring. The main reason 

for the high quality perfusion possibilities is the higher susceptibility 

effect of MultiHance® which is stronger than for 

conventional MR contrast media. 

KEY WORDS: Perfusion MR imaging, contrast media, brain 

tumors 

Poster 48 

Diffusion Changes of the Brain Parenchyma in 

Operation Site on Early Postoperative Diffusion- 

Weighted MR Imaging 

Ozturk, A. · Karli Oguz, K. · Cila, A. · Akalan, N. 



PURPOSE 

To evaluate diffusion changes in the brain parenchyma in 

operation site in first 24 hours following surgery. 


Twenty-one patients (8 female, 13 male; between 8 months 

and 80 years) who underwent surgery were included in the 

study. Nineteen of the subjects had surgery because of an 

intracranial tumor and 2 subjects with Chiari-type 1 malformation 

had foramen magnum decompression surgery. 

311 

Postoperative MR imaging performed in first 24 hours 

included diffusion-weighted imaging (DWI) (TR/TE; 

2800/78; matrix: 256 x 256; b of 0, 500, 1000 s/mm 2 ) and 

routine contrast-enhanced cranial MR imaging including 

T2* imaging. Diffusion changes in the brain parenchyma 

were noted. Any hemorrhagic changes in the brain parenchyma 

that would interfere with true parenchymal diffusion 

changes were noted. We also aimed to determine any effect 

of preoperative tumor size, tumor pathology, presence of 

residual tumor following surgery on these diffusion abnormalities. 

RESULTS 

Twelve of the tumors were supratentorial and 7 were 

infratentorial. Pathologic examination of the tumors revealed 

2 oligodendroglioma, 3 meningioma, 2 medulloblastoma, 4 

epandymoma, 2 hemangioblastoma, 1 pilocytic astrocytoma, 

1 DNET, 2 PNET, 2 mixed ganglioglial tumors. We found 

decreased diffusion in the parenchyma near the resection 

cavity of 14 patients who had tumor surgery. Diffusionweighted 

imaging of one patient showed increased ADC values 

and no diffusion abnormality in 4 patients. Among 4 

patients with no diffusion abnormality, 2 patients had signal 

abnormalities related to hemorrhage in the operation site. In 

two patients who underwent decompression surgery, there 

was restricted diffusion in cerebellar tonsils. 

CONCLUSION 

Most common abnormality in early postoperative DWI of 

the parenchyma in the operation site following surgery was 

restricted diffusion suggesting cytotoxic edema while 

increased diffusion and no abnormality may be seen occasionally. 

One should be catious and use other sequences 

including T2* images when evaluating diffusion abnormality 

on the parenchyma because hemorrhage interferes with 

diffusion changes. 

KEY WORDS: Diffusion changes, brain, postoperative 

Poster 49 

Carcinomas and Lymphomas Involving Cavernous 

Sinus: Evaluation with Line Scan Diffusion-Weighted 

Imaging 

Maeda, M. 1 · Maier, S. E. 2 · Takeda, K. 1 

1 Mie University School of Medicine, Tsu, JAPAN, 2 Brigham 

and Women’s Hospital, Boston, MA 

PURPOSE 

Broad categories of diseases involving the cavernous sinus 

can cause multiple cranial neuropathies. Malignant tumors 

such as malignant lymphomas and carcinomas occasionally 

involve cavernous sinus, because these malignant tumors are 

common in head and neck and tend to spread into cavernous 

sinus. The purpose of this study was to evaluate apparent diffusion 

coefficient (ADC) of tumors involving cavernous 

sinus by line scan diffusion-weighted imaging (LSDWI) and 

to determine the usefulness of this method for differentiating 

between the two tumors. 

Posters

Posters 


Four patients with malignant lymphoma (diffuse large B 

cell) and five patients with carcinomas (four patients of 

squamous cell carcinoma and one patient of undifferentiated 

carcinoma) were studied prospectively. All patients presented 

with acute symptoms of cavernous sinus syndrome. The 

mean maximum diameter of tumors was 19 mm in malignant 

lymphoma and 25 mm in carcinoma. The MR study was performed 

prior to treatment in all patients. Line-scan diffusionweighted 

imaging was performed using b values of 5 and 

1000 sec/mm 2 , and effective section thickness/gap of 3/0.5 

mm. Line-scan diffusion-weighted images were obtained in 

the coronal plane focusing on cavernous sinus. Apparent diffusion 

coefficient maps were created. Apparent diffusion 

coefficient values were calculated in each tumor and were 

compared between malignant lymphomas and carcinomas. 

Mann-Whitney U test was applied to detect any significant 

differences in mean ADC values of the two. 

RESULTS 

Line-scan diffusion-weighhted imaging could provide 

acceptable images free from susceptibility artifacts and permit 

measurements of ADC values in all cases. Figure 1 

shows the ADC map of malignant lymphoma involving 

bilateral cavernous sinuses (ADC 0.53 x 10 -3 mm 2 /s). The 

ADC value (mean + SD) was 0.51 + 0.06 x 10 -3 mm 2 /s in 

malignant lymphomas and 1.01 + 0.06 x 10 -3 mm 2 /s in carcinomas. 

A significant difference in ADC values was found 

between the two (p < 0.01). 

CONCLUSION 

Line-scan diffusion-weighted imaging permits quantitative 

assessment of diffusion in malignant tumors involving cavernous 

sinus. Malignant lymphomas showed significantly 

lower ADC value than carcinomas. This method appears 

useful for differentiating between malignant lymphomas and 

carcinomas involving cavernous sinus. 

REFERENCES 

1. Lee JH, Lee HK, Park JK, et al. Cavernous sinus syndrome: 

Clinical features and differential diagnosis with MR imaging. 

AJR Am J Roentgenol 2003;181:583-590 

2. Delpassand ES, Kirkpatrick. Cavernous sinus syndrome as the 

presentation of malignant lymphoma: Case report and 

review of the literature. Neurosurgery 1988;23:501-504 

3. Maier SE, Gudbjartsson H, Patz S, et al. Line scan diffusion 

imaging Characterization in healthy subjects and stroke 

patients. AJR Am J Roentgenol 1988;171:85-93 

KEY WORDS: Cavernous sinus, neoplasm, diffusion-weighted 

imaging 

312 

Poster 50 

MR Imaging Characteristics Associated with Clinical 

Outcome after Gamma Knife Radiosurgery for 

Meningiomas 

Zhao, Y. · Kuo, J. S. · Zee, C. S. · Yu, C. · Go, J. L. · 

Giannotta, S. L. · Apuzzo, M. J. 

Keck School of Medicine, University of Southern 

California Los Angeles 


PURPOSE 

To identify MR imaging parameters that correlate with clinical 

outcome after meningioma radiosurgery. 


Pretreatment and follow-up MR examinations of 77 patients 

(26 male, 51 female) with intracranial meningiomas treated 

with gamma knife radiosurgery from September 1994 to 

June 2003 were analyzed for treated tumor size, tumor 

enhancement, surrounding edema, distance from major 

draining sinuses, venous congestion, and posttreatment radiation-induced 

injury. All patients were followed with serial 

neurologic exams and annual MR imaging for a median of 4 

years (range 1-9 years). Statistical analysis was used to identify 

anatomical and imaging factors that correlate with clinical 

outcome. 

RESULTS 

Meningioma radiosurgery is highly effective and safe in 

selected patients. Increased tumor size and proximity to 

major draining sinuses were positively correlated with the 

incidence of radiation-induced injury, as noted on follow-up 

imaging studies and clinical symptoms. Symptomatic radiation-induced 

injury required medical and/or surgical therapies. 

Other analyzed parameters were not correlated statistically 

with adverse clinical outcomes. 

CONCLUSION 

Identification of anatomical and imaging factors that correlate 

with adverse clinical outcome suggests that selected 

meningioma radiosurgery patients require frequent follow up 

for early detection and treatment of radiation-induced injury. 

KEY WORDS: Meningioma, gamma knife, radiation change 

Poster 51 

Errors in Localization Due to Chemical Shift Artifact in 

an Oil-Based Stereotactic Frame 

Parrish, T. B. · Kepka, A. · Levy, R. M. · Walker, M. T. 


Chicago, IL 

PURPOSE 

The purpose of this study was to investigate the error 

induced in stereotactic localization based on MR images due 

to chemical shift artifact. These errors originate from the use 

of oil as the marker fluid in the frame. The oil-based marker 

will shift in the frequency (readout) direction relative to 

water (brain tissue) and result in targeting errors.


In order to investigate the magnitude of the chemical shift, 

we devised a Plexiglas phantom with holes drilled in a uniform 

pattern. Every other hole was filled with oil and the 

remaining holes were filled with doped water. The phantom 

was imaged at 1.5 T and 3 T with a T1-weighted MP-RAGE 

sequence at different bandwidths. The spacing between the 

holes was measured in the image using the radiation planning 

software and an in-house Matlab program. To demonstrate 

the effect in a clinical environment, images of 2 

patients were acquired with the standard bandwidth of 130 

hz/pixel and one with a high bandwidth of 500 hz/pixel. The 

frequency orientation was anterior to posterior. The imaging 

data were compared using radiation planning software. 

RESULTS 

The results of the phantom data demonstrated the 220 hz 

chemical shift at 1.5 T. The standard bandwidth sequence at 

1.5 T resulted in a 1.5 pixel (1.5 mm) shift in the readout 

direction. At 3 T, the shift was twice as large because chemical 

shift scales directly with field strength. Fig 1 visually 

demonstrates the chemical shift artifact at 1.5 T. The image 

on the left was acquired with the 130 hz/pix bandwidth (lbw) 

and the one on the right used the 500 hz/pix bandwidth 

(hbw). The solid line is aligned with the markers in the hbw 

sequence and shows how the markers have shifted in the lbw 

image. The dotted line demonstrates that the anatomy has 

remained the same. The short lines with arrows are referenced 

to the marker and point to different anatomy. 

CONCLUSION 

Stereotactic frames were designed to improve localization of 

procedures. However, if the frame uses an oil-based marker 

in conjunction with MR imaging, there is a systematic error 

in the frequency encoding direction. This error could lead to 

suboptimal targeting of a lesion which is to be biopsied or 

radiated. The amount of error increases at higher field 

strengths and lower bandwidths. In clinical practice, it 

should be noted that with standard imaging techniques, oilbased 

stereotactic frames are prone to targeting errors in the 

chemical shift direction. 

KEY WORDS: Stereotactic, brain surgery, localization 

313 

Poster 52 

In Vivo Vital Sign Investigation on Subjects Undergoing 

8.0 T High-Field Head MR Imaging 

Yang, M. · Christoforidis, G. A. · Beversdorf, D. · 

Schmalbrock, P. 

The Ohio State University 

Columbus, OH 

PURPOSE 

To determine whether basic physiologic signs are altered significantly 

after exposure to the time-varying gradient 

strength, radiofrequency (RF) power, and 8 T static magnetic 

field. 


Subjects were scanned using an 8.0 T high-field MR system 

and a custom-made 16-strut transverse electromagnetic 

(TEM) coil tuned at 340 MHz. Two groups of subjects 

included 65 with cerebral pathology (58 with pathologically 

proven brain tumors and 7 with cerebral vascular disease 

cases; mean age = 45 years, range 20-70 years, male:female 

= 32:33) and 18 normal volunteers (mean age 30 years, 

range 20-56 years, male:female = 11:7). Vital signs measured 

included mean arterial pressure (MAP), respiration rate 

(RR), heart rate (HR), oxygen saturation, and temperature. 

Vital signs were obtained before, during, and after scanning. 

The MR imaging procedure included gradient-echo (GRE) 

and rapid acquisition relaxation (RARE) sequences with 

specific absorption rates (SAR) of 0.086 W/kg and 0.95 

W/kg respectively. Scanning duration varied from 20-30 

minutes in the disease group and 45 to 90 minutes in the normal 

group. Subjects were interviewed and reported any 

unusual experiences. A paired Student t-test was used to 

compare subject’s physiologic index before and after imaging. 

RESULTS 

In normal group, the HR significantly decreased from 65.2 to 

60.3 (p < 0.001), other vital sign has no significant change 

(Table). Thirteen subjects (72%) thought the procedure had 

no unusual experience. Transient phenomena, such as nausea, 

cold, dizziness, and vertigo were reported sporadically. 

In the disease group, RR increased significantly from 14.5 to 

15.1 per minute (p < 0.001). Other vital signs did not significantly 

change (Table). Forty (60%) subjects felt normal. 

Others reported transient vertigo, nausea, or coldness. 

Intergroup comparison found that only the HR changed significantly 

(p < 0.001). 

Vital sign change in normal group and disease group 

Normal Normal Normal Disease Disease Disease Intergroup 

group group group group group group comparison 

Prescan Postscan p value Prescan Postscan p value p value 

Respiration 14.3 14.8 0.217 14.5 15.1 0.001 0.8727 

(breaths/min) 

Temperature (F) 97.1 97.2 0.666 97.3 97.2 0.398 0.8273 

HR (beats/min) 65.2 60.3 0.001 70.6 70.3 0.682 0.003 

MAP (mmHg) 90.0 93.4 0.051 96.4 98.1 0.081 0.3721 

Oxygen 97.2 97.4 0.798 96.8 96.7 0.791 0.7369 

saturation (%) 

CONCLUSION 

This study provides direct evidence that undergoing 8 T 

high-field MR imaging of the brain is safe in normal subjects 

and selected patients. 

Posters

Posters 

REFERENCES 

1. Chakeras DW, et al. Effect of static magnetic field exposure of 

up to 8 Tesla on sequential human vital sign measurements. 

JMRI 2003;18:346-352 

2. Shellock FG, Crues JV. MR procedure: Biological effects, 

safety and patients care. Radiology 2004;232:635-652 

KEY WORDS: High-field MR imaging, MR imaging, safety 

Poster 53 

8 T Phase Imaging of the White Matter Tracts of the Mid 

and Upper Brainstem 

Chakeres, D. W. · Abduljalil, A. · Schmalbrock, P. 

Ohio State University 

Columbus, OH 

PURPOSE 

We compare the axial anatomy of the mid and upper brainstem 

white matter tracts as imaged with magnitude and 

phase gradient-echo (GE) high-resolution 8 T images with 

published anatomy diagrams and histologic sections. 


We studied 6 normal human subjects. All subjects signed 

informed consent approved by an Investigational Review 

Board. The studies were performed on an 80 cm bore 

Magnex [General Electric (Abingdon, U.K.)]. Eight T magnet 

system with a water-cooled asymmetric gradient controlled 

by a Bruker Avance (Billerica, MA) console and a 21 

cm long transverse electromagnetic head coil. Gradient-echo 

axial inter-laced (two separate series) offset 1 mm MR 

images were acquired with imaging techniques of TR 600, 

TE 12, slice thickness of 2 mm, scan gap 2 mm, field of view 

of 18 cm, flip-angle approximated 20 o , bandwidth 50 kHz, 

acquisition time approximately 8 minutes, and a matrix of 

1024 x 768. In addition to the magnitude images, phase 

images were calculated after subtracting the slow varying 

phase due to the static field. The images were displayed as 

either routine magnitude images or pure phase images (Fig. 

1). We correlated the 8 T images to published axial histologic 

and anatomical diagrams. 

RESULTS 

Only one exam failed for technical reasons. Overall the 

image quality of the brainstem region was good and consistent. 

The imaging findings between all of the subjects were 

consistent. The magnitude images demonstrated most of the 

iron-bearing nuclei as low signal regions. The phase images 

better differentiated the white and gray matter structures 

(Fig. 1). The white matter tracts demonstrated higher intensity 

on the phase images while the gray matter structures 

were lower in intensity. White matter tracts well seen included: 

medial longitudinal fasciculus, lateral lemniscus, 

spinothalamic tract, medial lemniscus, superior cerebellar 

peduncle, pontocerebellar fiber, corticonuclear tract, corticospinal 

tract, cerebral peduncle, internal capsule, and the 

central tegmental tract. Gray matter regions could be defined 

by white matter margins, including: pontine nuclei, red 

nuclei, substantia nigra, medial geniculate, and the lateral 

geniculate. 

314 

CONCLUSION 

Gradient-echo 8 T images demonstrate many of the white 

matter tracts of the mid and upper brainstem in greater detail 

than any other imaging technique. Identification of the white 

matter tracts help define a number of the important nuclei. 

KEY WORDS: Brainstem, white matter tracts, high-field MR 

imaging 

Poster 54 

Why Is Volumetry Not Enough? Assessment of Shape 

Characteristics of Ventricular Systems from Time-Series 

Examinations in MR Imaging 

Preul, C. · Huebsch, T. · Tittgemeyer, M. · von Cramon, D. 

Max-Planck-Institute for Human Cognitive and Brain 

Sciences 

Leipzig, GERMANY 

PURPOSE 

From neuroradiologic experience it is evident that the adaptation 

of the ventricular system secondary to pathologic 

processes or surgery is not uniform. To describe changes 

entirely, information about volume and shape have to be 

considered in particular. The information encoded in the 

change of shape is addressed in this study. To exemplify the 

technique, we used time-series MR examinations of patients 

with surgically treated chronic or acute occlusive hydrocephalus. 

The analysis allows to mimic the pattern of ventricular 

adaptation after surgery. We argue that the alterations 

of the ventricular configuration depend on the dynamics of 

ventricular enlargement. 


Preoperative and postoperative MR imaging at different 

time-steps was performed in five patients with occlusive 

hydrocephalus. The etiology of aqueductal stenosis and 

time-course of ventricular enlargement was different in all 

patients. Third and lateral ventricles were segmented with an 

automated classification scheme. Ventricular surfaces were 

binarized, mapped to a spherical coordinate system and 

modelled by harmonic basis functions (Fig.). This approach 

allows to simplify the complex shape by stepwise filtering 

the details that form the surface. The ventricles can be compared 

directly on the level of the simplified shape. 

RESULTS 

Although the relative volumetric change was comparable 

(approximately 50% from pre to postoperative status for the 

case presented in the figure) within patients, analysis of 

shape reveals notable regional differences in the pattern of 

adaptation (e.g., the differences between pre and postopera-

tive condition is 20% at the posterior horn), whereas changes 

in the lateral wall of the ventricular body are discrete. 

Comparing the different cases of hydrocephalus, a fundamental 

difference between acute and chronic hydrocephalus 

is reflected by the analysis. This finding suggests a vital difference 

in the development of preoperative ventricular 

enlargement and postoperative readjustment between acute 

and chronic hydrocephalus. 

CONCLUSION 

The process of readjustment is not uniform: distinct differences 

between pre and postoperative shape appear in a comparison 

of the underlying simplified models. Additional to 

the mere volumetric description, this approach allows to 

identify regions that readjust differently to the altered pressure. 

Analyzing the different patterns of ventricular adaptation 

in patients with chronic or acute hydrocephalus suggests 

a theoretical concept that has to consider tissue properties of 

the surrounding parenchyma.The technique is not restricted 

to the assessment of ventricular adaptation after surgery, but 

can be used to monitor the ventricular system in time-series 

examinations in general. 

KEY WORDS: Ventricular system, analysis of shape 

Poster 55 

Comparison of High Resolution to Standard Resolution 

CT of the Head 

Riccelli, L. P. · Anderson, J. C. · Nesbit, G. M. · Weissman, 

J. L. · Hamilton, B. E. · Brown, P. 


Portland, OR 

PURPOSE 

Newly developed reconstruction filters allow high resolution 

CT scanning of the head on the Philips Brilliance 16 detector 

CT. This study presents initial evaluation of high compared 

to standard resolution head CT scans. The purpose of 

the study is to determine whether high resolution scans provide 

expected improvement in quality, using the ACR phantom 

and patient scans. 

315 


High resolution and standard resolution CT scans of high 

and low contrast portions of the ACR CT phantom were 

compared. Standard resolution axial noncontrast technique, 

with the standard resolution UB reconstruction filter was 

compared to high resolution axial technique with the new 

UA and UB high resolution filters. Technique parameters 

were similar, except for rotation time, which was reduced 

from 0.75 to 1.5 seconds at high resolution to achieve the 

same mAs. Neuroradiologists rated the number of line pairs 

confidently seen to evaluate high contrast resolution. Low 

contrast elements of the phantom also were evaluated, and 

rated as to the smallest size seen with confidence. A series of 

20 high resolution head CT studies done for routine clinical 

follow up was compared to prior standard resolution studies 

on the same patients. The pons, cerebellum, internal capsule/insular 

cortex, gray-white junction, overall diagnostic 

quality, and overall aesthetic quality were evaluated by four 

neuroradiologists as similar, better, or worse than standard 

scans. Bone algorithms were compared similarly on seven 

patients. Patients with known or suspected pathology are 

continuing to be evaluated at high and standard resolution. 

RESULTS 

Initial review of high resolution phantom scans indicates that 

on average 1.6 lp/cm more were resolved with high compared 

to standard resolution scanning (10.6 vs 9 lp/cm). With 

high resolution technique, conspicuity of low contrast phantom 

elements was similar. Bone algorithms at high resolution 

were considered sharper than standard resolution in all 

seven patient cases evaluated. High resolution UA images 

were subjectively considered better than high resolution UB 

images, primarily due to noisiness of UB images. Four neuroradiologists 

reviewed 20 sets of studies, comparing high 

resolution UA to standard resolution UB images. The image 

quality of the following structures at high resolution was 

rated as better/same/worse than standard resolution respectively: 

pons (36/29/15), cerebellum (50/17/13), internal capsule/insular 

cortex (55/14/11), gray-white junction 

(53/12/15), overall diagnostic quality (45/27/8), and overall 

aesthetic quality (55/17/8). Limitations of high resolution 

scanning include greater susceptibility to motion artifact due 

to increased scan time, with increased artifact along the inner 

table of the skull in some cases. 

CONCLUSION 

High resolution technique CT scans of the ACR CT phantom 

show improved high contrast resolution, and similar low 

contrast resolution, compared to standard technique. A series 

of 20 high resolution head CT studies compared to prior 

standard studies showed increased conspicuity of gray-white 

structures. Patients with known pathology or suspected subtle 

lesions are continuing to be evaluated to determine if 

diagnostic accuracy can be improved with high resolution 

technique. 

KEY WORDS: Brain; head, high-resolution acquisition, CT 

Posters

Posters 

316 

Poster 56 

Poster 57 

Brain Metabolic Changes following Cerebral Regional Cerebral Blood Flow Single Photon Emission 

Concussion: A Pilot Study Using Proton MR Tomography and Neuropsychological Tests in the 

Spectroscopy 

Evaluation of Postconcussion Syndrome after Mild Head 

Trauma: Functional Evaluation Six Months after Event 

Ludovici, A. · Garaci, F. G. · Marziali, S. · Colangelo, V. · 

Gaudiello, F. · Floris, R. · Simonetti, G. 


Rome, ITALY 

PURPOSE 

Current imaging does not allow quantification of neural 

injury after traumatic brain concussion and therefore limits 

appropriate patient management. In this study we assessed 

over time the neurochemical changes that could follow a 

cerebral concussion by using proton MR spectroscopy 

(MRS). 


Ten patients who had a concussion were studied by conventional 

MR imaging and MRS within 3 days, at 10 and at 45 

days after the trauma. Ten age-matched subjects served as 

controls. A multivoxel with point-resolved spectroscopy 

sequence (PRESS) was used. N-acetyl-aspartate/creatine 

(NAA/Cr), N-acetyl-aspartate/choline (NAA/Cho) and 

choline/creatine (Cho/Cr) ratios were calculated bilaterally 

in subcortical white matter of frontal and parietal lobe. 

RESULTS 

N-acetyl-aspartate/creatine was statistically significantly 

lower compared with controls at the first time-point. At 10 

days NAA/Cr showed a trend towards normalization. MR 

spectroscopy performed 45 days after the concussion 

showed a complete recovery of NAA/Cr values. 

Choline/creatine ratio was normal. N-acetylaspartate/choline 

was statistically significantly lower compared 

with controls at the first time-point. At 10 days 

NAA/Cho showed a complete recovery of normal values. 

CONCLUSION 

The NAA system may represent a unique metabolic construct 

of the intact human brain. MR spectroscopy can be a 

useful noninvasive method for in vivo monitoring of posttraumatic 

metabolism in brain concussion. Our results 

demonstrate a change in NAA/Cr ratios. Recovering of the 

above value suggests a temporary metabolic impairment. 

These findings may reflect a metabolic vulnerability of the 

brain following a concussion. This transitory status could 

worsen the effect of a second traumatic event. 

KEY WORDS: Cerebral concussion 

Ramos-Font, C. 1 · Rodriguez-Fernandez, A. 1 · Vilar-Lopez, 

R. 1 · Gomez-Rio, M. 1 · Garcia-Rivas, J. V. 1 · Castaneda- 

Guerrero, M. 1 · Perez-Garcia, M. 2 · Llamas-Elvira, J. M. 1 · 

Bellon-Guardia, M. 1 

1Virgen de las Nieves Universitary Hospital, Granada, 

SPAIN, 2Psychology University, Granada, SPAIN 

PURPOSE 

Postconcussion syndrome (PCS) is a clinical entity that some 

patients could develop after mild head trauma (HT). It usually 

is diagnosed clinically, on basis of patient symptoms, 

reported without abnormalities in neurostructural imaging 

studies (CT, MR imaging). In this situation, neurofunctional 

procedures: regional cerebral blood flow (rCBF) studied by 

single photon emission tomography (PET) and neuropsychological 

testing, could have high relevance, especially 

when it could exist malingering. The objective of this work 

is to study the possible correlation of the results obtained 

from clinic, rCBF single PET and neuropsychological testing. 


For this, [considering that this work is still in progress (was 

conceived as prospective series)] 27 patients have been 

included preliminarily (male: 63%; mean age: 32.3 ± 11.02 

years). All patients were clinically symptomatic of PCS 6 

months after a mild HT and none of them showed radiologic 

changes. Using the clinical assessment, rCBF single PET 

and neuropsychological testing were evaluated, considered 

in interesting blinded conditions. 

RESULTS 

We found different degrees of global heterogeneity pattern 

of rCBF among the observed patients that reached 70.4% of 

them. Specific regions of reduced uptake were present at 

prefrontal areas (usually bilateral: 66.6%) and transitional 

occipital-parietal cortex (right: 70.4%; left: 55.5%; both: 

40.7%). Only 4 patients were considered as fully normal. On 

basic clinical suspicions, three patients showed a high probability 

of malingering, rCBF single PET and neuropsychological 

testing were normal in two cases and with diffuse 

global frontal hypometabolism in one. 

CONCLUSION 

Instead of few patients for conclusions and in absence of a 

definitive diagnostic procedure (that could considered as 

“gold standard”), our preliminary results suggest that the 

combination of functional approach (rCBF and neuropsychological 

testing) could be use to assess the diagnosis of 

PCS. 

KEY WORDS: Postconcussion syndrome, regional cerebral 

blood flow, single photon emission tomography

Poster 58 

Quantitative Analysis of Coma Duration Prediction in 

Diffuse Axonal Injury Using MR Imaging 

Zheng, W. B. 1 · Liu, G. R. 1 · Li, L. P. 2 · Wu, R. H. 1 

1Second Hospital, Shantou University, Medical College, 

Shantou, CHINA, 2 317 

Poster 59 

Measurement of Blood Flow in Acute and Subacute 

Brain Contusion Using 3.0 T Pulsed Arterial Spin 

Labeling and Contrast-Enhanced Perfusion MR Imaging 

Ai, L. · Dai, J. · Li, J. 

Shantou University, Shantou, CHINA Beijing Tiantan Hospital 

Beijing, CHINA 

PURPOSE 

To evaluate the hypothesis that the apparent diffusion coefficient 

(ADC) values combined with clinical prognostic factors 

indicates the depth of shearing lesions in the brain structure 

and therefore relates to coma duration of diffuse axonal 

injury (DAI). 


Sixty adult patients in posttraumatic coma state had cerebral 

MR imaging between 2 hours and 20 days after injury. All 

scans were obtained on a GE 1.5 T MR scanner. Convention 

MR imaging consisted of spin-echo T1-weighted imaging 

and T2-weighted imaging. Diffusion-weighted imaging was 

performed using spin-echo echo-plannar sequence. Apparent 

diffusion coefficient maps were obtained and the mean ADC 

values of each region of interest (ROI ) were measured using 

MR imaging console software. The analysis methods were 

used backward stepwise logistic regression for the coma 

duration prognostication model established. The correlations 

between ADC values, Glasgow coma scale, age, the numbers 

of presence of apparent brain injury, pupillary abnormalities, 

and the duration of coma were investigated. 

RESULTS 

A model to prognosticate the duration of coma was established. 

Logistic regression analysis showed that clinical 

characteristics, such as initial score on the Glasgow coma 

scale, age, and pupillary abnormalities, the number of all 

lesions, ADC scores of the lesions were predictive of the 

duration of coma. 

CONCLUSION 

Posttraumatic coma duration of DAI could be predicted by 

cerebral MR imaging findings in the acute to subacute stage 

after head injury combined with clinical prognostic factors. 

Age, ADC, GCS, number of lesions, pupillary abnormalities 

are highly significant in predicting coma duration. The technique 

presented herein might provide a tool for in vivo 

detection of DAI for the coma duration at the early stages in 

patients with traumatic brain injury not only for the diagnosis 

but also for the treatment of these patients. 

KEY WORDS: Diffuse axonal injury, MR imaging, Coma 

duration prognostication model 

PURPOSE 

To evaluate relative cerebral blood flow (rCBF) in traumatic 

brain contusion using arterial spin labeling (ASL) and contrast-enhanced 

MR imaging methods. 


Thirty-three adult patients with 1 hour to 10 days traumatic 

brain contusion were examined on a 3.0 T MR medical system. 

Arterial spin labeling was performed in 33 patients and 

contrast-enhanced perfusion MR imaging was done in 27 

patients. Blood flow was measured in normal-appearing 

brain matter and peri-contusion regions with both methods. 

RESULTS 

Blood flow values in normal-appearing brain matter or pericontusion 

regions obtained by both methods correlated well 

(P < 0.001) and were different between normal-appearing 

brain matter and peri-contusion regions in both acute and 

subacute stages (P < 0.05). 

CONCLUSION 

Aterial spin labeling technique is a promising and useful tool 

in evaluating cerebral blood flow of acute and subacute traumatic 

brain contusion due to the benefit of safety and efficiency. 

KEY WORDS: Brain contusion, arterial spin labeling 

Poster 60 

Early Brain MR Imaging after Carotid Endoarterectomy 

in Asymptomatic Patients 

Sintini, M. · Gessaroli, M. · Tarantini, S. · Manzo, L. 

Infermi Hospital Rimini 

Rimini, ITALY 

PURPOSE 

To evaluate the frequency of appearance of new cerebral 

ischemic lesions with diffusion-weighted and conventional 

MR imaging after carotid endo-arterectomy in asymptomatic 

patients. 


From January 2003 to February 2004 133 consecutive 

patients aged 53-84 years (mean 71) with a stenosis > 70% 

of one or both carotid arteries had presurgical (the day 

before) and postsurgical (the day after) brain MR imaging. 

All examinations were performed with a 1.5 T machine 

(Philips Intera Master) employing the following technique: 

sagittal T2-weighted FSE (4500/100/3 4 mm thick), axial 

FLAIR (11000/140/2500/2 5 mm thick), axial diffusionweighted 

single shot (6500/93/4, b = 0 - 1000, 3 mm. thick). 

If lesions were present at postsurgical diffusion-weighted 

Posters

Posters 

imaging, we calculated their apparent diffusion coefficient 

(ADC) employing a manual ROI, and scheduled the patients 

for a final MR imaging about 3 weeks later. Surgery consisted 

in unilateral carotid endoarterectomy with semieversion, 

dacron-patch and shunt with the patient under general anesthesia, 

and with a first clamping time of 6-11 min, a shunting 

time of 12-16 min and a second clamping time 70%, 23 

> 90%, 49 bilateral (20 >70%, 9 > 90%, 9 > 70% on one side 

with controlateral occlusion, 6 > 90% with controlateral > 

70%, and 5 > 90% with controlateral occlusion). All atheromasic 

plaques were fibrocalcific (46 pure, 52 with subintimal 

hemorrhage, 23 ulcerated. 

RESULTS 

All MR examinations were of good quality and early postsurgical 

of 8 patients (6.6%) pointed out as a whole in diffusion-weighted 

imaging 20 little (maximum diameter < 1 

cm), sometime multiple (from 1 to 7) lesions with reduced 

ADC. In all cases the lesions that we considered as acute 

ischemic edema, affected the cerebral hemisphere of the 

same side of endoarterectomy and their location included 

cortical surface (16), sub-cortical white matter (2), and basal 

ganglia (2). The final MR imaging showed the development 

of 2 ischemic infarcts in one patient of the same size depicted 

in the early postsurgical study, whereas all other lesions 

disappeared. We did not find correlations with a) first clamping 

time, b) degree of carotid stenosis, c) features of atheromasic 

plaques. 

CONCLUSION 

The rate of occurrence of acute ischemic lesions in asymptomatic 

patients after carotid endoarterectomy performed with 

the aforementioned technique is low (6.6%), and MR imaging 

proved effective to disclose them; a final infarct is rare 

(0.8%). We believe that embolic material detached during 

surgery, but also air bubbles, may have caused temporary 

occlusion of little cerebral vessels; since no lesion in our 

patients affected eloquent regions of the brain, we don’t 

know if the absence of neurologic symptoms was just due to 

their location. In conclusion we believe that asymptomatic 

patients after carotid endoarterectomy do not need MR imaging 

before discharging. 

KEY WORDS: MR imaging, carotid disease 

Poster 61 

Normal Variants and Anomalous Branches of the 

Intracranial Circulation Demonstrated by MR 

Angiography at 3 T 

Parmar, H. A. 1,2 · Sitoh, Y. 2 · Hui, F. 2 

1 Hospital for Sick Children, Toronto, ON, CANADA, 

2 National Neuroscience Institute, Singapore, SINGAPORE 

PURPOSE 

The introduction of high-field 3 T MR scanners into clinical 

practice has added a new dimension in MR angiographic 

evaluation of the intracranial circulation. The synergistic 

combination of higher signal-to-noise ratio and improved 

background suppression at 3 T enables high-resolution imaging 

of the intracranial vessels, giving superior image quality 

with the identification of smaller and more distal vessels, 

compared to 1.5 T. This may lead to an increased detection 

318 

of both normal variants as well as vessel abnormalities. We 

present a review of the normal variations and anomalies of 

the intracranial circulation, detected on MR angiography 

(MRA) performed on a high-field 3 T clinical scanner using 

parallel imaging techniques. The salient imaging features of 

these anomalies and normal variations are discussed with 

relevance to clinical practice. 


We performed 843 MRAs on a 3 T clinical scanner 

(Gyroscan Intera, Philips Medical System, Einthoven, 

Netherlands) over last 7 months. All patients had strokerelated 

symptoms and were referred for evaluation of the 

intracranial circulation. Unenhanced three-dimensional (3D) 

time of flight MRA was performed using an 8-element head 

coil array and sensitivity encoding (SENSE) technique. 

Maximum intensity projection (MIP) images were generated 

in all cases. Maximum intensity projection and source 

images were reviewed independently by two radiologists. 

Other reconstruction methods like 3D surface shaded display 

(SSD) were used in some problematic cases to resolve the 

complex anatomy. 

RESULTS 

We observed 13 arterial fenestration involving basilar (8), 

vertebral (1), middle cerebral (1), posterior cerebral (1), and 

anterior cerebral (2) artery. There was no associated 

aneurysm seen in any patient. There were 5 patients with 

persistent carotico-basilar anastamosis: persistent trigeminal 

(3), persistent hypoglossal (1), and proatlantal intersegmental 

artery (1). Anomalies of the middle cerebral artery included 

duplicated (3), accessory (4), and early branching middle 

cerebral arteries (10). Anterior cerebral artery variants were 

observed in 4 patients including azygous (1), multichanneled 

(2), and bihemispheric anterior cerebral artery (1). Posterior 

cerebral artery variant including hyperplastic anterior 

choroidal artery (2), fetal origin of posterior cerebral artery 

(68) also were seen. 

CONCLUSION 

In conclusion, the intracranial circulation especially at the 

circle of Willis can show varied patterns of branching and 

course of vessels. It is important to recognize these normal 

variants and differentiate them from pathologic vessel diseases 

in order to guide therapy and vascular surgery. MR 

angiography performed on a high-field 3 T scanner, due its 

noninvasive nature and exquisite depiction of the vessels, 

effectively demonstrates these normal variants and anomalous 

branches. 

KEY WORDS: Intracranial circulation, MR angiography, normal 

variants

Poster 62 

Increased Signal in the Subarachnoid Space on Fluid- 

Attenuated Inversion Recovery Imaging Due to Delayed 

Clearance of the Gadolinium Chelate: A Potential 

Diagnostic Pitfall 

Morris, J. M. · Miller, G. M. 

Mayo Clinic 

Rochester, MN 

PURPOSE 

Hyperintense cerebral spinal fluid (CSF) in the subarachnoid 

space (SAS) on fluid-attenuated inversion recovery (FLAIR) 

imaging has been reported in numerous pathologic conditions 

including subarachnoid hemorrhage, meningitis, 

meningeal carcinomatosis, superior sagittal thrombosis, and 

stroke. It also has been reported in otherwise normal patients 

undergoing anesthesia with supplemental oxygen. We present 

a series of 9 patients with hyperintense CSF signal in the 

SAS on FLAIR imaging due to delayed clearance of the 

gadolinium chelate. 


The average age of the patients was 68 years (range 21-81 

years). The head MR examinations were performed for a 

variety of reasons including seizures, spells, TIA, and dizziness. 

All nine patients had a gadolinium-enhanced body, 

spine, or brain MR image prior to their head exam. The average 

time between the two studies was 2 days (range 1-6 

days). 

RESULTS 

Seven of the nine patients had an emergent lumbar puncture 

performed to rule out subarachnoid hemorrhage (SAH), 

infection, or carcinomatosis. Three patients had mildly elevated 

protein in their CSF. The CSF bacterial cultures, viral 

serologies, and remaining laboratory values were normal. 

Three patients had chronic renal insufficiency, two patients 

had acute renal failure, and the remaining 4 patients had normal 

renal function. Two patients had acute infarcts, one in 

the occipital lobe and one in the right parietal lobe, remote 

from the site of CSF hyperintensity. Seven patients had no 

other significant intracranial abnormalities on MR imaging 

to account for the CSF hyperintensity. One patient had follow-up 

head studies 1 day and 2 days after the initial exam 

documenting resolution of the hyperintense CSF abnormality 

(Fig. 1). Two previous case reports have documented CSF 

hyperintensity in the SAS on delayed postcontrast FLAIR 

images in patients with renal failure secondary to decreased 

gadolinium chelate clearance likely accounting for the 

hyperintensity in 5 or our patients. To our knowledge no 

reports to date have documented this finding in patients with 

normal renal function and without any other intracranial 

abnormalities. 

319 

CONCLUSION 

Given the sharp rise in volume of contrast-enhanced MR 

studies it is inevitable that some patients will have contrastenhanced 

MR imaging 24-48 hours before brain MR imaging. 

The neuroradiologist should be aware that delayed 

gadolinium chelate clearance can cause increased signal in 

the SAS on FLAIR imaging in elderly patients without 

abnormalities known to disrupt the blood-brain barrier, with 

or without a history of renal failure. 

KEY WORDS: FLAIR, subarachnoid space, gadolinium 

Poster 63 

Sensitivity and Specificity of CT Angiography in the 

Diagnosis of Intracranial Aneurysms: Comparison with 

Intraarterial Digital Subtraction Angiography 

Hekmatnia, A. · Basiratnia, R. · Nouri-Mahdavi, K. · Rezaei, 

M. · Saboori, M. · Sanei, H. · Ghadamgahi, M. · Hamedi, M. 

Isfahan University of Medical Sciences 

Isfahan, IRAN (ISLAMIC REPUBLIC OF) 

PURPOSE 

The prevalence of intracranial aneurysms varies from 1% to 

14% in the general population. Aneurysm rupture is associated 

with high mortality and morbidity rates and occurs in 1- 

2% of the patients per year. Fortunately, with a timely diagnosis, 

surgical or interventional repair of unruptured 

aneurysms can readily decrease their mortality and morbidity 

rates significantly. Objectives: The purpose of this study 

was to compare the sensitivity and specificity of CT angiography 

as a noninvasive, outpatient-based, and relatively 

cheap technique with those of intraarterial digital subtraction 

angiography (DSA) in the detection of intracranial 

aneurysms. 

Posters

Posters 


From August 2003 through August 2004, 28 patients (15 

females and 13 males) with subarachnoid hemorrhage with 

an age range of 13-71 years underwent CT angiography and 

then intraarterial DSA. CT angiography was performed 20 

seconds after injector-administration of 150 ml of 

Omnipaque (300 mg/ml) at a rate of 5 ml/sec into the antecubital 

vein. Patients with a finding of intracranial aneurysm 

on intraarterial DSA and/or CT angiography underwent therapeutic 

intervention. 

RESULTS 

CT angiography detected 12 patients with aneurysms ranging 

in size from 4 to 18 mm, whereas intraarterial DSA 

revealed two more patients with aneurysms (range 4-18 

mm). Sensitivity of CT angiography in detecting intracranial 

aneurysms was 86% with a specificity of 88%. 

CONCLUSION 

As compared to intraarterial DSA, CT angiography is a reasonable 

noninvasive alternative diagnostic technique in 

detecting intracranial aneurysms. 

KEY WORDS: CT angiography, digital subtraction angiography, 

intracranial aneurysm 

Poster 64 

Can CT Angiography Replace DS Angiography for 

Investigation of Patients with Subarachnoid 

Hemorrhage? A Double-Blind Study of 120 Patients 

Chakraborty, S. · Roughley, S. · Das, K. · Niven, S. · Nahser, 

H.C. · Nixon, T.E. 

Walton Centre for Neurology and Neurosurgery 

Liverpool, UNITED KINGDOM 

PURPOSE 

The aim of this study was to determine whether CT angiography 

(CTA) could replace digital subtraction angiography 

(DSA) as the definitive diagnostic tool for the assessment of 

cerebral aneurysm following subarachnoid hemorrhage 

(SAH). 


One hundred and twenty patients who underwent CTA and 

DSA for investigation of acute SAH were reviewed retrospectively. 

A pilot of 10 CTAs was reported by two experienced 

neuroradiologists which showed excellent (Kappa 1) 

interobserver agreement. Subsequently, all cases were 

reviewed by one of the two neuroradiologists, who were 

blinded to the name of the patients and the results of DSA. 

These reports then were compared with the result of DSA for 

number, position, and size of the abnormality. 

RESULTS 

Compared to DSA, CTA had sensitivity of 92% and specificity 

97%. The positive predictive value was 99% and negative 

predictive value was 83%. We illustrate the cases 

missed by CTA and also discuss the caveats of CTA examinations. 

320 

CONCLUSION 

The results of this study demonstrate that although the sensitivity 

and specificity of CTA are comparable with DSA in the 

detection of aneurysms, DSA may be necessary in patients 

with SAH when the CTA is negative. 

KEY WORDS: CT angiography, aneurysm, digital subtraction 

angiography 

Poster 65 

Application of 3D Time-of-Flight MR Angiography of 

Intracranial Arteries at 3.0 T 

Wei, C. 1 · Yuan, F. 2 · Long, M. 2 · Gu, X. 3 

1The Affiliated Hospital of Jiangsu University, Zhenjiang, 

CHINA, 2The First Center Hospital of Tianjin Medical 

University, Tianjin, CHINA, 3The Affiliated Hospital of 

Wujing University, Tianjin, CHINA 

PURPOSE 

The purpose of this study was to evaluate the advantage and 

the diagnostic potential of 3D TOF MRA at 3.0 T with 

SENSE technique by examination of patients at 1.5 T and 3.0 

T MR imaging (with SENSE and without SENSE). 


Fifteen healthy volunteers were examined at 3.0 T and 1.5T 

for circle of Willis using 3D TOF MRA. Twenty patients 

who were referred to MRA were performed of the intracranial 

arteries using 3D TOF MRA with SENSE (SENSE factor: 

2) and without SENSE at 3.0 T. All resulting source 

images were examined to determine blood-to-background 

contrast, blood signal-to-noise (SNR) and contrast-to-noise 

(CNR) ratios. A qualitative blinded examination of all brain 

source images and MIPs was made by a neuroradiologist to 

determine proximal and distal vessel visibility, degree of 

background suppression, and sharpness. 

RESULTS 

The average blood-to-background contrast of 15 volunteers 

was 2.6 + 0.5 at 1.5 T and 4.0 + 0.2 at 3.0 T; SNR were 55.2 

+ 5.8 at 1.5 T and 145.4 + 14.3 at 3.0 T ; CNR were 39.7 + 

6.4 at 1.5 T and 107.6 + 10.9 at 3.0 T. The average blood-tobackground 

contrast of 20 patients with SENSE and without 

SENSE were 4.5 + 0.2 and 3.3 + 0.5; SNR were 122.9 + 51.2 

and 96.4 + 11.8; CNR were 120.2 + 22.5 and 72.4 + 8.5, 

respectively. From the qualitative evaluation by a neuroradiologist, 

the only significant findings were an improvement at 

3.0 T in the visualization of the distal intracranial vessels 

(2.8 + 0.4)and in background suppression (0.8 + 0.4 at 3.0 T 

and 0.6 + 0.5 at 1.5 T).

CONCLUSION 

For 3D TOF MRA, the longer longitudinal T1 relaxation 

times at 3.0 T over 1.5 T lead to substantial increase in 

blood-to-background contrast. This increase in contrast is 

valuable for imaging distal intracranial vessels flowing 

through gray matter. SENSE used multiple MR imaging 

receive coil elements to encode spatial information in addition 

to traditional gradient encoding. SENSE MRA reduced 

scan time without losing the spatial resolution. 3D TOF 

MRA with SENSE technique is a rapid, convincing choice 

for neurovascular 3D TOF MRA. 

KEY WORDS: MR angiography, 3D time-of-flight, 3.0 T 

321 

Poster 66 

Sensitivity and Specificity of MR Angiography in the 

Diagnosis of Intracranial Aneurysms: Comparison with 

Intraarterial Digital Subtraction Angiography 

Hekmatnia, A. · Basiratnia, R. · Nouri-Mahdavi, K. · 

Hamedi, M. · Saboori, M. · Ghadamgahi, M. · Rezaei, M. 



PURPOSE 

The prevalence of intracranial aneurysms varies from 1% to 

14% in general population with a rupture risk of 1-2 % per 

year and is associated with high mortality and morbidity 

rates. Fortunately, with a timely diagnosis, surgical or interventional 

repair of unruptured aneurysms can readily 

decrease their mortality and morbidity rates significantly. 

The purpose of this study was to determine the sensitivity 

and specificity of 3D time of flight and phased-contrast MR 

angiography (MRA) as a noninvasive, outpatient-based and 

relatively cheap method compared to intraarterial digital 

subtraction angiography (DSA) in detection of intracranial 

aneurysms. 


From October 2002 until December 2003, 54 patients (36 

females and 18 males) with subarachnoid hemorrhage with 

an age range of 18 to 77 years underwent MRA and then 

intraarterial DSA respectively. MR angiography was performed 

using a head coil, with a field of view of 260, and a 

matrix scan of 512. One hundred 0.6 mm thick slices were 

obtained in each patient. Patients with a finding of intracranial 

aneurysm on intraarterial DSA and/or MRA underwent 

therapeutic intervention. 

RESULTS 

MR angiography detected 20 aneurysms ranging in size from 

3 to 15 mm, whereas intraarterial DSA revealed 22 (range 3- 

15 mm). Sensitivity of MRA in detecting intracranial 

aneurysms was 90.9% with a specificity of 88.8%. 

CONCLUSION 

As compared to intraarterial DSA, MRA is a reasonable noninvasive 

alternative diagnostic technique in detecting 

intracranial aneurysms. 

KEY WORDS: Aneurysm-intracranial, MR angiography, digital 

subtraction angiography 

Posters

Posters 

Poster 67 

Poster 68 

Improved Characterization of Cerebral Arteriovenous Morphologic Comparison of Human and Canine 

Malformation before Embolization by MR Angiography Intracranial Arteries 

at 3.0 T 

Seifert, P. 1 · Arends, J. 2 · Henry, B. 2 · Williams, S. 3 · Higgins, 

R. J. 4 · Holloway, K. 2 

322 

Heidenreich, J. O. 1 · Schilling, A. M. 1 · Lüdemann, L. 2 · Wolf, 

K. J. 1 · Bruhn, H. 2 

1Charité Campus Benjamin Franklin, Berlin, GERMANY, 

2Charité Campus Virchow Klinikum, Berlin, GERMANY 

PURPOSE 

The role of MR imaging and MRA in the diagnosis of cerebral 

AVM is well established. To define the risk of a cerebral 

hemorrhage prior to embolization knowledge of feeding 

arteries and draining veins is crucial. MR angiography at 1.5 

T is considered to be not sufficient in predicting the malformations 

vasculature. Towards this goal we evaluated MRA at 

3.0 T and compared its performance with MRA at 1.5 T and 

digital subtraction angiography (DSA) as gold standard. 


Fifteen patients who had angiographically proven cerebral 

AVMs underwent 3D TOF MRA at 1.5 T (Magnetom Vision, 

Siemens) (TR/TE = 37/6.5 msec; FA = 20°; FoV 200; 512 x 

256 matrix, 70 mm slab, 1 mm effective) and at 3.0 T (Signa 

3T, GE Healthcare) (TR/TE = 30/ 4.4 msec; FA = 20°; FoV 

220; 512 x 256 matrix, 90 mm slab, 1 mm effective) prior to 

embolization therapy. The exams were evaluated by two 

neuroradiologists independently regarding image quality on 

a 4-point scale and detectability of feeding arteries and 

draining veins. 

RESULTS 

At comparable acquisition times 3D TOF MRA at 3.0 T 

showed a notably superior spatial resolution and brain coverage. 

Arteriovenous malformation detection was possible 

with both methods in 100%. At 3.0 T image quality of MRA 

was excellent in 14 cases and reasonable in one case due to 

misplacement of the saturation pulse. At 1.5 T quality was 

poor in 3 cases, reasonable in 7 cases and good in 5 cases. 

Number of feeding artery depiction was improved markedly 

for each case at 3.0 T, especially in the posterior fossa and 

temporal lobe but less spectacular in thalamic and central 

locations of the AVM. Although the precision of DSA in 

characterizing the feeding arteries particularly in superselective 

catheter positions was not yet reached, the information 

gained was judged sufficient for embolization planning and 

risk evaluation. Detection of venous drainage was insufficient 

in patients with superficial drainage due to saturation 

pulses. Also at both scanners a singular large deep draining 

vein was missed in two patients due to slow flow. 

CONCLUSION 

The gain in signal-to-noise at 3.0 T improves the characterization 

of vasculature in brain AVM as compared to MRA at 

1.5 T. More feeding arteries could be detected in every single 

patient, which increases the safety and predictability of 

each embolization procedure. It still does not give the same 

detailed information on vasculature as conventional angiography. 

Characterization of venous drainage is critical for 

superficial drainage and even larger veins, when the flow is 

reduced. 

KEY WORDS: AVM, 3.0 T, neuroradiology 

1 Boston Scientific, Oregon House, CA, 2 Boston Scientific, 

Fremont, CA, 3 Medical Education and Research Institute, 

Memphis, TN, 4 School of Veterinary Medicine, University of 

California Davis, Davis, CA 

PURPOSE 

Canine models are a valuable tool in the development of 

treatments for intracranial vascular diseases, such as 

aneurysm embolization and atherosclerotic lesion stenting. 

However, as there has been little published morphologic 

information on human intracranial arteries, our understanding 

of the comparability of human and canine intracranial 

arteries is limited. In order to aid in preclincal model development 

for intracranial stenting, this study compared the 

morphology of human and canine intracranial arteries. 


Canine (n = 3) and human cadaver (n = 3) intracranial arteries 

were compared histologically. Human arteries examined 

included the common carotid, internal carotid, the anterior 

and middle cerebral arteries (MCA), and the basilar artery. 

Canine arteries examined included the vertebral, basilar, 

communicating, internal carotid, and rostral cerebral arteries. 

Both human and canine arteries were stained with elastin 

trichrome and hematoxylin and eosin stains. 

RESULTS 

In human arteries, it was observed that large extracranial 

arteries are elastic arteries and exhibit a media replete with 

elastin fibers. As they continue distally, they transition to 

muscular arteries characterized by a media largely devoid of 

elastin fibers but possessing an internal elastic lamina (IEL) 

and an external elastic lamina (EEL). As the vasculature 

enters the skull proper, the muscular arteries lose their EEL. 

Elastin fibers in the media and adventitia of human intracranial 

arteries are sparse. The figure below shows the human 

MCA as an example of an intracranial target artery (A). 

There is a prominent IEL, but no EEL or elastin in the media 

or adventitia. Canine intracranial arteries are histologically 

similar to human intracranial arteries. There was no EEL in 

any of the canine arteries examined. There were no elastin 

fibers observed in the media or adventitia of canine specimens. 

In summary, canine intracranial arteries are similar in 

structure to human intracranial arteries and are appropriate 

models for intracranial stenting. In particular, the canine 

basilar artery (B) has a similar histologic structure to human 

intracranial arteries and is accessible and large enough for 

stenting.

CONCLUSION 

Intracranial arteries were morphologically similar between 

human and canine. Both species included an IEL but no 

EEL, and little or no elastin fibers in the media or adventitia. 

This study identified the canine basilar artery as an appropriate 

preclinical model for the evaluation of intracranial 

stenting. 

KEY WORDS: Intracranial artery morphology, human, canine 

Poster 69 

Effect of Scan Direction of Multidetector-Row CT for 

Image Quality of Three-Dimensional Head CT 

Angiography 

Araki, Y. 1 · Suzuki, Y. 1 · Tomitaka, E. 1 · Yoshimatsu, S. 1 · 

Hirai, T. 2 · Yamashita, Y. 2 

1National Hospital Organization Kumamoto Medical Center, 

Kumamoto, JAPAN, 2Kumamoto University Graduate 

School of Medical Sciences, Kumamoto, JAPAN 

PURPOSE 

In three-dimensional (3D) head CT angiography, CT attenuation 

of the intracranial vessels and overlap of the intracranial 

arteries and veins affect the image interpretation. Since 

scan time of multidetector-row CT is quite short, scan direction 

might be important for obtaining adequate image qualities. 

However, there are no reports that evaluate the effect of 

scan direction on the image quality of 3D head CT angiography. 

The purpose of this study was to determine whether 

scan direction of multidetector-row CT affects the image 

quality of 3D CT angiography for the intracranial arteries. 


Eighteen 3D head CT angiograms of 18 patients who underwent 

multidetector-row CT were evaluated prospectively. 

They had been assigned randomly to be given one of two 

protocols: cranio-caudal direction (protocol A, n = 9) and 

caudal-cranial direction (protocol B, n = 9). A CT scanner 

with ten detector rows (SOMATOM Sensation 10; Siemens, 

Erlangen, Germany) was used. The row data were acquired 

with following parameters: 120 kV, 100 eff.mAs, a collimation 

of 0.75 mm × 10, a field of view of 15 cm, a rotation 

speed of 0.5 second per rotation, a feed/rotation of 5.3 mm, 

a slab thickness of 5 cm, and a scan time of 5.59 second. 

Automated detection of bolus arrival in the internal carotid 

artery was used in all cases; the threshold was 100 HU. Two 

radiologists independently graded vessel visualization of the 

intracranial major arteries and veins on a 4-point scale: 1, 

unacceptable image quality; 2, marginally acceptable image 

323 

quality; 3, good image quality; and 4, excellent image quality. 

Vessel attenuation values in the intracranial major arteries 

and veins were measured quantitatively. For data analysis, 

Student’s t-test was performed. 

RESULTS 

In qualitative grading, there were no statistically significant 

differences between the two readers and between the two 

protocols. The mean attenuation value of the basilar artery 

for protocol A was significantly higher than that for protocol 

B (p < 0.05). In the internal carotid and middle cerebral 

arteries and the transverse and straight sinuses, however, 

there were no statistically significant differences between the 

two protocols. 

CONCLUSION 

Cranio-caudal scan direction with multidetector-row CT 

might be beneficial in the evaluation of 3D CT angiograms 

for the posterior circulation. 

KEY WORDS: CT angiography, multidetector-row CT, 

intracranial aneurysms 

Poster 70 

Gender Findings in Patients with Intracranial Dural 

Arteriovenous Fistula 

Klurfan, P. · Shelef, I. · Gunnarsson, T. · TerBrugge, K. · 

Willinsky, R. 

University of Toronto 


PURPOSE 

Intracranial hemorrhage due to intracranial dural arteriovenous 

fistulas (DAVFs) is associated with the presence of retrograde 

leptomeningeal venous drainage. In this study, the 

authors performed a retrospective analysis of 280 patients 

with intracranial DAVFs. Gender was reviewed in relationship 

to the Borden type as well as hemorrhagic and nonhemorrhagic 

presentations. 

Posters

Posters 


The clinical records and imaging studies of 280 patients with 

intracranial DAVFs explored and/or treated at our institution 

between 1984 and November 2004 were reviewed. The following 

data were recorded: age, gender, Borden grade, and 

hemorrhagic presentation. The chi-square test was used to 

determine statistical significance. 

RESULTS 

There were 144 males (51.4%) and 136 females (48.6%). 

Among the male group 56 (38%) had Borden type 1, 25 

(17%) were Borden type 2 and 65 (45%) Borden type 3. In 

the female group, 80 (59%) patients had Borden Type 1, 30 

(22%) had Borden type 2 and 19 (14%) had Borden type 3. 

In the male group with either Borden type 2 or 3, 36 (40%) 

patients presented with hemorrhage, while in the female 

group with either Borden type 2 or 3, 5 (10%) of the patients 

presented with hemorrhage. 

CONCLUSION 

Overall the intracranial DAVFs were divided equally in gender. 

Males had a significantly higher incidence of either 

Borden type 2 or 3 compared to females. This male predominance 

was most striking in the Borden type 3. In addition to 

the hemorrhage risk related to the Borden grade, the male 

gender is a strong risk factor for hemorrhagic presentation 

compared to females. 

KEY WORDS: DAVF 

Poster 71 

Subarachnoid Hemorrhage of Uncommon Etiology: 

Detection of Causative Vascular Lesion with Contrast- 

Enhanced 3D Gradient-Echo Technique 

Yamura, M. 1 · Hirai, T. 1 · Kitajima, M. 1 · Korogi, Y. 2 · 

Hayashida, Y. 1 · Yamashita, Y. 1 




JAPAN 

PURPOSE 

The purpose of this study was to evaluate the usefulness of 

contrast-enhanced three-dimensional (3D) gradient-echo 

(GRE) techniques to detect uncommon vascular lesions presented 

with subarachnoid hemorrhage. 


Six patients (2 males and 4 females, 19-68 years old) who 

presented with subarachnoid hemorrhage were evaluated by 

conventional spin-echo MR imaging, MR angiography, and 

contrast-enhanced 3D magnetization prepared rapid acquisition 

gradient-echo (MPRAGE) images and/or 3D fast imaging 

with steady-state precession (FISP) images. The vascular 

lesions included perimedullary AVF in four, cerebral 

aneurysms associated with Behçet’s disease in one, and 

unknown cause in one. Two neuroradiologists evaluated 

detectability of vascular lesions and the findings of 3D GRE 

images were compared with those of catheter angiography. 

324 

RESULTS 

Only 2 of 6 patients had abnormal vascular findings on conventional 

spin-echo MR imaging and/or MR angiography. In 

5 of 6 patients, abnormal vessels adjacent to the brain or 

spinal cord were detected by contrast-enhanced 3D GRE 

images. These 5 lesions were confirmed by catheter angiography 

and surgery. In one patient who had no abnormal vessels 

on 3D GRE images, catheter angiography did not show 

abnormal findings. 

CONCLUSION 

Contrast-enhanced 3D GRE images were useful for detecting 

uncommon vascular lesions presented with subarachnoid 

hemorrhage. 

KEY WORDS: Subarachnoid hemorrhage, contrast-enhanced 

MR imaging, 3D gradient-echo technique 

Poster 72 

Imaging of Stents Using Angiographic CT 

Benndorf, G. · Strother, C. M. · Naeini, R. · Morsi, H. · 

Klucznik, R. · Mawad, M. 

Baylor College of Medicine 

Houston, TX 

PURPOSE 

To evaluate the feasibility of angiographic CT (ACT) for 

visualization of small devices used for endovascular therapy. 


Angiographic CT was performed immediate postprocedure 

in three patients who underwent intracranial (2) and 

extracranial (1) stent placement for treatment of atherosclerotic 

lesions. Using a bi-plane angiographic system (Axiom 

Artis dBA, Siemens Medical Solutions, Erlangen), rotational 

radiography was performed with new commercially available 

software (Dyna-CT) and the following parameters: 10- 

20 sec, 0.4° increment, 512 matrix in projections, 220° total 

angle, 20°/s, ~ 15-30 frames/s, total of 300-537 projections. 

Image postprocessing was performed to correct scattered

adiation, beam hardening, and ring artifacts on a commercially 

available workstation (Leonardo, Siemens Medical 

Solutions, Erlangen). 

RESULTS 

Angiographic CT provided excellent contrast resolution to 

visualize precisely 2 intracranial stents as well 1 stent in the 

head and neck area. Three-dimensional orientation of each 

stent could be well appreciated using multiplanar reformatted 

images. Visualization of the degree of stent deployment 

and wall apposition was superior to conventional angiography. 

Low-contrast resolution further provided additional 

information about wall calcification proximal and distal to 

the stented vessel segment. 

CONCLUSION 

Angiographic CT is superior to conventional angiography in 

visualization of small endovascular devices such as intracranial 

stents. Representing the first angiographic technique for 

identifying the calcified arterial wall in an angiographic 

suite, it may become an important diagnostic tool for imaging 

pre and poststenting as well for follow up. 

KEY WORDS: Angiographic CT (ACT), flat detector, 

intracranial stents 

Poster 73 

3D CT Angiography for Triage of Intracranial Aneurysm 

to Surgical Versus Endovascular Treatment 

Kim, P. E. · Kim, A. K. Y. · Fassihi, A. A. · Larsen, D. W. · 

Teitelbaum, G. P. · Go, J. L. · Zee, C. 


Los Angeles 


PURPOSE 

CT angiography (CTA) provides high spatial resolution and 

extensive 3D information not available by conventional 

catheter digital subtraction angiography (DSA). We studied 

the role of CTA to determine suitability of patients for coil 

embolization to treat aneurysm(s) following subarachnoid 

hemorrhage (SAH) between March and August 2004. 


Fourteen patients (50% female; 50% Hispanic; 29% Asian) 

presented on average post-bleed day 2 and underwent admission 

brain CTA. Twenty-one aneurysms were identified 

(43% MCA; 33% ACOM). Eighty-one percent of aneurysms 

were deemed not suitable for coil embolization based on 

location and neck-to-dome ratio. By CTA, four aneurysms 

were considered suitable for coil embolization on basis of 

neck-to-dome ratio, location, and anatomical variation. All 

were coiled successfully. One ACOM aneurysm was determined 

to be unsuitable for coil embolization by its broad 

neck and underwent trial coil embolization without success. 

RESULTS 

CT angiography shortened the time from admission to diagnosis 

of the presence, type, and location of aneurysm(s) in 

patients postaneurysmal SAH. Upon admission, CTA is performed 

simultaneously with head CT in patients suspicious 

for SAH. The presence of an aneurysm in some cases can 

325 

pre-empt lumbar puncture in highly suspicious candidates 

without overt SAH on CT. Seventy-one percent of patients (n 

= 10) bypassed cerebral angiography and proceeded directly 

to clip ligation. Average nonionic contrast use for CTA was 

119 ml compared to 90 ml for angiography. 

CONCLUSION 

CT angiography affords a rapid screening method for triage 

of intracranial aneurysms to the appropriate treatment. 

Analysis of the location and dome-to-neck size ratio of the 

aneurysm is performed easily and allows for rapid triage of 

patients to endovascular coiling or clip ligation while often 

avoiding the risks of diagnostic neuroangiography. 

KEY WORDS: CT angiography, aneurysm, triage 

Poster 74 

Head CT Angiography: Anatomical Review 

Bonfante, E. · Nguyen, H. · Sitton, C. W. · Cacayorin, E. · 

Hochhauser, L. · Weir, R. 

Memorial Hermann Hospital 

Houston, TX 

PURPOSE 

The objective of this exhibit is to illustrate normal vascular 

anatomy and variants, with emphasis on the venous system 

in multidetector CT angiography (CTA) of the head. 


We present axial source images and multiplanar MIP reconstructions 

obtained in an 8-detector scanner to demonstrate 

anatomical structures and analyze potential pitfalls. 

RESULTS 

The simultaneous opacification of the arterial and venous 

structures warrants a detailed knowledge of the vascular 

anatomy for an appropriate interpretation of the images. 

Careful review of the source images and multiplanar overlapping 

MIP slabs is required. In our experience, volume 

rendering images have limitations because they are timeconsuming 

and operator-dependent. 

CONCLUSION 

Multidetector CTA of the head allows rapid and noninvasive 

evaluation of the intra and extracranial vascular anatomy and 

pathology. 

KEY WORDS: CT angiography, anatomy, venous 

Posters

Posters 

Poster 75 

New Concept for Aneurysmal Flow Imaging with 16, 32- 

Row Multislice Helical CT 

Hayakawa, M. 1 · Ida, Y. 2 · Negoro, M. 1 · Katada, K. 1 · 

Oshima, M. 3 · Torii, R. 3 · Irie, K. 1 · Kanno, T. 1 · Sano, H. 1 · 

Shojima, M. 3 

1 2 Fujita Health University, Aichi, JAPAN, Fujita Health 

University Hospital, Aichi, JAPAN, 3The University of 

Tokyo, Tokyo, JAPAN 

PURPOSE 

Due to the extremely high speed of 16, 32-row Multislice 

helical CT (MSH CT), when a large cerebral aneurysm with 

slow blood flow is examined, scanning may be completed 

before the aneurysm is completely filled with contrast medium. 

In the present study, we attempted to develop a new 

method in which the CT values within the aneurysm are 

divided into small intervals in order to obtain information 

concerning aneurysmal blood flow. This results were compared 

with computer simulation and angiography. 


Three patients were scanned by 16, 32-row MSH CT scanner 

(Aquilion, Toshiba Corporation). Three-dimensional 

reconstruction was performed using the volume-rendering 

method. Computer simulation or DSA were performed with 

same patients. Three giant or large aneurysms were selected 

at this study. 

RESULTS 

All volume rendering of aneurysmal flow imaging were similar 

with computer simulation and DSA. 

326 

CONCLUSION 

Multislice helical CT is feasible to fast scan. It is not only 

fast examination but also could visualize aneurysmal flow 

image with volume rendering. This method is very simple 

and low cost compared with computer simulation or DSA. 

Usefulness of visualizing the aneurysmal flow image was 

described any paper with model and computer simulation. 

Our method suggests new trend for 3D CT angiography. 

KEY WORDS: Intracranial aneurysm, multislice helical CT, 

flow image 

Poster 76 

Another Approach to Determining Sensitivity of CT 

Angiography for Aneurysm Detection 

Kim, P. E. · Fassihi, A. A. · Go, J. L. · Zee, C. 


Los Angeles 


PURPOSE 

Digital subtraction angiography (DSA) is considered the 

gold standard for detection of intracranial aneurysms. CT 

angiography (CTA) has been reported to be less sensitive for 

smaller aneurysms than DSA in both older and recent 

reports. However, analysis of the image architecture of

angiograms using the most current CTA methods, particularly 

near-isotropic voxel acquisitions, suggests that CTA 

should be at least equal to DSA in sensitivity for detection of 

aneurysms measuring as small as 1-1.5 mm. With the addition 

of 3D information, which is virtually absent in DSA, the 

overall robustness of CTA as a methodology for aneurysm 

detection theoretically surpasses DSA. Because many institutions 

have now adopted CTA in lieu of DSA for initial 

evaluation of these patients, some of which are treated without 

undergoing DSA, a comprehensive comparison of these 

two methods is problematic. However, because the incidence 

of multiple aneurysms detected by angiography has been 

widely reported historically, comparing the rate of multiple 

aneurysms detected by CTA may yield inferences regarding 

its sensitivity vis-à-vis DSA. 


All patients were referred for evaluation of subarachnoid 

hemorrhage as well as aneurysms discovered incidentally by 

other means such as MRA. Patients underwent CTA using a 

16-channel CT scanner. One hundred twenty-five cc of contrast 

were administered at 4 cc per second for each patient 

using automatic bolus timing. Images were acquired at 0.62 

mm collimation with ~0.3 mm overlap (effective pitch ~ 

0.5). Postprocessing was performed on a Vitrea 2 workstation 

(Vital Images, Plymouth, MN) which included 2D planar 

maximal intensity projection (MIP) reconstructions formatted 

in multiple planes, 3D MIP and 3D volume rendering. 

Postprocessing and image interpretation was performed 

by a neuroradiologist experienced in CTA. The overall incidence 

of multiple intracranial aneurysms as well as the incidence 

in patients with subarachnoid hemorrhage were evaluated 

and compared to historical reports that used conventional 

angiography and DSA. Total aneurysm incidence, as 

well as the percentage of patients with multiple aneurysms 

was compared with historical data from the literature. 

Twenty-two patients also underwent DSA, and these results 

were compared. 

RESULTS 

Of the 75 patients examined, 20 showed no evidence of 

aneurysm. Thirty-four (45%) had one aneurysm. Twelve 

(16%) had 2 aneurysms, 5 (6.7%) had 3 aneurysms. Two 

(2.7%) patients had 4 and 2 (2.7%) had 5 aneurysms. Of the 

55 patients who had a at least one aneurysm, 21 (38.2%) 

were found to have multiple aneurysms. The largest 

aneurysm evaluated measured 12.1 mm and the smallest 1.4 

mm. In the 22 patients who underwent DSA, no additional 

aneurysms were detected. 

CONCLUSION 

The rate of multiple aneurysm detection by CTA compares 

favorably with historical angiographic data. In the limited 

number of patients who underwent DSA, no additional 

aneurysms were identified by conventional angiography. 

Because the image architecture of the most advanced CT 

angiograms, which is a combination of submillimeter spatial 

resolution and 3D spatial relationship information (absent in 

DSA), is theoretically more robust than DSA, CTA may be 

more sensitive than DSA for the detection of intracranial 

aneurysms. 

KEY WORDS: CT angiography, multiple aneurysm, digital 

subtraction angiography 

327 

Poster 77 

Willisian Collaterals May not Predict Intraoperative 

Electroencephalogram 

Endarterectomy 

Changes during Carotid 

McGarvey, M. L. 1 · Ances, B. M. 1 · Thurber, A. B. 1 · 

McNicholl, S. T. 1 · Kornegay, A. L. 1 · Mylett, K. A. 1 · Messé, 

S. R. 1 · Litt, B. 1 · Liebeskind, D. S. 1,2 

1 University of Pennsylvania, Philadelphia, PA, 2 University 

of California Los Angeles, Los Angeles, CA 

PURPOSE 

Intraoperative electroencephalogram (EEG) monitoring may 

reveal subtle changes associated with cerebral ischemia during 

carotid endarterectomy. Willisian collateral flow patterns 

on cerebral angiography may be predictive of ischemic risk, 

yet the role of noninvasive approaches such as MRA remains 

unclear. We investigated the predictive value of willisian collaterals 

on TOF MRA for intraoperative EEG changes during 

carotid endarterectomy. 


Time-of-flight MRA of the circle of Willis was acquired in 

73 cases of carotid stenosis (39 left, 34 right) prior to 

endarterectomy (median age 69 years, range 42-91 years; 39 

men, 34 women). Patency of willisian segments and the 

proximal cerebral arteries was categorized on a 3-point scale 

in blinded fashion with respect to clinical and electrographic 

data. Intraoperative monitoring by a neurologist classified 

abnormal EEG based on a 50% or greater attenuation of 

background preclamp activity. Multivariate logistic regression 

analyses were used to predict EEG abnormality based 

on clinical and MRA data. 

RESULTS 

Electroencephalogram abnormalities were identified in 

10/73 cases, including 2/10 with contralateral ICA occlusion. 

Electroencephalogram changes were unrelated to age 

or gender. ACOMM patency did not correlate with a lack of 

EEG changes. Ipsilateral PCOMM patency (17/73 cases) 

also was unrelated to subsequent EEG abnormalities, 

although a potential protective effect was noted in cases of 

contralateral ICA occlusion (p = 0.14). Patency of other vascular 

segments was not predictive of EEG changes, although 

an inverse trend was noted suggesting that EEG changes 

were less likely with diminished conspicuity of the ispilateral 

ICA. 

CONCLUSION 

Time-of-flight MRA demonstration of willisian segments 

may not be predictive of intraoperative EEG abnormalities 

during carotid surgery. Willisian collateral function may 

require serial examination, yet leptomeningeal collateralization 

may account for the diminished likelihood of EEG 

changes with more severe carotid stenoses. 

REFERENCES 

1. Rutgers DR, Blankensteijn JD, van der Grond J. Preoperative 

MRA flow quantification in CEA patients: flow differences 

between patients who develop cerebral ischemia and 

patients who do not develop cerebral ischemia during crossclamping 

of the carotid artery. Stroke 2000;31(12):3021-3028 

Posters

Posters 

2. Lee JH, Choi CG, Kim do K, Kim GE, Lee HK, Suh DC. 

Relationship between circle of Willis morphology on 3D 

time-of-flight MR angiograms and transient ischemia during 

vascular clamping of the internal carotid artery during 

carotid endarterectomy. AJNR Am J Neuroradiol 

2004;25(4):558-564 

3. Kim GE, Cho YP, Lim SM. The anatomy of the circle of Willis 

as a predictive factor for intra-operative cerebral ischemia 

(shunt need) during carotid endarterectomy. Neurol Res 

2002;24(3):237-240 

KEY WORDS: Collateral, EEG, endarterectomy 

Poster 78 

Evaluation of Intracranial Arterial Stenosis: Comparison 

between Time-of-Flight and Wide-Range Contrast- 

Enhanced MR Angiography 

Kim, B. · Jung, S. · Lee, J. · Ahn, K. · Yoo, W. · Byun, J. 

The Catholic University of Korea 


PURPOSE 

Time-of-flight MR angiography (TOF MRA) and contrastenhanced 

MRA (CE MRA) are noninvasive imaging modalities 

for evaluation of arterial stenosis. Innovative techniques 

and improvements in image quality has enabled high-resolution, 

wide-range field-of-view of CE MRA covering from 

the aortic arch up to the circle of Willis. We evaluated CE 

MRA covering from the aortic arch up to the circle of Willis 

and TOF MRA of circle of Willis compared with DSA for 

detection of intracranial arterial stenosis. 


Two readers, blinded to outcome of other studies, independently 

evaluated stenoses on DSA, TOF MRA, and CE MRA 

of 25 patients (18 male and 7 female; 32 ~ 80 years old; 

median, 62 years). Circle of Willis including proximal anterior 

and middle cerebral arteries, distal internal carotid 

artery, vertebrobasilar and proximal posterior cerebral arteries 

were evaluated for the stenosis. Stenosis of 50% or more 

was categorized into positive result. Results of both modalities 

were compared with the corresponding DSA findings. 

RESULTS 

Intracranial arterial stenosis greater than 50% was identified 

in 24 of 400 intracranial arteries at DSA. Time-of-flight 

MRA had a sensitivity of 75.0%, a specificity of 94.7%, and 

an accuracy of 87% for the identification of intracranial 

stenosis, whereas the CE MRA yielded values of 87.5%, 

92.0%, and 86%, respectively. 

CONCLUSION 

Contrast-enhanced MRA covering from aortic arch to the 

circle of Willis has reliable accuracy in evaluation of 

intracranial arterial stenosis comparable to that of TOF MRA 

and DSA. 

KEY WORDS: MR angiography 

328 

Poster 79 

Bone Subtraction CT Angiography Techniques for Skull 

Base Vascular-Related Pathology: Initial Clinical 

Experience 

Nguyen, D. T. D. 1 · Choi, J. J. 2 · Lindisch, D. J. 2 · Iuliano, E. 

M. 1 · Sarcone, A. 2 · Mun, S. K. 2 · Cleary, K. R. 2 

1 Georgetown University Hospital, Washington, DC, 

2 Georgetown University, Washington, DC 

PURPOSE 

Assessing bone subtraction CT angiography (bsCTA) technique 

in determining intracranial vascular pathology within 

and adjacent to skull base osseous structures. 


We selected patients who have possible vascular pathology 

at the skull base level to perform bsCTA. Noncontrast and 

contrast-enhanced CT scans were sequentially performed 

with identical scanning protocols on a 4-slice multidetector 

CT unit. A commercial software was used to perform the 

bone subtraction. The subtracted vessel-only volume was 

then imported into a dedicated 3D workstation for final multiple 

volumetric rendering viewings. The bone subtraction 

CTA volume dataset then was compared to traditional digital 

subtraction angiography (DSA), the current gold standard, 

and 3D rotational DSA (rDSA) for image correlation. 

RESULTS 

The appreciation of the aneurysm morphologies such as 

overall size, neck origin, and neck size were better seen 

when bsCTA technique was applied to intracranial CTA. 

These important parameters showed similar image correlation 

on bsCTA as they were on the 3D rDSA images. 

CONCLUSION 

Traditional nonsubtracted CTA is limited at the level of skull 

base where osseous structures are the predominant factor. 

The bsCTA technique enhances visualization of vascular 

pathology within this level. This subtraction technique is 

robust, operator nondependent, and requires little additional 

postprocessing effort. Most importantly, it adds additional 

diagnostic certainty to our noninvasive CTA armamentarium. 

KEY WORDS: CT angiography, aneurysm, bone subtraction

Poster 80 

Three-Dimensional Dynamic MR Subtraction 

Angiography Using Sensitivity Encoding for the 

Evaluation of Intracranial Arteriovenous Malformations 

Taschner, C. A. 1 · Gauvrit, J. 1 · Oppenheim, C. 2 · Meder, J. 2 · 

Leclerc, X. 1 

1Hôpital Salengro, University Hospital Lille, Lille, 

FRANCE, 2Hôpital Sainte-Anne, Paris V University, Paris, 

FRANCE 

PURPOSE 

To develop an 3D dynamic MR digital subtraction angiography 

(MR DSA) using sensitivity encoding for the evaluation 

of cerebral arteriovenous malformations (cAVMs). 


Nineteen patients with 19 angiographically proven cAVMs 

(16 supratentorial and 3 infratentorial) were assessed by conventional 

catheter angiography (CCA) and 3D dynamic MR 

DSA. Three-dimensional contrast-enhanced gradient-echo 

sequence with sensitivity encoding based on parallel imaging 

technique was performed and provided an acquisition of 

a volume of 20 dynamic images repeated 18 times every 1.7 

seconds. Three-dimensional dynamic MR DSA were analyzed 

independently by two radiologists in a blinded fashion 

with regard to cAVM nidus and venous drainage. 

Conventional catheter angiography was used as reference. 

RESULTS 

All MR examinations were assessable. Interobserver agreement 

was excellent for the detection of nidus and for the 

evaluation of nidus size (κ = 1 and 0.875 respectively) but 

moderate for the visualization of the venous drainage (κ = 

0.56). All nidus detected on CCA were clearly depicted on 

3D dynamic MR DSA. The evaluation of size of nidus by 

both techniques was similar. On 3D dynamic MR 

angiograms, veins were analyzed correctly in 17 out of 19 

cAVMs. 

CONCLUSION 

Our preliminary study demonstrates that the 3D dynamic 

MR DSA using sensitivity encoding technique with a high 

spatial resolution is suited for the assessment of cAVMs. 

KEY WORDS: MR subtraction angiography, sensitivity 

encoding, arteriovenous malformations 

Poster 81 

Selection of Optimal Treatment Strategy for Intracranial 

Aneurysms: Proposal of a New Classification System 

Based on Helical CT Angiography 

Villablanca, J.P. · Martin, N. · Jahan, R. · Frazee, J. · Sayre, J. 

University of California Los Angeles, Center for the Health 

Sciences 


PURPOSE 

To employ helical CT angiography (CTA) to identify 

aneurysm characteristics predictive of successful endovascular, 

neurosurgical, and combined treatment strategies. 

329 


Between May 1997 and March 2004, 658 patients underwent 

prospective CTA for suspected intracranial aneurysm usign a 

routine clinical protocol. Two-dimensional and 3D data 

analysis was performed using standardized protocol. Twodimensional 

and 3D images were used selectively for quantitation, 

definition of multidimensional neck and sac size and 

geometry, dome-to-neck ratio, presence of arterial incorporations, 

relationship to local arteries and bony anatomy, 

intraluminal thrombus and mural calcification. 

RESULTS 

Circle of Willis (523), mycotid (7), and posterior circulation 

aneurysms (127) were included. Based on morphologic 

analysis we propose a 4-tier classification scheme based on 

2D and 3D CTA. Type I aneurysms are triaged to endovascular 

treatment and are characterized by neck diameter of < 

= 4 mm, a dome-to-neck ratio of > = 2, lack of extensive 

intraluminal thrombus or fully incorporated arterial branches, 

and may have mural calcifications and may be intra or 

extradural. Type II aneurysms are triaged to neurosurgical 

clipping. These aneurysms possess neck diameters > 4 mm, 

dome-to-neck ratios < 2, absence of mural calcification or 

thrombus at the neck, an accessible neck via craniotomy, and 

possible incorporation of arterial branches into the aneurysm 

neck or sac. Type III aneurysms require combined endovascular 

and neurosurgical approaches due to the presence of 

key arterial incorporations, neck calcifications, extensive 

thrombus and/or mural calcification extending to the neck, 

are fusiform in shape, or are not accessible via craniotomy. 

Combined approaches include intracranial to extracranial 

bypass followed by neurosurgical trapping of the aneurysm 

or endovascular occlusion of the parent artery. Type IV 

aneurysms are untreatable by current techniques, generally 

due to the presence of multiple incorporated arterial segments 

into the body of the aneurysm. 

CONCLUSION 

CT angiography provides precise information about intracranial 

aneurysms, surrounding arteries, and local bony anatomy. 

This data can be used to triage patients to neurosurgical, 

endovascular, and combined modality treatments. 

KEY WORDS: CTA, aneurysm, treatment planning 

Posters

Posters 

Functional 

82-102 

Poster 82 

Functional MR Imaging Evaluation of Deep Brain 

Stimulation Therapy for a Variety of Neurologic 

Disorders 

Phillips, M. D. 1 · Baker, K. B. 1 · Lowe, M. J. 1 · Kopell, B. 2 · 

Malone, D. 1 · Greenberg, B. D. 3 · Borg, B. 1 · Tkach, J. A. 1 · 

Rezai, A. R. 1 

1 2 Cleveland Clinic Foundation, Cleveland, OH, Medical 

College of Wisconsin, Milwaukee, WI, 3Brown Medical 

School, Providence, RI 

PURPOSE 

To review the MR imaging safety and functional MR imaging 

(fMRI) patterns of activation produced by deep brain 

stimulation (DBS) in a variety of neurologic disorders. 


All imaging was performed on a 3 T Siemens Allegra MR 

unit (Erlangen, Germany). Prior to any human experiments 

extensive safety testing using a previously described gel 

phantom was performed using the exact imaging sequences 

and wiring for DBS leads. A total of 10 patients with bilateral 

DBS placement for a variety of disease processes (7 with 

Parkinson’s disease, 1 with a central tremor, 1 with obsessive 

compulsive disorder and 1 with depression) were examined 

using fMRI. Subjects were tested with percutaneously 

extended bilateral DBS electrodes on the first or second 

postoperative day. The externalized lead system was extended 

through the waveguide to an external pulse generator in 

the MR imaging control room. Scanning consisted of threedimensional 

anatomical data set with leads disconnected 

from the pulse generator and BOLD fMRI with a single lead 

connected to the pulse generator. BOLD images were 

acquired using prospective motion correction. BOLD images 

were acquired for each DBS lead separately for all conditions. 

Functional MR imaging examinations were performed 

with a block style paradigm consisting of 5 stimulator off 

and 4 stimulator on 32 second epochs with 10 seconds 

placed between the stimulator off and stimulator on conditions 

in order to gradually ramp the stimulator to the optimal 

stimulation level. Gradual ramping was employed for patient 

comfort and to decrease patient motion. Images acquired 

during the ramping process were discarded prior to image 

analysis. The MR imaging time series at each pixel was fit 

using least squares to a boxcar reference function plus a 

slope and intercept. All subjects received neurologic examinations 

immediately prior to and after the fMRI evaluations. 

RESULTS 

All imaging sequences tested produced less than 2 °C of 

heating. All subjects tolerated the fMRI examinations well 

with no change in neurologic examination from the pre- to 

post-fMRI evaluation. Good activation was acquired in all 

subjects. Patterns activation differed according to placement 

of the DBS electrodes and underlying neurologic abnormalities. 

In the case of multiple subjects with a similar disease 

efficacious stimulation produced a consistent pattern of acti- 

330 

vation. In cases when nonefficacious stimulation was compared 

to effective stimulation the patterns of activation produced 

on fMRI were different. 

CONCLUSION 

Functional MR imaging can be performed safely in patients 

receiving DBS for a variety of disease processes. Deep brain 

stimulation is becoming increasingly prevalent as a method 

for treating intractable neurologic diseases. Functional MR 

imaging offers a unique opportunity for the in vivo study and 

understanding of the mechanisms of action for DBS as well 

as assessing the clinical efficacy in an objective noninvasive 

manner. 

KEY WORDS: Functional MR imaging, deep brain stimulation 

Poster 83 

Automated versus Manually Optimized Selection of 

Metabolite Peak Boundaries in MR Spectroscopy 

Evaluation of Enhancing Brain Lesions 

Foerster, B. · Petrou, M. · Pang, Y. · Sundgren, P. C. 


Ann Arbor, MI 

PURPOSE 

To explore the difference in 2D CSI Cho/Cr, Cho/NAA, and 

NAA/Cr ratios calculated with automated selection vs manually 

optimized selection of metabolite peak boundaries in 

patients with enhancing brain lesions. 


Two-dimensional CSI MR spectroscopy (PRESS, TE 

144/TR 1000) was performed in 21 patients (11 male, 8 

female aged 4-54 years, mean 33.4 years). Each patient had 

newly found contrast-enhancing lesions in the pons, posterior 

fossa, or supratentorial region at the site of a previously 

diagnosed and treated primary brain neoplasm. Histologic 

types included glioma, astrocytoma, ependymoma, and 

medulloblastoma. Within the VOI, smaller voxels (1 x 1 x 1 

cm) were manually placed in the contrast-enhancing lesion. 

In each patient, we computed the metabolite ratios (Cho/Cr, 

Cho/NAA, and NAA/Cr) using the automated vender-provided 

software (Functool 2000 by GE) to generate default 

metabolic peak limits/boundaries and integrate the area 

under the curve for each of the metabolites. These “default” 

metabolite peak limits/boundaries are often systematically 

shifted, voxel by voxel, due to magnetic field inhomogeneities. 

We therefore labeled the ratios generated by the 

default limits/boundaries as the “uncorrected” metabolite 

ratios. To reduce the systematic error, we manually adjusted 

the peak limits/boundaries to cover the whole metabolite 

peak area in each patient. We then re-computed the metabolite 

ratios (Cho/Cr, Cho/NAA, and NAA/Cr) for each of the 

patients using the adjusted set of values. We labeled this set 

of ratios as the “corrected” metabolite ratios. The overall 

mean and standard deviation for each of the metabolite ratios 

(Cho/Cr, Cho/NAA, and NAA/Cr) then were calculated and 

analyzed using the Student T test.

RESULTS 

The mean corrected Cho/Cr and Cho/NAA ratios were significantly 

lower than the mean uncorrected Cho/Cr and 

Cho/NAA ratios (2.08 vs 2.37, p = 0.006) (2.88 vs 3.34, p = 

0.009). There were no significant differences between the 

mean corrected and the mean uncorrected NAA/Cr ratios. 

CONCLUSION 

The significant difference observed between the mean corrected 

and mean uncorrected Cho/Cr and Cho/NAA ratios 

demonstrates the importance of manually selecting the 

boundary limits of the metabolites, particularly choline. We 

often found that the automatic determinations of the metabolite 

peak boundaries are not optimally centered. The automated 

boundary settings can falsely elevate the Cho/Cr and 

Cho/NAA ratios which could potentially affect interpretation 

of the spectroscopy results. We would therefore recommend 

manually setting the metabolite boundary parameters to 

obtain the most accurate values. 

KEY WORDS: MR spectroscopy, neoplasm, postradiation 

changes 

Poster 84 

Magnocellular and Parvocellular Visual Pathways Have 

Different Bold Signal Time Courses in Human Primary 

Visual Cortex 

Liu, C. J. · Bryan, R. N. · Woo, J. · Liu, G. T. · Elliott, M. A. 



PURPOSE 

The magnocellular and parvocellular pathways (M and P 

pathways) are the major pathways of the visual system, 

accounting for most of the axons that leave the retina and the 

perceived vision. Histologically, the M and P pathways are 

distinct. Physiologically the M pathway is considered insensitive 

to color when the luminance is balanced, has higher 

contrast sensitivity, and is responsive to lower spatial and 

higher temporal frequencies, whereas the P pathway is color 

sensitive, has lower contrast sensitivity, and is responsive to 

higher spatial and lower temporal frequencies. We hypothesize 

that the histologic and physiologic differences of the 

two visual pathways would also manifest as differences in 

the signal time course of blood oxygen level dependent functional 

MR imaging (BOLD fMRI) in the commonly activated 

region of human primary visual cortex (V1). The differences 

in BOLD signal time course may provide insight into 

the metabolic requirements of the two pathways. 


Eleven fMRI sessions on six subjects were performed on a 

Siemens Trio 3 T scanner. All subjects gave informed consent. 

Gradient-echo EPI was used for functional imaging. 

The stimulus that preferentially activated the M pathway was 

a black and white reversing checkerboard with 18% contrast, 

2 degree check size, and reversing rate of 19 Hz. The stimulus 

that preferentially activated the P pathway was an isoluminant 

red and green reversing checkerboard, 0.5 degree 

check size, and reversing rate of 2 Hz. The two stimuli were 

presented in a pseudorandom manner, with 20 presentations 

per session. Analysis of functional activation was performed 

331 

with SPM2. The data sets were normalized to a standard 

space, and the regions of activation were determined using 

an unbiased hemodynamic response function (HRF) with 

finite impulse response functions. Regions commonly activated 

by both the M and P stimuli in V1 were determined 

and used for the BOLD signal time course analysis. The contrast 

elicited by the stimulus, time-to-peak (TTP), and the 

full width at half maximum (FWHM) of the HRF, were 

measured for both stimuli. The P signal time course metrics 

were normalized to the values of the M signal time course to 

minimize intersubject differences, and two-tailed one sample 

t tests were performed to determine statistically significant 

differences in the metrics. 

RESULTS 

All data sets exhibited robust response to the M and P stimuli 

and showed characteristic regions of activation. Within 

V1, while the percent contrast elicited by the stimulus and 

FWHM of the HRF were not statistically different for the M 

and P stimulus, the TTP of the HRF was statistically significant, 

with the P stimulus generating TTPs that were on average 

12% faster than the M stimulus (P = 0.0037). 

CONCLUSION 

We have demonstrated the ability to differentiate the M and 

P stimuli in a commonly activated region. As the BOLD 

response is dependent on the ratio of oxyhemoglobin and 

deoxyhemoglobin in the blood, the difference in the BOLD 

time course between the two stimuli suggests that the oxygen 

demand of the two pathways may be different. 

KEY WORDS: BOLD time course, magnocellular/parvocellular 

pathways, oxygen metabolism 

Poster 85 

Visualization of 3D White Matter Tractography by 

Diffusion-Tensor MR Imaging: Comparison between 

Single-Shot Fast Spin-Echo-Based Sequence and Single- 

Shot Echo-Planar Sequence 

Adachi, Y. 1 · Hori, M. 2 · Nakata, Y. 2 · Ishigame, K. 2 · 

Kumagai, H. 2 · Araki, T. 2 

1Fujiyoshida Municipal Hospital, Yamanashi, JAPAN, 

2University of Yamanashi, Yamanashi, JAPAN 

PURPOSE 

The purpose of this study was to investigate and compare the 

susceptibility effect influenced on three-dimensional (3D) 

white matter tractography of single-shot fast spin-echo 

(SSFSE) diffusion-tensor MR imaging (DTI) and the vast 

majority method of DTI, single-shot echo-planar sequence 

(EPI) in the brain at 1.5 T MR imaging. 


A total of 30 patients with neurologic disease participated in 

this study. All MR imaging were performed on a 1.5 T MR 

imager (Signa Lx, GE Medical Systems, Milwaukee, WI) 

with a birdcage head coil. After conventional T2-and T1weighted 

transverse scan, SSFSE-DWI and EPI-DWI were 

performed. Imaging parameters of SSFSE-DWI were as follows: 

TR/TE = 20000/73 ms, matrix 128 x 128, bandwidth = 

32 kHz, FOV = 260 x 260 mm, slice thickness/gap = 5/0 mm 

and b value of 0 and 1000 s/mm 2 with the maximum b value 

Posters

Posters 

applied in 13 directions. Imaging parameters of EPI-DWI 

were as follows: TR/ TE = 10000/ 70 ms, bandwidth = 116 

kHz with the maximum b value applied in 15 directions. As 

written above, FOV, spatial resolution, slice thickness, slice 

gap and b value of EPI-DWI were the same as SSFSE-DWI. 

Total scan time is 4 min. 43 sec and 3 min. for SSFSE-DWI 

and EPI-DWI, respectively. Subsequently, 3D tract projection 

of white matter fibers were demonstrated by using a 

software (dTV 1.5, developed by Image Computing and 

Analysis Laboratory, Department of Radiology, The 

University of Tokyo Hospital, Japan.). SSFSE-DT images 

were each compared with the corresponding EPI-DW 

images using two qualitative criteria, presence of distortion 

artifacts and demonstration of major white matter fibers (i.e., 

cortico-spinal tracts and splenium). This evaluation was performed 

by three experienced neuroradiologists. Distortion 

artifacts were graded on a 5-point scale: 1 signified that brain 

structure was not identified because of distortion artifact; 2, 

brain structure was unclear partially because of distortion 

artifact; 3, presence of distortion artifact but surrounding 

brain structure was identified; 4, presence of a little distortion 

artifacts; 5, no distortion artifacts. Demonstration of 

major white matter fibers were graded also on a 5-point 

scale: 1 signified that no demonstration of major white matter 

fibers; 2, some short fibers were visualized along the 

major white matter fibers; 3, major white matter fibers were 

identified but not completely; 4, major white matter fibers 

were visualized with some extra fibers; 5, major white matter 

fibers were clearly visualized without extra fibers. 

RESULTS 

In all patients, both SSFSE-DTI and EPI-DTI were effective 

in demonstrating white matter fiber-tracking. Mean scale for 

distortion artifact and demonstration of major white matter 

fibers were 4.5 and 3.7 at SSFSE-DWI and 2.3 and 4.2 at 

EPI-DWI, respectively. In particular, SSFSE-DWI were 

preferable methods for estimating the white matter fibers in 

patients who underwent some neurosurgical procedures (i.e., 

tumor resection or clipping intracranial aneurysm) because 

of less artifacts and distortion in the images. 

CONCLUSION 

SSFSE-DTI is a promising method to obtain less artifacts 

and distortions, compared with EPI-DWI in some clinical 

cases. This sequence is useful in particular for the patients 

after neurosurgical procedure. 

KEY WORDS: Diffusion tensor, tractography, single-shot fast 

spin-echo 

Poster 86 

Bidirectional Spatial Filtering: Refining Methods for 

Locating Brain Activation in the Presence of 

Parenchymal Abnormalities 

Walker, S. A. · Tanabe, J. L. · Miller, D. 

University of Colorado 

Denver, CO 

PURPOSE 

Functional MR imaging (fMRI) has become an important 

tool for presurgical mapping of brain tumors. While typical 

experiments are conservative in estimating brain activation 

332 

to avoid type I errors (errors of specificity), presurgical mapping 

increases the importance of accurately locating activated 

brain, especially in the presence of distorted anatomy. We 

have adapted the standard method of spatial filtering to 

include an “edge stopping” function which prevents apparent 

brain activity from blurring across anatomical boundaries 

and into regions of tumor. This has the effect of more 

accurately locating brain activation bordering regions of 

pathology without sacrificing signal-to-noise ratio throughout 

the rest of the brain (1). 


Functional MR data were acquired in a 30-year-old patient 

with known low-grade glioma in the right frontal-parietal lobe 

using a 1.5 T MR system. Simulated data were superimposed 

on images of nonactivated brain using the same block paradigm 

with additive signal equal to between 1% and 5% of the 

mean physiologic background. Before statistical analysis, 

either a standard spatial Gaussian filter or a bidirectional filter 

was applied (both with 8 mm FWHM). The bidirectional filter 

modifies a Gaussian filter by adding an “edge stopping” 

function, effectively smoothing voxels that are near each other 

and similar in intensity, but avoids smoothing across tissue 

boundaries and in regions of sharp activation (2). 

RESULTS 

The figure shows a comparison of Gaussian versus bidirectional 

filtering for a simulated 5% activation signal in 5 voxels 

superimposed on nonactivated brain. The bidirectional 

filter (left) introduced 2 false-positive voxels remote from 

the neoplasm and the motor gyrus as well as two false-positive 

voxels within the surrounding tumoral edema. The 

Gaussian filter (right) introduced 13 false-positive voxels at 

the tumor margin. This demonstrates that the bidirectional 

filter can more accurately localize activation in the presence 

of distorted anatomy. 

CONCLUSION 

Bidirectional filtering can improve the localization of activated 

brain in regions of abnormal tissue such as brain neoplasm. 

This gain is achieved at the same time improving the 

significance of and possibly sensitivity to activation signal in 

the rest of the brain. 

REFERENCES 

1. Parish TB, Gitelman DR, Labar KS, Mesulam MM. Magn Reson 

Med 2000;44: 925-932 

2. Barash D. IEEE Transactions on Pattern Analysis and Machine 

Intelligence 2002;24(6):844-847 

KEY WORDS: Bidirectional, spatial, filtering

333 

Poster 87 

Poster 88 

Automated Functional Imaging: An Interactive Modular Functional MR Imaging Evidence of Functionality and 

Software for Automated Postprocessing of Standardized 

Clinical and Experimental Functional MR Imaging 

Plasticity in Polymicrogyric Cortex 

Araujo, D. · de Araujo, D. B. · Pontes-Neto, O. M. · Rosset, 

Stippich, C. · Nennig, E. 

S. · Leite, J. P. · Sakamoto, A. C. · Santos, A. C. 

University of Heidelberg 

University of Sao Paulo 


Ribeirao Preto - Sao Paulo, BRAZIL 

PURPOSE 

To create an interactive software for clinical and experimental 

functional MR imaging (fMRI) data that automatically 

drives the postprocessing of standardized fMRI measurements 

using commercial postprocessing software 

(BrainVoyager (R) ) and that facilitates the data management 

including data transfer and sorting, automated anonymization 

of patient data and documentation of relevant imaging 

parameters and clinical findings in self-updating lists, automated 

storage of all functional and morphologic MR data 

and print out of CD covers. 


Automated functional imaging (AFI) was designed initially 

for BrainVoyager 2000 (Version 4.9) to facilitate the timeconsuming 

offline postprocessing and data management of 

standardized clinical routine fMRI measurements. Based on 

optimized clinical fMRI protocols this modular interactive 

software was programmed in a mixture of the scripting languages 

(Visual Basic Script, Python 2.3, AutoIt 2.64) and 

controls BrainVoyager via a Microsoft COM-based interface. 

Automated functional imaging performs the following 

steps fully automatically: data transfer from MR terminal to 

PC, data sorting, renaming, anonymization, administration, 

documentation, data analysis using standard parameters, 

generation of fMRI activation maps in 2D, 3D and Talairach 

space, data archiving and creation of CD/DVD covers. Some 

steps intrinsically require manual support and cannot be 

automated. These steps are supported interactively by AFI: 

editing of patient data, coregistration of functional and 

anatomical images and specification of the reference points 

for spatial normalization (Talairach transformation). 

RESULTS 

On a standard PC (2.4 GHz, Pentium) the average total evaluation 

time for a standard clinical fMRI study (4 different 

measurements, 3D data set) is 70 minutes, which is approximately 

50% of the time required for the manual evaluation 

by an expert (range: 140 to 170 minutes). More importantly 

the personnel expenditure time decreases to 16% (23 minutes) 

as the remainder runs automatically. 

CONCLUSION 

Automated functional imaging supports and accelerates the 

complex, time-consuming postprocessing of fMRI data by 

automization, standardization, and simplification. Compared 

to the available online postprocessing software tools the full 

functionality of BrainVoyager can be utilized including all 

overlay and export options (e.g., to neuronavigators). The 

adaptation for “SPM” is planned. 

KEY WORDS: Functional MR imaging, data processing, 

automation 

PURPOSE 

Malformations of cortical development (MCD) are one of 

the most common pathologic findings in refractory epilepsy 

related to foci outside the temporal lobes. MR imaging can 

reliably show these alterations and provide information for 

surgical resection. However, there is increasing evidence that 

neurons in MCD areas are functional and may show different 

organization patterns in comparison to normal brains. 

Our objective was to analyze the patterns of brain plasticity 

and function in polymicrogyric cortex in epileptic patients 

by blood oxygenation level dependent (BOLD) functional 

MR imaging (fMRI). 


Six adult patients (4 women), with clinical and neurophysiologic 

evidence of extratemporal epilepsy, and a radiologic 

diagnosis of polymicrogyria were scanned in a 1.5 T 

Siemens magnet with BOLD fMRI sequences according to 

language (word generation) and motor (finger tapping) protocols. 

All patients were right-handed. Cognitive impairment 

was never severe as to prevent collaboration on the trials. A 

set of 3D structural images was acquired also for coregistration. 

Functional images were processed through Brain 

Voyager . 

RESULTS 

Statistical maps showed significant activation of polymicrogyric 

cortex in the motor strip in the 5 cases in which it was 

present in that location. In two cases language tasks did not 

show any significant activation. In four cases the motor activation 

was bilateral and asymmetric, being more widespread 

in the side of larger abnormalities. In one case the motor activation 

was bilateral and symmetric. This patient had symmetric 

bilateral perysilvyan polymicrogyria. In one case the 

polymicrogyric cortex was restricted to the base of the temporal 

lobe and activation was consistent with the pattern 

found in normal brains. Word generation provided localization 

of Broca’s area in 3 of the 4 patients in whom activation 

was noted. One patient had a large parieto-occipital polymicrogyria 

and word generation showed widespread activation 

in the left hemisphere. 

CONCLUSION 

Functionality was shown in polymicrogyric areas. In some 

cases there was evidence of reorganization to ipsi and contralateral 

areas. Surgical resection may be performed but 

there is need for functional planning due to reorganization 

and functionality of polymicrogyric cortex. 

KEY WORDS: Functional MR imaging, polymicrogyria, plasticity 

Posters

Posters 

Poster 89 

Comprehensive Funtional MR Protocol in Combination 

of Word-Generation, Object-Naming, and Visual 

Memory: Pilot Study 

Kim, B. · Jung, S. · Kim, J. · Jeon, S. · Byun, J. 



PURPOSE 

To assess the feasibility of comprehensive functional MR 

imaging (fMRI) protocol for the language and memory lateralization. 


The protocol was designed to be performed in under 25 minutes 

in standard 1.5 T MR unit. We used three stimulation 

tasks in single protocol to test 7 healthy adult volunteers (6 

right-handed, 1 ambidextrous): word-generation, objectnaming, 

and visual memory. The SPM program was used for 

the postprocessing of images and the threshold for significance 

was set at p < 0.01. A lateralization index was calculated 

from the number of activated pixels in frontal, temporoparietal, 

and medial temporal regions. The results for 

lateralization were compared among stimulation tasks and 

those of group analysis. 

RESULTS 

The functional map from the word-generation task showed 

left dominance in six subjects. Language lateralization was 

better with the word-generation task than object-naming in 

which only two subjects showed left dominance. On group 

analysis, lateralization was better with the word-generation 

task than object-naming. Four of seven subjects showed 

medial temporal activation on visual memory task. 

334 

CONCLUSION 

Word-generation, object-naming, and visual memory tasks 

are feasible to be performed in single protocol. Combined 

results of this comprehensive fMRI will be adequate for a 

comparative study with Wada test. 

KEY WORDS: Functional MR imaging

335 

Poster 90 

strate a possible need of using both protocols to infer lan- 

Language Lateralization in Epilepsy with Functional 

MR Imaging: A Comparison of Word Generation with 

guage lateralization in epileptic patients in order to get more 

conclusive laterality maps. 

Categorization 

KEY WORDS: Language, functional MR imaging, paradigm 

Escorsi-Rosset, S. · David, A. · Santos, A. C. · Sakamoto, A. 

C. · de Araujo, D. B. 

University of Sao Paulo 

Ribeirao Preto - Sao Paulo, BRAZIL 

PURPOSE 

Recent studies have documented the efficacy of functional 

MR imaging (fMRI) for determining hemispheric dominance 

in patients with epilepsy, as demonstrated by good 

correspondence with Wada results. Most of these have used 

tasks involving word generation and object naming. In the 

present study we examined fMRI activation in both word 

generation with letters and categories in a group of patients 

with temporal lobe epilepsy (pre or postsurgery). The primary 

purpose was to ascertain whether the task would differ 

in lateralization concordance with Wada results for such 

patients and to determinate whether the tasks elicit different 

activation in frontal and temporal language areas. 


Fifty epileptic patients and twenty nonsymptomatic subjects 

were included in the study. All images were acquired in a 1.5 

T scanner (Siemens, Magneton Vision). The functional data 

set was obtained in EPI BOLD-like sequences, and was 

superposed onto high spatial resolution images, covering 

both hemispheres, with a 1 mm 3 voxel size. Six blocks of 

approximately 30 seconds of rest were interchanged with 

five blocks activity, either word generation or categorization. 

Clinical, demographic, neuroimaging, neurophysiologic, and 

cognitive performance (IQ, Boston Naming Test) were analyzed 

and compared with both tasks in fMRI. Individual data 

were analyzed with a General Linear Model (GLM) 

approach using the Brain Voyager 4.6 package. 

RESULTS 

Both tasks were capable of detecting lateralization in every 

subject to a different degree. Final statistical maps showed a 

similar location of activation for healthy volunteers, in both 

paradigms. Activation was primarily found on left inferior 

frontal gyrus, medial frontal gyrus, and mesial frontal 

regions, including supplementary motor area. As for the 

epileptic group, besides differences in laterality due to neurophysiologic 

mechanism related to synchronous discharges, 

and, in some cases, to surgical treatment, there was, in some 

exams, a difference in response between both paradigms. 

Some subjects showed a greater response to categorization 

while others to word generation. 

CONCLUSION 

We found language statistical maps to be fundamentally similar 

in location and laterality in healthy volunteers. However, 

either in patients or healthy volunteers, there was an evident 

difference in response to one of the paradigms. The difference 

between paradigms has, apparently, nothing to do with 

cognitive impairment of patients. Different groups of subject, 

with similar cognitive impairment, responded differently 

to the tasks, being paradigm selective. Our results demon- 

Poster 91 

Trigeminal Neuralgic Pain Investigated by Functional 

MR Imaging 

Schlieter, M. · Kress, B. · Nennig, E. · Durst, A. · Schramm, 

P. · Tronnier, V. · Sartor, K. · Stippich, C. 

University of Heidelberg Medical Center 


PURPOSE 

In this clinical trial pain-associated brain activation was 

assessed in patients with chronic trigeminal neuralgic pain 

using functional MR imaging (fMRI) during standardized 

fully automated tactile stimulation of the face within the 

receptive fields of the maxillary (V2) and mandibular (V3) 

division of the trigeminal nerve. 


Six patients with trigeminal neuralgic pain on the right side 

and 6 patients with trigeminal neuralgic pain on the left side 

underwent standardized block-designed BOLD fMRI at 1.5 

T during pneumatically driven fully automated innocuous 

tactile stimulation of the upper and lower lips (1). In each 

patient the stimulation was performed for the symptomatic 

and for the asymptomatic side separately using two different 

fMRI measurements. After spatial normalization and overlay 

of functional on anatomical images the fMRI activation 

maps (volume time courses, vtc-files) of each individual 

patient were analyzed for pain associated and somatosensory 

BOLD activation. Data analysis: BrainVoyager (R) . In 

addition the patients were imaged with a high-resolution 3D 

T2-weighted (CISS) sequence and a 3D T1-weighted 

(VIBE) sequence for volumetric analysis of the trigeminal 

nerve. 

RESULTS 

Innocuous tactile stimulation elicited BOLD activation also 

in pain-processing brain areas, namely in the anterior cingulate 

of both hemispheres, in the insula and in the thalamus 

contralateral to the stimulated side. Most thalamic neurons 

that respond to tactile stimulation of the face are located in 

ventroposterolateral (VPL) nucleus (2) and neurons that 

respond to noxious stimuli are found in large numbers in the 

dorsomedial (DM) nucleus (3). Activation was observed in 

both medial and lateral thalamic nuclei. The stimulus related 

somatosensory activation projected on the ipsilateral spinal 

(spV) nucleus of the trigeminal nerve, the thalamus contralateral 

to the stimulated side, the secondary somatosensory 

cortex bilaterally, and on the contralateral face representation 

of the primary somatosensory cortex. There was a tendency 

to a lower SI/SII activation compared to the activation 

in pain-processing areas in patients with hypestesia within 

the receptive fields of V2/V3. The volumetric data indicated 

a loss of nerve volume at the affected side in10 of 12 cases. 

Posters

Posters 

CONCLUSION 

BOLD fMRI is a capable tool to measure brain activation in 

chronic pain and to assess changes in the cerebral trigeminal 

pain pathway and pain processing. The ability to investigate 

functional changes in the brains of patients with trigeminal 

neuralgic pain may provide a basis for the objective longitudinal 

evaluation of chronic pain and therefore to prove the 

efficacy of analgesics and pain-relief therapy. 

REFERENCES 

1. Stippich C, et al. Neuroscience Letters 1999;277:25-28 

2. Apkarian AV. Segregation of nociceptive and non-nociceptive 

networks in the squirrel monkey somatosensory thalamus. J 

Neurophysiol 2000;84:484-494 

3. Dostrovsky JO. Role of thalamus in pain. Prog Brain Res 

2000; 129:245-257 

KEY WORDS: Trigeminal neuralgia, somatosensory cortex 

Poster 92 

Application of Diffusion Tensor MR Imaging in 

Differentiating Intracranial Tumor from Perifocal T2 

Hyperintensity 

Lin, C. · Yen, P. 

Buddhist Tzu Chi General Hospital 

Hualien, TAIWAN REPUBLIC OF CHINA 

PURPOSE 

To differentiate vasogenic peritumoral edema from tumor 

infiltration by using MR diffusion tensor imaging (DTI), 

measuring the tumor infiltration index (TII) in various 

tumors. 


MR DTI was performed in 23 patients with 25 intracranial 

tumors including 7 meningioma, 4 brain metastases, 3 

glioblastoma multiforme, 2 glioma, 2 oligodendroglioma, 2 

neuroma, 2 ependymoma, 1 suprasellar tumor, 1 PNET, and 

1 gliosis. We measured both fractional anisotropy (FA) and 

average diffusion coefficient (ADC) in tumors and peritumoral 

regions which has white matter signal abnormality in 

T2-weighted images. Tumor infiltration index was calculated 

by measuring the change in FA. 

RESULTS 

From the scatter plot distribution of these cases, we found a 

linear relationship between FA and ADC for both extraaxial 

tumors (include meningioma, neuroma, and suprasellar 

tumor) and metastases, which were regarded to have minimal 

tumor infiltration compared with glioma group. Tumor 

infiltration index was calculated from the linear regression. 

By testing the TII of perifocal edema associated with 

extraaxial tumors, metastases (mean, 0.02 ± 56) and gliomas 

(mean, 92.22 ± 13), we found a significant difference (p < 

.005) among the three. There were no significant differences 

while comparing both the ADC and FA among tumors. 

CONCLUSION 

Calculating TII of perifocal edema could be applied to distinguish 

vasogenic edema from tumor infiltration. This is 

helpful in differential diagnosis when we encountered a soli- 

336 

tary intraaxial tumor with profound perifocal edema and 

could be applied for preoperative planning ( including surgery 

and radiosurgery). 

KEY WORDS: Diffusion tensor imaging, perifocal edema, 

tumor infiltration 

Poster 93 

Line Scan Diffusion Tensor Imaging of the Brain Stem 

Wada, A. 1 · Fukuba, E. 1 · Kajitani, T. 1 · Uchida, K. 1 · Katsube, 

T. 1 · Uchida, N. 1 · Maier, S. E. 2 · Kitagaki, H. 1 

1Shimane University Faculty of Medicine, Izumo City 

Shimane, JAPAN, 2Brigham & Women’s Hospital, Boston, 

MA 

PURPOSE 

Diffusion tensor imaging with line scan method was performed 

to delineate white matter fiber tracts of the brain 

stem. 


In five healthy volunteers, diffusion tensor images covering 

brain stem were performed on a 1.5-T MR system (Signa 

Cvi: General Electric Medical Systems) with following 

parameters: b factors, 0 and 1000 s/mm 2 ; TR/TE/NEX, 

3300/52.78/1; field of view, 22 x 22 cm; effective section 

thickness/gap, 3/0 mm; and acquisition matrix, 128 x 128. 

The total imaging time was 19 minutes 46 seconds for 16 

images. Diffusion tensor postprocessing was performed on 

personal computer by using dTV (freeware developed by 

Tokyo University: http://www.ut-radiology.umin.jp/people/masutani/dTV.htm). 

Brain stem nerve tracts were defined 

from color-coding maps and reconstructed by continuous 

fiber tracking method on dTV. 

RESULTS 

The brain stem nerve tracts which were able to be identified on 

color-coding maps were as follows: at medulla oblongata level 

(Fig. A); pyramid, medial lemniscus, at pontine level (Fig. B); 

inferior cerebellar peduncle, corticospinal tract, transverse 

fibers of pons, middle cerebellar peduncle, medial lemniscus 

and central tegmental fasciculus, medial and dorsal longitudinal 

fasciculus, facial nerve, acoustic nerve, trigeminal nerve, 

superior cerebellar peduncle, corticospinal tract and corticopontine 

tract, superior cerebellar peduncle, at midbrain level; 

superior cerebellar peduncle decussation, oculomotor nerve, 

cerebral peduncle. The direction road of corticospinal tract, 

medial and dorsal longitudinal fasciculus, and medial lemniscus 

were able to be identified with fiber tracking method.

CONCLUSION 

Line scan diffusion tensor imaging provides stable image 

quality with minimal susceptibility artifacts and is useful 

method for identifying and evaluating various white matter 

fiber tracts at brain stem region. 

KEY WORDS: Brain stem, diffusion tensor imaging, line scan 

Poster 94 

A Parietal-Frontal Network Evoked by Somatosensory 

Oddball Stimuli Studied with Magnetoencephalography, 

and Its Cross-Modal Consistency 

Lee, R. R. 1 · Huang, M. X. 1 · Miller, G. A. 2 · Thoma, R. J. 3 · 

Hanlon, F. M. 3 · Harrington, D. L. 3 · Weisend, M. P. 3 · Canive, 

J. M. 3 

1University of California San Diego/Veteran’s 

Administration Medical Center, San Diego, CA, 2University of Illinois at Urbana-Champaign, Urbana, IL, 3Veteran’s Administration Medical Center/University of New Mexico, 

Albuquerque, NM 

PURPOSE 

Previous studies using fMRI and event-related potentials 

have found that a distributed parietal-frontal neuronal network 

is activated in normals during both auditory and visual 

“oddball” tasks. The common cortical regions in this network 

are inferior parietal lobule (IPL)/supramarginal gyrus 

(SMG), anterior cingulate cortex (ACC), and dorsolateral 

prefrontal cortex (DLPFC). The purpose of this study is to 

determine whether the same network is activated by oddball 

tasks using somatosensory stimuli. 


Nine right-handed normal subjects (7 male, 2 female) with 

mean age 39.3 years (SD = 13.9 years) were studied with a 

median-nerve oddball paradigm, using 0.25 ms square-wave 

pulses presented every 1000 ms to one or the other median 

nerve at the wrist, with the “frequent” side stimulated 85 

times for every 15 times of the “rare” (oddball) side. Four 

blocks of data were obtained, with 2 blocks being right-wrist 

“oddball,” and 2 blocks being left-wrist “oddball.” The 

patients were instructed to mentally count only the rare (oddball) 

stimuli. One hundred and twenty-two-channel wholehead 

magnetoencephalography (MEG) was performed to 

measure the brain’s evoked magnetic responses. An automated 

multiple dipole analysis technique, the Multi-Start 

Spatio-Temporal algorithm, localized multiple neuronal generators, 

which were superimposed on the patient’s brain MR 

images; and identified their source time-courses. The 

sources also were normalized in Talairach coordinates. 

RESULTS 

IPL/SMG, ACC, and DLPFC were localized reliably in the 

MEG median-nerve oddball responses (Fig. 1), with SMG 

activation significantly preceding ACC and DLPFC activation. 

337 

Fig. 1. Ten neuronal generators showing activation during 

the 15-500 ms interval in a representative subject’s right 

median-nerve oddball responses (rare minus frequent stimuli). 

The clusters indicate the localization uncertainty determined 

by Monte-Carlo analysis. The sources (left to right): 

left SI; midline SMA; left dPMA (dorsal premotor area); left 

SMG; right SII; left vPMA (ventral premotor area); left 

DLPFC; midline ACC; left SII; and left thalamus. (Note: 

patient’s left on reader’s left.) 

CONCLUSION 

The same parietofrontal neuronal network that shows activation 

during auditory and visual oddball tests is activated by 

a median-nerve oddball paradigm. Regions uniquely related 

to somatosensory oddball responses (e.g., primary and secondary 

somatosensory areas, and dorsal premotor/primary 

motor area) were localized also; and were activated before 

the parietofrontal network. Since this parietal-frontal network 

supports attentional allocation during performance of 

the task, this study may provide a novel method, as well as 

normative baseline data, for examining attention-related 

deficits in the somatosensory system of patients with neurologic 

or psychiatric disorders. 

KEY WORDS: Magnetoencephalography, oddball, 

somatosensory 

Poster 95 

A Clinical Role for Functional MR Imaging in 

Evaluation of Psychiatric Patients with Executive 

Impairment 

Afchani, O. · Flannery, M. · Thulborn, K. R. 

University of Illinois at Chicago 

Chicago, IL 

PURPOSE 

The potential of functional MR imaging (fMRI) in the evaluation 

of psychiatric patients with executive and cognitive 

dysfunction was evaluated by comparison to normal volunteers. 


Brain activation patterns were obtained for patients (n = 21, 

mean age 40(SD18) years, 76% male) referred for imaging 

evaluation of executive dysfunction and for normal subjects 

(n = 11, mean age 29 (SD4) years, 55% male) using fMRI 

(gradient-echo, echo-planar imaging, TR = 3000 ms, TE = 

30 ms, voxel size = 3 x 3 x 3 mm 3 ) and a visually guided saccade 

oculomotor paradigm at 3.0 T. The paradigm was a 

block design of unpredictable visually guided saccades 

(VGS) that alternated blocks (30s) of eye movement to a tar- 

Posters

Posters 

get moving randomly between 5 locations along the horizontal 

meridian (2 Hz) with central fixation across 6.5 cycles 

starting and ending with fixation. Head motion was less than 

0.5 voxel and statistical thresholds were comparable for all 

cases. Activation was measured in left and right frontal, parietal, 

temporal and occipital lobes and percentage fraction of 

total activation was calculated. Comparisons between patterns 

in patients and normal subjects were performed with tstatistics. 

Perfusion MR imaging with dynamic susceptibility 

contrast bolus tracking (gradient-echo, echo-planar imaging, 

TR = 1000 ms, TE = 30 ms, voxel size = 3 x 3 x 3 mm 3 ) 

was performed to exclude vascular disease. Maps of relative 

cerebral blood volume and tissue transit time were calculated 

by a gamma variate fitting routine. Conventional MR 

imaging using T1- and T2-weighted pulse sequences was 

evaluated for anatomical abnormalities by two neuroradiologists. 

RESULTS 

No significant global or regional perfusion or anatomical 

abnormalities were found in any of these patients. Four distinct 

brain activation patterns were found in the patients that 

differed from normal subjects as shown in the table. Group 

A showed reduced parietal activation whereas Group B 

showed reduced activation in both frontal and parietal lobes. 

Group C showed only reduced frontal activation. Only group 

D shown activation that increased in frontal but decreased in 

occipital lobes. Significant results (* p < 0.01, ** p < 0.05) 

are given in the distribution patterns of percent activation 

[mean(standard deviation)] of all activation in Table 1. 

Table: Activation Distribution Patterns for Patients and 

Normal Controls 

Group n Frontal Parietal Temporal Occipital 

Lobes Lobes Lobes Lobes 

Normals 11 13(6) 20(7) 6(2) 61(10) 

A 9 15(4) 8(4)** 6(2) 71(5) 

B 5 5(2)* 4(2)** 4(2) 88(2)** 

C 3 3(1)* 17(2) 5(5) 75(5) 

D 4 29(11)** 20(3) 4(2) 46(7)* 

CONCLUSION 

Even in the absence of anatomical and physiologic abnormalities, 

psychiatric patients with executive dysfunction 

show abnormalities in functional maps compared to normal 

subjects. As cognitive behavior must have a biological basis 

in brain function, it is reassuring that fMRI performed with 

a simple paradigm on a clinical service is sufficiently sensitive 

to reflect such different behavior. The heterogeneity of 

the functional maps may be important in the evaluation of 

psychiatric patients presenting with the clinical impression 

of executive impairment for triaging to different therapies. 

KEY WORDS: fMRI, psychiatric disease, executive dysfunction 

338 

Poster 96 

“MR Tumoral Activity Mapping” in the Radiologic 

Evaluation of Glioblastoma Multiforme: A Voxel-Based 

Grading System 

Gelmez, S. · Diren, B. · Incesu, L. · Kuruoglu, E. · Türe, U. 

Ondokuz Mayis University, School of Medicine 

Samsun, TURKEY 

PURPOSE 

Pathologically, some parts of glioblastoma multiforme can 

be of high grade while the other parts may be of low grade 

based on certain histopathologic features. Conventional MR 

imaging is insufficient in predicting glioma grade by itself. 

Based on different aspects of tumoral activity, grading can be 

performed by MR spectroscopy and MR perfusion imaging. 

Both methods have been shown to correlate reliably with 

tumor grade determined histopathologically. In this study, 

comparative evaluation of these techniques in tumor grading 

was performed in voxel-to-voxel basis and in the light of 

obtained data “tumoral activity maps” were generated showing 

the malignant and benign parts of the tumor in the slice 

of interest. 


Twenty-one pathologically proven glialblastoma multiforme 

(GBM) patients were evaluated. Imaging was performed 

with a 1.5 T MR unit (Symphony; Siemens AG, Erlangen, 

GERMANY). Multivoxel 2D proton chemical shift imaging 

(1500/135 TR/TE) which uses a point-resolved spectroscopy 

sequence with a volume of interest (VOI) of 8 x 8 cm region 

placed within 16 x 16 cm field of view on a 1 cm transverse 

section was performed for each patient. After that dynamic 

contrast-enhanced T2*-weighted echo-lanar images 

(1200/54 TR/TE; FA: 45 o ) were acquired during the first pass 

of a standard dose (0.1 mmol/kg) bolus of gadolinium- 

DTPA. Slice positioning was performed carefully to include 

the slice on which spectroscopy examination was performed. 

The voxels that comprise the tumor and the peritumoral 

edema were analyzed comparatively with both techniques. 

Grading was performed on two point scale (low and high 

grade) based on relative cerebral blood volume and metabolite 

ratios [choline (Cho)/creatinine (Cre) and Cho/N-acetyl 

aspartate (NAA)] derived from perfusion and spectroscopy 

data respectively . Threshold values reported in the literature 

for relative cerebral blood volume (rCBV) and metabolite 

ratios were used to determine the grade. Voxels containing 

vessels were excluded. Tumoral activity maps were generated 

for the slice of interest by combining the data derived 

from both techniques. 

RESULTS 

A total of 1324 voxels were evaluated. Comparative analysis 

was performed on 280 voxels that comprise tumor and peritumoral 

edema. Cho/Cre and Cho/NAA elevation suggesting 

high-grade tumor were found in 182 and 186 voxels respectively. 

Both of these ratios were elevated in 160 of them. In 

214 voxels, rCBV were elevated. Kappa analysis was used to 

measure the strength of agreement for these findings 

between two tests. All kappa values were associated with 

statistically significant value of p < 0.05. The greatest 

strength of agreement was found to be moderate comparing 

the grading based on rCBV and Cho/NAA-Cho/Cre combination 

(Kappa = 0.63; p < 0.001).

CONCLUSION 

This study suggests that rCBV maps and MR spectroscopy 

are complementary techniques that may help determine the 

grade of gliomatous tumors preoperatively. Combination of 

data derived from both of these techniques can be used to 

generate tumoral activity maps which can be used to guide 

biopsies and be helpful for the surgeon to determine which 

parts of the tumor should be removed preferentially in the 

operations that primarily aim macroscopic total removal. 

KEY WORDS: Spectroscopy, perfusion, MR imaging 

Poster 97 

Decreased Fractional Anisotropy of Middle Cerebellar 

Peduncle in Crossed Cerebellar Diaschisis: Diffusion 

Tensor Imaging-Positron Emission Tomography 

Correlation Study 

Kim, J. · Lee, S. · Lee, J. · Kim, Y. · Kim, D. 

Yonsei University College of Medicine 


PURPOSE 

To evaluate the clinical usefulness of diffusion tensor MR 

imaging for the assessment of chronic stroke with crossed 

cerebellar diaschisis. 


Twenty-two patients (15 male, 7 female: mean age, 60.5 

years) with chronic stroke (13 ischemic and 9 hemorrhagic 

stroke) were evaluated by diffusion tensor MR imaging and 

18 F-fluorodeoxyglucose (FDG) positron emission tomography 

(PET). Fractional anisotropy (FA) and color-coded vector 

maps were generated. To evaluate afferent fiber systems 

to the cerebellum, the FA of the bilateral middle cerebellar 

peduncle (MCP) was measured. Changes of FA values in the 

MCP were compared against visually assessed PET results. 

RESULTS 

In patients with a chronic infarct involving more than one 

third of the unilateral hemisphere, MCP of the contra-lesional 

side showed an FA value of 0.526 ± 0.017, which was significantly 

lower than that of the ipsi-lesional side MCP 

(0.540 ± 0.013) (one tail paired t-test, p = 0.017). On FDG 

PET scan, decreased glucose metabolism was observed in 

the affected cerebellum in 19 patients. 

CONCLUSION 

Diffusion tensor imaging can visualize an altered corticocerebellar 

circuit in the case of chronic stroke with crossed 

cerebellar diaschisis, which is hardly demonstrated by conventional 

MR images. 

KEY WORDS: Diaschisis, diffusion tensor imaging, stroke 

339 

Poster 98 

Measurement of Sensorimotor Cortical Function following 

Constraint-Induced Therapy in a Child with 

Hemiplegic Cerebral Palsy Using Magnetoencephalography 

and Functional MR Imaging 

Gaetz, W. C. 1 · Sutcliffe, T. L. 1 · Logan, W. J. 1 · Shroff, M. 1 · 

Fehlings, D. L. 2 · Cheyne, D. 1 

1Hospital for Sick Children, Toronto, ON, CANADA, 

2Bloorview MacMillan Children’s Center, Toronto, ON, 

CANADA 

PURPOSE 

To determine if constraint-induced therapy (CIT) in children 

with hemiplegic cerebral palsy results in functional improvement 

and changes in the cortical representation of the affected 

hand as measured by functional MR imaging (fMRI) and 

magnetoencephalography (MEG). 


Constraint-induced therapy has been successful in improving 

motor skills in children with hemiplegic cerebral palsy; 

however, no prior studies have demonstrated cortical reorganization 

in children. A case study with a prospective cohort 

AB design with the subject acting as their own control was 

completed. The individual was 8 years of age with a diagnosis 

of right-sided hemiplegic cerebral palsy. Constraintinduced 

therapy included casting of the unaffected arm and 

hand for 3 weeks with weekly occupational therapy. 

Constraint-induced therapy took place between week 3 and 

week 6. Clinical measures (outlined in results) and neuroimaging 

were completed at baseline, week 3, and week 6. 

Further clinical measurements were completed at week 10. 

The motor activation task completed during fMRI and MEG 

was finger extension and standardized for all neuroimaging 

assessments. Functional MR imaging was performed with a 

GE 1.5 T LX MR imaging system equipped with high-speed 

gradient. The motor task was performed in a block design 

using rest as control. Magnetoenchephalography data following 

finger extension was collected using a 151 channel 

whole-head MEG system (Omega, CTF Systems Inc.) for a 

period of 2.5 seconds duration (1.5 seconds premovement). 

Magnetoencephalography analysis consisted of electromyogram 

event-related analysis of source strength using the synthetic 

aperture magnetometry (SAM) beamformer algorithm. 

RESULTS 

Constraint-induced therapy resulted in clinically significant 

improvement as measured by the Quality of Upper 

Extremity Skills Test (pretreatment scores 64.99, 69.16; 

posttreatment score 77.81), the Pediatric Motor Activity Log 

(pretreatment frequency of use 1.0, 0.6, quality 0.8, 0.5; 

posttreatment frequency of use 4.7, quality 4.2), and the 

Canadian Occupational Performance Measure goals of zipper 

fastening and shoelace tying (pretreatment zipper performance 

5, shoelace performance 5; posttreatment zipper 

performance 10, shoelace performance 6). There was no significant 

change with the Pediatric Evaluation of Disability 

(PEDI) functional skills domain (pretreatment scores 71, 67; 

posttreatment 68) and PEDI caregiver assistance domain 

(pretreatment scores 39, 39; posttreatment 38). Functional 

MR imaging and MEG demonstrated increased activation of 

the motor cortex in the contralateral (affected) hemisphere 

postconstraint therapy. 

Posters

Posters 

CONCLUSION 

Constraint-induced therapy resulted in clinically significant 

improvement in motor function of the involved hand and 

increased contralateral activation on fMRI and MEG. Four 

additional subjects have been recruited into the study. 

KEY WORDS: Magnetoencephalography (MEG), functional 

MR imaging, cerebral palsy 

Poster 99 

Atypical Language Lateralization in Children with 

Epilepsy: A Functional MR Imaging Study 

Donner, E. J. · Evans, J. W. · Malone, S. A. · Taylor, M. J. · 

Logan, W. J. 



PURPOSE 

Functional MR imaging (fMRI) is a noninvasive technique 

used for language lateralization in children being considered 

for epilepsy surgery. Medically refractory epilepsy, characterized 

by frequent epileptic seizures and persistent interictal 

epileptiform discharges as demonstrated by electroencephalography 

(EEG) may be associated with functional 

and/or structural CNS lesions. Atypical language patterns, 

including mixed hemispheric dominance and right hemisphere 

dominance in right-handed individuals, may develop 

more readily in children with medically refractory epilepsy. 


Functional MR imaging for language lateralization was performed 

in 14 right-handed children, aged 7-16 years, 9 

female. All children were diagnosed with medically refractory 

epilepsy with right or left, frontal or temporal epileptic 

foci documented by video-electroencephalography and were 

being considered for epilepsy surgery. Standardized language 

paradigms, adjusted for age and developmental level, 

were performed and included verb generation, letter fluency, 

picture naming, and word repeat. 

RESULTS 

Ten of the 14 children demonstrated atypical language patterns 

on fMRI testing. Atypical patterns included mixed 

hemisphere dominance characterized by frontal lobe activations 

contralateral to temporal lobe activations and right 

hemisphere language dominance. Seven of the eight children 

with left hemisphere epileptic foci demonstrated atypical 

language patterns, while three of the six children with right 

hemisphere epileptic foci also demonstrated atypical language 

patterns. 

CONCLUSION 

We found that patterns of atypical language activation are 

demonstrated frequently by fMRI in children with medically 

refractory epilepsy. Notably, this phenomenon is not confined 

to children with epileptic foci affecting the dominant 

(left) hemisphere. Atypical language dominance in children 

with medically refractory epilepsy may reflect neuroplasticity 

and relocalization of eloquent cortex due to frequent 

seizures. Atypical language activation may also reflect 

340 

developmental aspects of language lateralization in children. 

Further fMRI studies investigating the developmental trajectory 

of language dominance in children are needed. 

KEY WORDS: Functional MR imaging, epilepsy, language 

Poster 100 

Improved Conspicuity of the Optic Radiations in the 

Newborn Brain Using Combined Diffusion 

Tensor/Anatomical MR Tractography 

Nielsen, J. F. · Panigrahy, A. · Chen, V. J. · Nelson, M. D. · 

Erberich, S. G. 

Children’s Hospital Los Angeles 


PURPOSE 

Diffusion-tensor imaging (DTI) is sensitive to subtle white 

matter (WM) injury in newborns and enables novel investigations 

of structure-function relationships in the developing 

brain. Diffusion-tensor imaging data is complex, and in common 

clinical practice only the direction of maximum diffusion 

(principal diffusion direction) (PDD) is displayed, using 

2D color-coded fractional anisotropy (FA) maps (Fig. a). 

However, 1) PDD noise in the newborn subcortex makes it 

difficult to “trust” the directional information in such low FA 

regions, 2) the PDD contains no reliable information about 

the connectivity between voxels, and 3) spurious directional 

information is displayed inside CSF and gray matter (GM). 

Here we demonstrate improved conspicuity of the optic radiations 

in newborns using enhanced color-coded directional 

maps that overcome these limitations. 


IRB-approved DTI (EPI; 1.7 x 1.7 x 3 mm3; TE/TR = 

83/10000 msec, b = 700 s/mm2) was performed in 10 fullterm 

newborns. Patients were sedated and placed in an MRcompatible 

incubator with neonatal headcoils. Our method is 

based on novel multimodality probabilistic (Monte Carlo) 

tractography, which combines coregistered DTI and anatomical 

(T2-weighted) MR image volumes such that fiber generation 

into GM and CSF is suppressed, and is designed to 

be robust with respect to noise in the DTI data. Tractography 

was performed from multiple starting points distributed uniformly 

throughout the brain volume, and results were visualized 

using 2D color-coded maps with the color and brightness 

of each voxel based on the orientation and density, 

respectively, of the fibers passing through the voxel. The 

directional maps were evaluated by two pediatric neuroradiologists. 

RESULTS 

Tractography-based directional maps (Fig. b) were found to 

be highly effective in suppressing spurious directional anatomy 

inside CSF. In addition, tractography-based maps 

appeared more uniform (less “speckled”) in the newborn 

subcortex, which correctly reflects the increased uncertainty 

in the DTI measurements in these low FA regions. Longrange 

connectivity could be assessed by the relative intensities 

of the various WM tracts. Clinical evaluation showed 

improved conspicuity of the optic radiations, largely independent 

of DTI noise level.

CONCLUSION 

Relevant information from DTI and anatomical MR imaging 

can be combined to produce tractography-based color-coded 

directional maps that increase the conspicuity of the optic 

radiations in newborns by suppressing CSF/GM artifacts and 

enhancing highly connected WM structures. This postprocessing 

method improves the clinical utility of DTI in newborns, 

and provides a basis for structure-function analysis 

using combined functional MR imaging and DTI of the 

neonatal visual system. 

KEY WORDS: Diffusion tensor imaging, tractography, optic 

radiation 

Poster 101 

Diffusion Tensor Imaging Evaluation of the Abnormal 

Fiber Connections in Callosal Dysgenesis and 

Holoprosencephaly: Initial Experience 

Benardi, B. · Makki, M. · Thomas, E. J. 

Children’s Hospital of Michigan 

Detroit, MI 

PURPOSE 

The corpus callosum (CC), is the major white matter (WM) 

tract connecting cerebral hemispheres. In case of severe 

defect, the developing CC axons, soon after reaching the 

midline and finding no bridge to cross begin to form the 

Probst’s bundles. Diffusion tensor imaging color map shows 

WM fiber orientation in vivo and it is capable of tracking 

fibers, thus is useful in understanding of normal interhemispheric 

connections as well as the evaluation of the abnormal 

connections in callosal dysgenesis (CD). Holoprosencephaly 

is a complex brain malformation resulting from incomplete 

cleavage of the prosencephalon. A form of agenesis of the 

CC is a part of the holoprosencephalic spectrum differing 

from the isolated CD. The purpose of this investigation was 

to study in vivo the abnormal interhemispheric connectivity 

associated with different types of CD and the abnormal connectivity 

in holoprosencephaly. 


Three CD patients and two cases of lobar and semi-lobar 

holoprosencephaly were studied. Forty axial diffusion tensor 

echo-planar images were acquired with 6 noncollinear directions. 

Apparent diffusion coefficient (ADC) and fractional 

anisotropy (FA) values were measured. White matter fiber 

tracking was performed based on fiber assignment by continuous 

tracking algorithm. Source and target ROIs were 

fixed with FA threshold (0.2) and angulation value (60º) to 

start and end the connectivity. Regions of interest were 

341 

drawn on diffusion-weighted imaging, ADC, and FA maps 

over hemispheric/interhemispheric structures: superiorfrontal 

gyrus, temporal stem, superior parietal lobule, thalamus, 

external/internal capsule, PB, abnormal cingulum, and 

anterior commissure. In holoprosencephaly cases frontal 

abnormal interhemispheric WM, and residual posterior CC 

also were evaluated. 

RESULTS 

The limited number of patients and normal controls does not 

allow any statistical comparison of the ADC and FA values. 

However, diffusion tensor imaging provided promising 

information on the aberrant hemispheric fiber connections. 

Different patterns of PB are associated to different types of 

CD that can be investigated with tractography. In holoprosencephaly 

cases, almost normal CC was found posterior and 

continuous with the WM joining the two cerebral hemispheres. 

This resultant CC was not associated with normal 

anterior commissure or fornices. 

CONCLUSION 

Diffusion tensor imaging shows different PB extensions and 

topography in different types of CD. The time of the developmental 

impairment produces different fiber arrangements 

and different associated interhemispheric anomalies. 

Diffusion tensor imaging may contribute to the comprehension 

of the topographic representation in the PB and to confirm 

if the problem inducing CD is in some of the substrates 

of axonal guidance at the midsagittal plane and not in the 

cells where callosal axons originate or in the growth cones 

themselves. Dislocation or other fiber track anomalies due to 

associated cortical dysplasia may be assessed. Preliminary 

data suggest that diffusion tensor imaging may be of value 

for a better comprehension of CD and a promising method to 

analyze connectivity in holoprosencephaly with its variability. 

Longitudinal studies may display developing of the connectivity 

and WM organization that could be associated to 

different genetic program defects. 

KEY WORDS: Diffusion tensor imaging, callosal dysgenesis, 

tractography 

Poster 102 

Clinical Value of Earlier Detection of Disease Progression 

in Children with Diffuse Pontine Tumors by MR 

Spectroscopy 

Karimi, S. · Thakur, S. · Dunkel, I. · Holodny, A. I. · Huang, 

W. 


New York, NY 

PURPOSE 

To determine if MR spectroscopy measurements can provide 

information for earlier detection of disease progression in 

patients with diffuse pontine tumors (DPTs), and consequently, 

if MR spectroscopy can play an important role in 

their clinical management. 


Two pediatric patients (7-year-old male and 10-year-old 

female) with DPT underwent MR/MR spectroscopic imaging 

of the brain/brain stem. The patients had MR/MR spectro- 

Posters

Posters 

scopic examinations prior to treatment (baseline), 5.5 weeks 

(follow-up 1), and 10 weeks (follow-up 2) after completion of 

radiation treatment on a 1.5 T scanner. The MR spectroscopic 

imaging data were acquired using a PRESS sequence with TE 

= 144 ms, processed using GE's Functool software and overlaid 

onto thin FLAIR images. Peak area ratios of Cho/NAA 

and Cho/Cr were analyzed at baseline, follow-up 1 and follow-up 

2 for the voxels located in the same anatomical locations. 

Our home-developed image processing software was 

used to measure contrast-enhancing volume. 

RESULTS 

Figures 1 and 2 depict changes in Cho/NAA, Cho/Cr and 

contrast-enhancing volume over time for the two patients, 

respectively. There was no contrast enhancement in both 

patients at the baseline. Patient 1 presented only with 

headaches, which clinically suggested a less aggressive 

tumor. Patient 2 presented with multiple cranial neuropathies. 

However, baseline ratios of Cho/NAA and Cho/Cr 

in both lesions were elevated significantly, suggesting 

aggressive tumors. Following treatment, both patients 

showed excellent clinical improvement at follow-up 1; however, 

Cho/Cr and Cho/NAA continued to rise in both lesions 

at follow-ups 1 and 2 indicating progression. Both patients 

showed clinical deteriorations at follow-up 2 and patient 2 

died shortly after follow-up 2. MR imaging contrast 

enhancement appeared at follow-up 1 and increased at follow-up 

2 for both patients. Since radiation necrosis cannot 

be differentiated readily from tumor progression based on 

contrast enhancement, the MR spectroscopic data were critical 

in showing disease progression and predicting clinical 

deterioration particularly at follow-up 1 when both patients' 

symptoms had improved dramatically. 

CONCLUSION 

Although patient number is limited, this longitudinal study 

suggests that compared to conventional postcontrast MR 

imaging and clinical symptomatology, MR spectroscopy is a 

better indicator of the presence of aggressive disease at presentation 

and in the early detection of disease progression in 

patients with DPT. MR spectroscopic imaging may be of 

critical importance in clinical management of these patients. 

Earlier detection of progression allows patients to be more 

appropriate candidates for investigational therapies which 

may lead to an improved overall status and longer life 

expectancy. 

KEY WORDS: Diffuse pontine tumors, pontine glioma, MR 

spectroscopy 

342 

Head and Neck 

103-120 

Poster 103 

Direct Quantification of Brain Motion 

Soellinger, M. · Rutz, A. · Kozerke, S. · Boesiger, P. 

ETH Zurich and University Zurich 

Zurich, SWITZERLAND 

PURPOSE 

Several diseases of the central nervous system, like hydrocephalus, 

are associated with abnormal cerebrospinal fluid 

(CSF) dynamics in the brain. Cerebrospinal fluid pulsation is 

mediated by brain motion which in turn is caused by pulsation 

of blood in the cerebral arteries (1, 2). In order to 

explore brain motion, several studies using MR velocity 

measurements have been presented (3). Tagging techniques 

have been employed also to measure brain tissue displacement. 

The CSPAMM method (4) has been used previously in 

conjunction with the HARP algorithm (5). It is the objective 

of the current work to present results of a highly motion-sensitive 

method with sufficient spatial resolution for brain 

motion quantification using complementary displacement 

encoding with stimulated echoes (C-DENSE). 


Complementary displacement encoding with stimulated 

echoes was implemented based on a CSPAMM sequence 

with additional motion decoding gradients (6). The motion 

encoding/decoding gradients were applied in measurement 

direction. The measurement direction was rotated in consecutive 

experiments to encode motion along all three spatial 

axes. The encoding sensitivity was chosen region-dependent 

ranging from 0.5 to 0.1 mm/cycle. Remaining scan parameters 

were: temporal resolution: 70 ms, in-plane resolution: 1- 

2 mm 2, slice thickness: 4 mm. Time-invariant phase errors 

due to eddy-currents were eliminated by subtraction of the 

first time frame from the following ones. The scan time was 

2 to 4 minutes. Five healthy volunteers were examined on a 

Philips 1.5 T system (Philips Medical Systems, Best, The 

Netherlands). 

RESULTS 

Displacement maps (Fig. 1) could be acquired in all volunteers 

over the whole cardiac cycle and encoded in three 

directions. The results are in good agreement with the brain 

motion velocity data found in ref. 2 with maximal caudal 

motion in the brain stem (e.g., 0.18 mm). During systole, significant 

motion in medial direction could be observed in the 

thalamus as well as in the caudate nucleus (e.g., 0.09 mm). 

Motion was strongly reduced in the corpus callosum (Fig. 1).

CONCLUSION 

A highly motion-sensitive method has been implemented for 

studying brain motion with sufficient spatial resolution. Very 

small motion in the brain hemispheres can be observed and 

further knowledge of brain dynamics can be derived. 

REFERENCES 

1. Alperin N, et al. Magn Reson Med 1996;35(5):741-754 

2. Greitz D, et al. Neuroradiology 1992;34(5):370-380 

3. Feinberg DA. Radiology 1992;185(3):630-632 

4. Fischer SE, et al. Magn Reson Med 1993;30(2):191-200 

5. Osman NF, et al. Magn Reson Med 1999;42(6):1048-1060 

6. Gilson WD, et al. Magn Reson Med 2004;51(4):744-752 

KEY WORDS: MR imaging, hydrocephalus, brain motion 

Poster 104 

Influence of 1.5 and 3 T MR Imaging on Lid Implants 

Bauknecht, H. · Schrom, T. · Berghaus, A. · Scherer, H. · 

Klingebiel, R. 

Charité Berlin 


PURPOSE 

Due to its high resolution, MR imaging has been used 

increasingly as a diagnostic procedure, particularly in tumors 

of the head and neck. Next to the 1.5 T systems 3.0 T systems 

have gained in frequency and importance. For therapy 

of lagophthalmos in facial palsy gold (99.99%), platinum 

(99.95%) and platinum (97%)/iridium (3%)-alloy weights 

will be implanted into the upper eyelid. Investigations about 

the MR imaging compatibility of the different eyelid 

implants with 1.5 and 3.0 T-MRT-systems are still missing. 


Three different eyelid implants made of pure gold (99.99%), 

pure platinum (99.95%) and a platinum (97%)/iridium (3%) 

alloy were tested in vitro. As parameter temperature changes, 

movements of the implants and picture artifacts of the different 

implants were determined at a 1.5 and 3.0 T MR system. 

The judgement of the deflection forces of the implants 

were evaluated by thread test, lying in a petri dish, and floating 

in a water bath. 

RESULTS 

No rise in temperature was recorded in any implant while 

testing with the 1.5 and 3.0 T MR system. On the water surface 

all three implants changed their positions and floated 

towards the magnetic field. This effect increased in the 3.0 T 

system. An increase in the frictional resistance of the 

implants showed no movement in the 1.5 and 3.0 T system. 

Also in the thread test no movement of the implants could be 

observed. All implants made little artifacts near which gold 

showed the least artifacts. 

CONCLUSION 

MR imaging at 1.5 or 3.0 T may be performed safely in 

patients with eyelid implants made of pure gold (99.99%), 

pure platinum (99.95%) and a platinum (97%)/iridium (3%) 

alloy with respect to warming up or dislocation of the 

implant. 

KEY WORDS: MR imaging, lid implant, iridium 

343 

Poster 105 

Clinical Application of 3D CT Angiography Using 

Multidetector Row CT 

Endo, Y. · Matsumoto, M. · Sakuma, J. · Suzuki, K. · Sasaki, 

T. · Kodama, N. 

Fukushima Medical University 

Fukushima, JAPAN 

PURPOSE 

We have developed several 3D images using multidetector 

row CT scanner. We evaluated the usefulness of the following 

3D CT images (cranio-cervical 3D CTA, 3D brain image, 

ECG gated 3D CTA and the separate demonstration of arterial 

phase and venous phase on 3D CTA). 


A multidetector CT scanner with 16 detectors was used. 1) 

Cranio-cervical 3D CTA: The continuous scanning from the 

upper edge of sternum to the vertex was performed with an 

injection of contrast material at a speed of 3 ml/sec (a total 

volume of 100 ml), with a 5 mm slice thickness. 2) 3D brain 

image: A subtraction method was used to extract brain from 

the 3D CT image. The bone image was increased in thickness 

by 6-8 pixel layers by using the "dilate" function of the 

workstation. The "dilate" function adds a specified number 

of pixel layers of original data to an image. The dilated bone 

image (mask) was subtracted from the 3D CT image. 3) ECG 

gated 3D CTA: In the patients with cerebral aneurysm, ECG 

gated 3D CTA was performed. After acquisition of helical 

data set, data for image reconstruction are selected every 

10% of the R-R interval of the ECG signal and the cine 

image of the aneurysm was developed. 4) Separate demonstration 

of arterial phase and venous phase on 3D CTA: First, 

dynamic CT scanning was performed to obtain the optimal 

scan timings for arterial phase and venous phase. Based on 

the results of the dynamic CT scan, the scannings for arterial 

phase and venous phase were performed for 5 seconds 

after a bolus injection of contrast medium at a rate of 6-7 

ml/sec (total 30-35 ml) for 5 seconds followed by a chaser 

bolus of 18-21 ml of saline at 6-7 ml/sec. 

RESULTS 

1) 3D CTA of the whole course from the origin of the internal 

carotid and vertebral arteries to their intracranial part was 

obtained with a single scan. Cranio-cervical 3D CTA 

allowed us to diagnose multiple lesions in the cervical and 

intracranial arteries. 2) 3D brain image demonstrated major 

sulci and gyri, providing us the useful anatomical information 

for the surgery. 3) The cine image of ECG gated 3D 

CTA showed the movement of cerebral aneurysm with cardiac 

cycle. However, it was difficult to evaluate the thickness 

or fragility of aneurysmal wall because of artifact due to helical 

scan. Moreover, there were discrepancies between the 

movement on cine image and the intraoperative findings. 4) 

Our method made it possible to develop the separate demonstration 

of arterial phase and venous phase on 3D CTA in a 

single procedure. The fused 3D CT images of arterial phase 

and venous phase facilitated the understanding of the 

anatomical relationship among the lesion, the arteries, veins, 

and bony structures, allowing a safer surgical approach. 

Posters

Posters 

CONCLUSION 

Although ECG gated 3D CTA needs a further improvement 

for clinical application, multidetector CT supplies several 

useful 3D images. Multidetector CT will be an essential 

modality for neurosurgical field. The indication of conventional 

cerebral angiography for central nervous system disorders 

will be limited. 

KEY WORDS: Multidetector CT, 3D CT angiography, 3D 

image 

Poster 106 

Practical Use of MR Investigation of the 

Temporomandibular Joint 

Peterova, V. · Meszarosova, M. · Fikackova, H. · Mazanek, J. 

Charles University 

Prague, CZECH REPUBLIC 

PURPOSE 

The temporomandibular joint (TMJ) dysfunction might be 

cause of atypical facial pain, which can be treated by 50 Hz 

stimulation of masseter muscles, knowing the exact state of 

TMJ MR imaging. We evaluated the benefit of MR imaging 

of TMJ with specific measurements of joint movement with 

the aim to compare the objective state before and after treatment 

of TMJ disorder. 


We prospectively examined both TMJ in five healthy persons 

before and after stimulation of masseter muscles. All 

individuals were investigated on 1.5 T MR scanner in mode 

turbo spin echo (TSE) during closed mouth and its opening 

in head coil; the four semipositions were fixed by plastic 

cylinders of fixed size. We measured specific distances of 

the joint and angles: angle α denotes medial slope of the 

condyle, angle β expresses external rotation of the condyle 

and angle γ evaluates craniocaudal slope of condyle. Further 

we measured the total rotation of mandible head, the craniocaudal, 

ventrodorsal, and mediolateral displacement of 

condyle, distance of ventral surface of condyle from the 

external auditory meatus, distance of ventral meniscal surface 

from ventral condylar surface, and the total latitude 

(width) of articular cleft. 

RESULTS 

Evoked spasm of masseter caused deviation to the stimulation 

side in closed-mouth mandible, dorsolateral shift and 

rotation of mandible head on the side of stimulation and 

symmetric reverse finding on the oposite side; the position of 

the disk did not change. In dynamic study evoked spasm of 

masseter brings the deviation of mandible to the stimulation 

side and restriction of the maximal extent of mouth opening, 

the disk kinematics was not influenced. 

CONCLUSION 

Dynamic MR investigation of TMJ in T2-weighted images 

in TSE mode brought as the best imaging protocol the evaluation 

of the shape, movement and displacement of the joint 

components during closed and opened mouth. The determination 

of the total rotation and mediolateral displacement of 

344 

mandible head provides the greatest benefit in the measurement 

of condyle movement and should be done before treatment 

by unilateral stimulation of masseter muscle. 

KEY WORDS: Temporomandibular joint, MR imaging, measurements 

Poster 107 

3D Volumetric Imaging of the Labyrinth 

Lee, B. C. P. 1,2 · Clary, R. 1,2 

1Saint Louis Children’s Hospital, St. Louis, MO, 

2Washington University Medical School, St. Louis, MO 

PURPOSE 

To evaluate the role of high-resolution 3 and 1.5 T 3D MR 

imaging and 3D volume rendering in evaluation of sensorineural 

deafness in children. 


Thirty-two children age 6 months to 11 years with sensorineural 

deafness referred for potential cochlear implants 

were studied on Siemens 1.5 Sonata (22 cases) and 3 T Trio 

scanners (10 cases). The resolution for CISS was 0.5 x 0.5 x 

0.5 mm 3 (FOV150, 48 slices, thickness 0.5 mm, 512 matrix). 

An additional 3D T2 sequence was performed on the 3 T 

scanner: resolution 0.3 x 0.3 x 0.6 mm 3 (FOV170, 48 slices, 

thickness 0.3 mm, 320 matrix, 5 averages). Scan times were 

4.44 and 15.42 min respectively. 1.5 mm thick T2 oblique 

sagittal scans of the internal auditory canals and routine 4 

mm T2 axial scans of the entire brain and 2 mm T2 coronal 

scans also were performed. All patients were sedated and 

contrast was not administered. Three-dimensional volumerendered 

reconstructions were performed in all cases using a 

Voxar workstation. The precise configuration of the cochlear 

and semicircular canals were evaluated. Automatic volumetric 

measurement of enlarged endolymphatic duct and sacs 

were performed using a simple thresholding technique. 

RESULTS 

The endolymphatic duct and sacs were enlarged in 8 cases 

with volumes between 1.3-2.5 mm 3 (Fig. A). Absence of the 

semicircular canals was detected in 3 cases (Fig. B). The 

basement membrane of the cochlea was seen in all cases, and 

the precise turns of the cochlear was evaluated better by freehand 

rotation/translation of the 3D volumetric data than on 

individual 2D images. The facial and all divisions of the 

eight nerve were visualized in all the cases. The oblique 

reformatted views were similar in quality to the direct T2 

oblique views. All reconstructions, including mpr and 3D 

renderings were better with the 3D T2 technique.

CONCLUSION 

Although images from 1.5 T scanner is used currently for 

evaluation of sensorineural deafness, images from 3 T scanner 

is superior. We believe a comprehensive evaluation of 

the labyrinth is possible in under 16 minutes on the 3 T scanner 

using a 3D T2 technique. These images have better resolution 

than 3D CISS. 

KEY WORDS: Inner ear, MR imaging, 3D 

Poster 108 

18F-Fluoro-2-Deoxyglucose Positron Emission 

Tomography and CT/MR Imaging in Oral Cavity 

Carcinoma: Results of a Prospective Study of 101 

Patients 

Ng, S. H. · Wong, A. M. · Yen, T. C. · Liao, C. T. 

Chang Gung Memorial Hospital 

Taiwan, TAIWAN REPUBLIC OF CHINA 

PURPOSE 

To prospectively evaluate the use of 18 F-fluorodeoxyglucose 

positron emission tomography (FDG PET), CT/MR imaging, 

and their combination in the initial staging of squamous 

cell carcinoma (SCC) of the oral cavity with histologic correlation. 


A total 101 patients with oral cavity SCC planned for dissection 

were recruited into this study. For correlative analysis 

of nodal staging of FDG PET, CT/MR imaging, and the 

histopathologic examination, the neck was divided into levels 

based on the imaging-based nodal classification. Three 

nuclear medicine physicians and two head and neck radiologists 

interpreted FDG PET and CT/MR imaging separately 

in a blinded fashion without knowledge of each others findings, 

and then they joined to read FDG PET and CT/MR 

imaging together. A total of 402 neck levels were dissected. 

Histopathologic analysis was used as the gold standard for 

assessment of the sensitivity and specificity of these modalities. 

We used McNemar test to compare the results, and used 

the receiver operating characteristic curve (ROC) method 

with area under the curve (AUC) calculation to evaluate their 

discriminative power. 

345 

RESULTS 

Of our 101 patients, FDG PET correctly revealed the primary 

tumor in 100 and CT/MR imaging correctly revealed 

87. In 402 neck levels resected in our 101 patients, 80 neck 

levels contained metastatic neck disease in a total of 44 

patients. On a level-by-level basis, the sensitivity of FDG 

PET was 32.8 % higher than that of CT/MR imaging (80% 

vs 53.8% P < 0.005) while the specificity of FDG PET was 

2% lower than that of CT/MR imaging (91.6% vs 93.5%, P 

= 0.327). The AUC obtained from ROC curve showed that 

FDG PET was significantly superior to CT/MR imaging for 

total nodal detection (0.91 vs 0.79, P = 0.0026). On the other 

hand, the sensitivity and specificity of the combined use of 

FDG PET and CT/MR imaging showed 1.6% and 3.4%, 

respectively, higher than those of FDG PET alone (81.3% vs 

80%, P = 0.99; 95.0% vs 91.6%, P = 0.003, respectively). 

The AUC obtained from ROC curve showed that FDG PET 

was significantly superior to CT/MR imaging for total nodal 

detection (0.911 vs 0.794, P = 0.0026) while the combined 

use of FDG PET and CT/MR imaging was slightly superior 

to FDG PET alone without statistical significance (0.913 vs 

0.911, P = 0.1034). 

CONCLUSION 

In the initial nodal staging of oral cavity SCC, FDG PET is 

superior to CT/MR imaging because the former had a significantly 

higher sensitivity. The combined use of FDG PET 

and CT/MR imaging shows little increment of sensitivity but 

significant increment of specificity than PET alone. 

KEY WORDS: Oral cavity cancer, positron emission tomography, 

lymph node metastases 

Poster 109 

Effect of Radiation Therapy on the Prevalence of Carotid 

Disease in 15- to 30-Year-Old Patients with Hodgkin’s 

Disease 

Basiratnia, R. · Hekmatnia, A. · Nouri-Mahdavi, K. · 

Pourmoghaddas, A. 

Ghadamgahi, M. 

· Hamedi, M. · Rezaei, M. · 



PURPOSE 

Hodgkin’s disease is a relatively common malignancy for 

which external beam radiation therapy is often used. 

Potential vascular complications of radiation therapy include 

intimal thickening, medial hyalinization, and cellular infiltration 

in adventitia with plaque formation, fibrosis, and 

thrombus formation in small vessels. Early detection of 

asymptomatic carotid stenosis may prevent strokes. We studied 

the prevalence of carotid stenosis in irradiated patients 

with Hodgkin’s disease of the cervical lymph nodes as compared 

with a control group. 


There were 50 patients (mean age 26.3 years) with a history 

of Hodgkin’s disease of the cervical lymph nodes who had 

undergone radiotherapy (3500-4000 cGy) 5-15 years earlier 

and at the time of this survey were < 30 years old. The control 

group consisted of 50 patients (mean age 22.6 years) 

who had been hospitalized for other reasons and had no his- 

Posters

Posters 

tory of radiation. Patients with diabetes or hyperlipidemia 

and smokers were excluded. The carotid arteries were 

scanned with a 7.5 MHz ultrasound probe in three areas: the 

supraclavicular area (point A), the carotid bifurcation (point 

B), and the proximal part of the internal carotid artery (point 

C). Intima-media thickness (IMT) of the patients in the far 

point of these three areas on the longitudinal view was compared 

with that of the control group. 

RESULTS 

Mean IMT in the patients was 0.58 mm, 0.75 mm, and 0.68 

mm at points A, B, and C, respectively. Mean IMT in the 

control group was 0.44 mm, 0.58 mm, and 0.51 mm at points 

A, B, and C, respectively. Mean IMT at each point was significantly 

higher in the patients than in the control group (p 

< 0.05). 

CONCLUSION 

Radiation therapy of patients with Hodgkin’s disease of the 

cervical lymph nodes is a significant predisposing factor for 

early atherosclerosis of the carotid arteries. 

KEY WORDS: Hodgkin’s disease, radiation therapy, atherosclerosis 

Poster 110 

Effect of Radiation Therapy on the Prevalence of Carotid 

Disease in 30- to 60-Year-Old Patients with Head and 

Neck Malignancy 

Basiratnia, R. · Hekmatnia, A. · Nouri-Mahdavi, K. · Rezaei, 

M. · Pourmoghaddas, A. · Sanei, H. · Hamedi, M. · 

Ghadamgahi, M. 



PURPOSE 

Radiation therapy of the head and neck is known to be a predisposing 

factor for early atherosclerotic changes of the 

carotid arteries. The intima-media thickness (IMT) of the 

carotid arteries generally is used as an index of atherosclerotic 

involvement. We measured carotid IMT in patients with 

a history of radiation therapy for head and neck cancers with 

a control group. 


Twenty-four patients (12 men and 12 women) with a mean 

age of 41 years (range 30-60 years) had a history of radiation 

therapy for head and neck malignancy (3600-6000 cGy) 3 to 

13 years earlier. The cancers included adenoid cystic carcinoma 

of the parotid glands (6 patients), Hodgkin’s disease 

(4), laryngeal carcinoma (3), thyroid carcinoma (3), 

nasopharyngeal fibrosarcoma (2), acinar cell carcinoma of 

the parotid gland (2), pleomorphic adenoma of the submandibular 

gland (2), adenoid cystic carcinoma of the mouth 

floor (2). The patients had no other predisposing factors for 

atherosclerosis. Intima-media thickness of the carotid arteries 

was measured with a 7.5 MHz ultrasound probe at the far 

point of three areas: the midportion of the common carotid 

artery (point A), the carotid bifurcation (point B), and the 

proximal part of the internal carotid artery (point C). A com- 

346 

parison then was made with the corresponding values of a 

group of 22 age-matched and sex-matched healthy volunteers. 

RESULTS 

Mean IMT in the patients was 0.75 mm, 0.87 mm, and 0.79 mm 

at points A, B, and C, respectively. Mean IMT in the control 

group was 0.60 mm, 0.62 mm, and 0.57 mm at points A, B, and 

C, respectively. Mean IMT at each point was significantly higher 

in the patients than in the control group (p < 0.05). 

CONCLUSION 

Radiation therapy of patients with head and neck malignancy 

is a significant predisposing factor for early atherosclerosis 

of the carotid arteries. Thus, it may be advisable to monitor 

these patients with color Doppler ultrasound of the 

carotid arteries. 

KEY WORDS: Radiation therapy, head and neck malignancy, 

atherosclerosis 

Poster 111 

Parathyroid Tumors: Scintigraphy (Technetium-99m 

MIBI), US, CT, and MR Imaging with Diagnostic Value 

Park, D. 1 · Kim, S. 1 · Kim, Y. 1 · Kim, Y. 1 · Park, C. 1 · Tae, K. 1 

· Lee, S. 2 · Hahm, C. 2 

1 Hanyang University Guri Hospital, Guri, REPUBLIC OF 

KOREA, 2 Hanyang University Hospital, Seoul, REPUBLIC 

OF KOREA 

PURPOSE 

This study aims to evaluate the imaging findings and the 

diagnostic value of scintigraphy (technetium-99m MIBI), 

US, CT, and MR in the patients with primary hyperparathyroidism. 


Thirteen patients (3 males, 10 females, 37-70 years old, 

mean: 51.2 years old) with primary hyperparathyroidism 

were studied preoperatively with technetium 99m MIBI 

scintigraphy (13 patients), US (13 patients), CT (11 patients) 

and MR imaging (6 patients). They were histologically 

proven to be adenomas (11 patients), carcinoma (1 patient) 

and 2 hyperplastic glands (1 patient). We retrospectively 

analyzed the imaging findings and the diagnostic value, 

comparing with the surgical and pathologic outcomes. 

RESULTS 

All parathyroid tumors showed hot uptakes on scintigraphy, 

hypoechoic or hypodense characteristics on US and CT, and 

isointense signals on T1- and T2-weighted MR images, 

when compared to the thyroid gland. Those which had no or 

minimal Doppler color flow signals on US, demonstrated no 

or minimal enhancement on CT and MR imaging after contrast 

injection. In particular, parathyroid hyperplasia 

revealed multiple masses, while carcinoma showed an illdefined 

large mass. The overall diagnostic localization accuracy 

for scintigraphy was 92%, US 88%, MR imaging 83%, 

and CT 77%. Technetium MIBI scintigraphy had the highest 

accuracy rate for localizing the parathyroid gland tumors.

CONCLUSION 

Parathyroid tumors show similar imaging findings except for 

multiple masses found in parathyroid hyperplasia and illdefined 

large mass in carcinoma. In patients with parathyroid 

tumor, scintigraphy is the most effective localization 

technique and US is a good confirmatory localization test. 

KEY WORDS: Parathyroid 

Poster 112 

CT and MR Imaging Features of Salivary Duct 

Carcinomas: Review of 17 Consecutive Cases 

Weon, Y. · Kim, H. · Baek, C. · Son, Y. · Kim, S. · Jeong, H. 

· Ko, Y. · Byun, H. 

Samsung Medical Center 


PURPOSE 

Salivary duct carcinoma (SDC) is a rare distinctive salivary 

gland neoplasm characterized by intraductal and infiltrating 

components and the grave prognosis. The aim of this study 

was to describe CT and MR imaging features of SDC. 


CT (n = 13) and MR (n = 6) images of 17 patients (12 men 

and 5 women; mean age, 59 years; range, 44-72 years) with 

pathologically proved SDC were reviewed retrospectively. 

Particular attention was paid to the location, margin, presence 

of necrosis or calcification, invasion of the adjacent tissue, 

and lymph node metastasis. The density and the signal 

intensity of the tumor as well as the pattern of enhancement 

wererecorded also. 

RESULTS 

Thirteen lesions (76%) with one in the distal parotid duct were 

located in the parotid gland and four in the submandibular 

gland. The size of the tumor was 2.8 cm (1.2-5.0 cm) and 

greater than 3 cm in 8 cases (47%). Fourteen lesions (82%) 

were irregular in margin, while three were well marginated. 

Tumor necrosis was noted in 11 cases (65%) and calcifications 

were demonstrated in 7 cases (41%). The adjacent tissue invasion 

was shown in 10 (59%) cases and metastasis to the regional 

lymph nodes was found in 8 cases (47%). Compared with 

the adjacent muscle, the tumor appeared hypodense on precontrast 

CT scans either homogeneously (n = 6) or heterogeneously 

(n = 7). On the postcontrast CT scans, the tumor was poorly 

enhanced in 10 of 13 cases. In all six cases in which MR imaging 

was undertaken, the tumor was isointense to muscle on T1weighted 

imaging and iso to hyperintense to muscle on T2weighted 

imaging. All cases showed heterogeneous enhancement 

after gadolinium injection. 

CONCLUSION 

On CT and MR imaging, most SDCs appeared as a mass 

with irregular margins, most commonly arising from the 

parotid gland. Necrosis, calcifications, extraglandular extension, 

and regional lymph node metastasis were frequently 

associated findings. Knowledge of the CT and MR imaging 

findings of SDC may help build up the understandings of 

this highly aggressive tumor of the salivary gland. 

KEY WORDS: Salivary duct carcinoma, CT, MR imaging 

347 

Poster 113 

Hyperintense Cystic or Necrotic Cervical Lymph Node 

on T1-Weighted MR Image: Characteristic Findings of 

Nodal Metastasis of Thyroid Carcinoma 

Park, D. 1 · Kim, S. 1 · Kim, Y. 1 · Kim, Y. 1 · Park, C. 1 · Lee, S. 2 

· Hahm, C. 2 

1Hanyang University Guri Hospital, Guri, REPUBLIC OF 

KOREA, 2Hanyang University Hospital, Seoul, REPUBLIC 

OF KOREA 

PURPOSE 

The purpose of this study was to evaluate the hyperintense 

cystic or necrotic cervical lymph node on T1-weighted MR 

image and to ascertain the usefulness of MR imaging in the 

diagnosis of cystic nodal metastasis of thyroid carcinoma. 


The MR findings of cystic or necrotic cervical lymph nodes 

were analyzed retrospectively in 15 patients (males 9, 

females 6, 8-85 years old, mean: 42 years old), with regard 

to the signal intensity of cystic or necrotic portion on T1weighted 

MR image. They were histologically proven to be 

thyroid carcinomas (5 patients: 4 papillary carcinoma, 1 

anaplastic carcinoma, M:F = 4:1, 30-66 years old, mean: 

40.6 years old), or other neoplastic, benign or inflammatory 

lesions (10 patients: 5 tuberculous lymphadenopathy, 3 lymphoma, 

1 nasopharyngeal squamous cell carcinoma, 1 

benign reactive hyperplasia, M:F = 5:5, 8-85 years old, 

mean: 42.7 years old). 

RESULTS 

In all 5 patients with cystic nodal metastases from thyroid 

carcinoma, T1-weighted MR image showed hyperintense 

signal, especially in the cystic portion. Cystic nodal metastasis 

on MR was demonstrated in 5 (45%) of 11 patients with 

thyroid carcinoma. This characteristic signal intensity pattern 

was not found in any of the other 10 patients, in whom 

pathologic exam revealed cystic nodes from other malignancy 

(4 patients), tuberculous lymphadenopathy (5 patients) 

and benign reactive hyperplasia (1 patient). 

CONCLUSION 

MR imaging is useful for diagnosing cystic change within 

cervical lymph node; the presence of a hyperintense cystic 

node on T1-weighted image strongly indicates metastasis 

from thyroid carcinoma, especially papillary carcinoma. The 

hyperintense signal on T1-weighted image is thought to be 

due to the high colloid or thyroglobulin content within cystic 

nodes suggesting thyroid origin, as previously reported. 

KEY WORDS: Thyroid cancer, neck node, MR imaging 

Posters

Posters 

Poster 114 

Role of Ultrasound and Ultrasound-Guided Fine-Needle 

Aspiration in the Follow Up of Patient Operated on for 

Thyroid Cancer 

Lee, Y. · Lee, N. · Kim, J. 

Anam Hospital, Korea University College of Medicine 


PURPOSE 

The purpose of this study was to determine the role of ultrasound 

(US) and US-guided fine-needle aspiration in the follow 

up of patient operated on for thyroid cancer. 


We examined the patients operated on for thyroid cancer in 

our hospital with US and US-guided fine-needle aspiration 

over a 3-year period at our institution. All of them underwent 

thyroid surgery and radioactive isotope therapy with initial 

diagnosis of papillary carcinoma. US-guided fine-needle 

aspiration was carried out if recurrence was suspected during 

US examination. Whole body iodine-131 scan or serum thyroglobulin 

level was also checked concurrently. 

RESULTS 

There were 14 US-guided fine-needle aspirations in 14 

patients. Four local recurrences at previous thyroid beds and 

five regional recurrences at cervical lymph nodes were found 

after US and US-guided fine-needle aspiration. All these 

patients were treated by surgery. One recurrence was detected 

in case of undetectable serum thyroglobulin level. Other 

five lesions with US-guided fine-needle aspirations were 

confirmed benign lymph node or postoperative fibrosis. 

CONCLUSION 

Getting familiar with US findings of normal postoperative 

neck and various loco-regional recurrences in previously 

treated patient with thyroid cancer will provide guidance to 

the radiologist performing cervical US examination. 

KEY WORDS: Thyroid, fine-needle aspiration, lymph node 

Poster 115 

Cervical Tuberculous Lymphadenopathy: Conventional 

and Diffusion-Weighted MR Imaging Findings Focusing 

on Necrosis 

Park, D. 1 · Kim, S. 1 · Kim, Y. 1 · Kim, Y. 1 · Park, C. 1 · Lee, S. 2 

· Hahm, C. 2 

1Hanyang University Guri Hospital, Guri, REPUBLIC OF 

KOREA, 2Hanyang University Hospital, Seoul, REPUBLIC 

OF KOREA 

PURPOSE 

The purpose of this study is to evaluate the conventional and 

diffusion-weighted MR findings focusing on necrosis of cervical 

tuberculous lymphadenopathy and to compare the 

imaging findings with clinical findings. 

348 


MR imaging studies of 5 patients (1 male, 4 females, 8-85 

years old, mean: 43.2 years old) with histologically proven 

cervical tuberculous lymphadenopathy were analyzed retrospectively 

with regard to homogeneity, signal intensity, and 

enhancement of diseased nodes after contrast injection. The 

imaging findings were grouped by patterns and correlated 

with clinical signs or symptoms. 

RESULTS 

Two imaging patterns of cervical tuberculous lymphadenopathy 

were seen according to the signal intensity of 

its necrotic portion on T2-weighted images. In three of five 

patients, type I nodes were inhomogeneous with a strong 

peripheral enhancement after contrast injection. Enhancing 

areas were of intermediate intensity on T1-weighted images 

and hypointense on T2-weighted images. Unenhanced areas 

were relatively hypointense on T1- and T2-weighted images. 

In the remaining two patients, type II nodes were like type I 

nodes but hyperintense in unenhanced areas on T2-weighted 

images. They were found as larger conglomerated nodes, 

often forming abscess-like cavities, as opposed to small, separately 

found necrotic nodes in Type I. Diffusion-weighted 

image showed hypointense in unenhanced areas of all nodes 

and intermediate intensity in enhancing areas. Patients with 

cervical tuberculous lymphadenitis had similar clinical signs 

and symptoms with similar clinical history, irrespective of 

nodal type. 

CONCLUSION 

Cervical tuberculous lymphadenitis specifically accompanies 

inhomogeneous, ring-enhancing nodes containing central 

unenhanced areas with hypointensity on T2- and diffusion-weighted 

images. Hyperintensity of unenhanced area 

on T2-weighted image suggests progression toward abscesslike 

cavity with larger conglomerated nodes. However, such 

nodal appearance on MR imaging is not correlated with the 

clinical manifestation, history, or even the maturation degree 

of caseation necrosis within the tuberculous nodes. 

KEY WORDS: Neck node, MR imaging, tuberculosis 

Poster 116 

Radiation Necrosis as a Mimic of Intracranial Tumor 

Extension following Treatment of Paranasal Sinus 

Malignancies 

Hu, L. S. · Defatta, R. · Yetkin, Z. · Mendelsohn, D. B. 

University of Texas, Southwestern Medical Center 

Dallas, TX 

PURPOSE 

To summarize salient imaging features of brain radiation 

necrosis as a complication of paranasal sinus tumor treatment 

on conventional and dynamic perfusion MR imaging 

and PET scanning. 


Three patients, two with squamous cell carcinoma of the ethmoid 

sinus and one with high-grade sarcoma of the maxillary 

sinus, underwent radiation therapy (n = 3) with concomitant 

surgical resection (n = 2) and/or chemotherapy (n = 

2). None of the patients had evidence of intracranial exten-

sion at the time of initial treatment. Total radiation doses 

ranged from 6480 to 15,200 cGy. Conventional MR images 

(FLAIR, T2-weighted, pre and postcontrast T1-weighted 

sequences) of the head and neck were obtained for all 

patients (n = 3), along with dynamic perfusion MR (n = 1) 

and PET scanning (n = 2). The diagnosis of radiation necrosis 

was confirmed by brain tissue biopsy (n = 1) and serial 

MR examinations (n = 2) showing interval improvement in 

brain lesions in the absence of coinciding therapy (i.e., additional 

radiation or chemotherapy). 

RESULTS 

On posttreatment imaging studies, development of intraparenchymal 

lesions occurred in all patients, ranging from 6 

to 33 months status postradiation, and manifested by contrast 

enhancement and abnormal T2-weighted and FLAIR signal. 

The lesions occurred in the frontal lobes in all cases, particularly 

the white matter abutting and underlying the gyrus 

rectus and orbitofrontal gyrus, and were bilateral in two 

cases. Subsequent development of temporal lobe lesions 

with contrast enhancement occurred in two of the three cases 

and ranged from 7 to 19 months after development of the 

frontal lobe abnormalities. In none of the cases did the original 

tumor site show evidence of residual or recurrent disease 

or direct contiguity with the frontal lobe lesions, making 

neoplastic recurrence less favored. The diagnosis of radiation 

necrosis was further substantiated by evidence of 

hypoperfusion on MR imaging and decreased metabolic 

activity on PET. 

CONCLUSION 

Radiation injury resulting from paranasal sinus tumor therapy 

occasionally occurs within areas remote from the original 

tumor site. Ocular manifestations comprise the majority of 

described cases (1). Radiation injury can, however, involve 

the brain parenchyma (2, 3) and is often mistaken for tumor 

recurrence (4). Recognition of the pattern of radiation necrosis 

is important for correct image interpretation. These findings 

should not be mistaken for tumor recurrence, as accurate 

distinction has implications on further management and 

prognosis. Perfusion MR and PET imaging can be included 

in the armamentarium of techniques to further support the 

diagnosis of radiation necrosis and avert unnecessary craniotomy 

for tissue diagnosis. 

REFERENCES 

1. Morita K, Kawabe Y. Late effects on the eye of conformation 

radiotherapy for carcinoma of paranasal sinuses and nasal 

cavity. Radiology 1979;130(1):227-232 

2. Garcia-Torres E, Piekarski JD, Doyon D, et al. Developmental 

x-ray computed tomographic and magnetic resonance studies 

of swinging frontal cerebral radionecrosis. J Radiol 

1987;68(6-7):413-420 

3. Satran R, Lapham LW, Kido DR. Late cerebral radionecrosis 

after conventional irradiation of cerebral tumors. Rev Neurol 

1984;140(4):249-255 

4. Eyster EF, Nielsen SL, Sheline GE, Wilson CB. Cerebral radiation 

necrosis simulating a brain tumor. J Neurosurg 

1974;40(2):267-271 

KEY WORDS: Radiation, necrosis, paranasal 

349 

Poster 117 

Usual Skull Base Lesions: Unusual Imaging 

Manifestations 

Vibhute, P. G. · Naidich, T. P. · Lee, J. S. · Som, P. M. 


New York, NY 

PURPOSE 

The skull base harbors many lesions of diverse type. In most 

of these, the characteristic lesion location and imaging features 

suggest a reasonable, limited differential diagnosis. 

Rarely, however, atypical lesion appearance and/or location 

may confuse the diagnosis and confound the radiologist. We 

present four pathologically proven cases of usual skull base 

lesions with atypical imaging appearances. 


Retrospective review of the institutional database for the 

years 2000-2004 disclosed four common skull base lesions 

that presented markedly atypical imaging features. In each 

case, surgical resection of the lesion provided a firm unanticipated 

histologic diagnosis and gross photographs of the 

lesion for comparison with the imaging studies. 

RESULTS 

The four lesions proved to be (1) a large right jugular fossa 

schwannoma, which exhibited irregular erosion and destruction 

of the jugular foramen, relatively low T2 signal intensity, 

and foci of hemorrhage; (2) an expansile left petrous apex 

cholesterol granuloma, which extended posteriorly to bevel 

and expand the internal auditory canal; (3) an intraosseous 

arachnoid cyst of the basisphenoid, which extended into the 

sphenoid sinus, Meckel’s cave and the petrous apex; and (4) 

a benign fibro-osseous lesion of the clivus, which was of the 

cystic pagetoid form. In each case the intraoperative photographs 

and histologic sections of the lesion helped to explain 

the imaging features. 

CONCLUSION 

Atypical imaging features may obscure the nature of common 

lesions of the skull base. Such “zebras” increase the difficulty 

of imaging diagnosis, but add “zest” to the life of the 

radiologist. 

KEY WORDS: Skull base, petrous bone, clivus 

Poster 118 

64-Slice Multidetector CT of the Temporal Bone: Normal 

Sagittal Anatomical Atlas 

Lane, J. I. · Lindell, E. P. · Witte, R. J. · Driscoll, C. L. W. 

Mayo Clinic 

Rochester, MN 

PURPOSE 

Direct sagittal CT imaging of the temporal bone was advocated 

almost two decades ago as an imaging plane of significant 

interest to surgeons since it follows the plane of surgical 

exposure. This technique was cumbersome, requiring a 

specialized head holder and modifications in routine patient 

positioning. With the advent of multidetector CT, the poten- 

Posters

Posters 

tial to reconstruct orthogonal planes of section from a single 

acquisition has been realized. The development of 64-slice 

CT has permitted the routine display of temporal bone anatomy 

in the sagittal plane utilizing isometric voxels without 

loss of resolution. Familiarity with anatomy in the sagittal 

plane can be aided with use of a standard atlas. 


Sixty-four-slice CT examinations of the temporal bone were 

acquired in a direct axial plane using helical technique (120 

kV, 350 mA, pitch .85, rotation time 1 sec, slice thickness .6 

mm) Sagittal reconstructions were oriented in the plane of 

the long axis of the tympanic portion of the facial nerve. 

MicroCT images of the cadaver specimen were acquired on 

a research scanner (35 kV, 50 mA, 18 sec exposures x 360 

views, total scan time 13 hours). Sagittal reconstructions are 

displayed alongside of sagittal CT microscopy images of a 

human cadaveric temporal bone (20 um voxel size) in the 

identical plane for comparison. 

RESULTS 

Reconstructed CT images in the sagittal plane acquired using 

helical technique on a 64-slice scanner compare favorably to 

previously published images acquired in the direct sagittal 

plane. Images displayed in conjunction with microCT data 

demonstrate normal anatomical appearance of facial nerve 

canal, vestibular aqueduct, vestibule, internal auditory canal, 

tegmen tympani, round window, semicircular canals, common 

crus, and chorda tympani canal in the sagittal plane. 

CONCLUSION 

Reconstructed images in the sagittal plane from a 64-slice 

scanner can replace direct sagittal scanning. Familiarity with 

normal sagittal temporal bone anatomy can be aided with use 

of this standard CT atlas. 

KEY WORDS: Multidetector CT, CT microscopy, temporal 

bone 

Poster 119 

Contribution of High-Resolution Multiplanar Reformats 

of the Skull Base to the Detection of Skull Base Fractures 

Flis, C. M. · Sibtain, N. A. · Connor, S. E. J. 

King’s College Hospital 

London, UNITED KINGDOM 

PURPOSE 

To investigate the contribution of routine review of submillimetric 

multiplanar reformats (MPR) to the diagnosis of 

skull base fractures. 


We prospectively analyzed all CT heads (GE Lightspeed 16 

scanner) referred following cranial trauma under established 

(National Institute for Clinical Excellence) guidelines over a 

6-month period (n = 407). The reformatted 5 mm (thick) 

axial sections and subsequently the high-resolution multiplanar 

reformats (HRMPRs) were viewed on an ADW 4.1 

workstation using bone windows and algorithm. All skull 

base fractures and related features (e.g., intracranial air) 

were recorded by the consensus of two radiologists. They 

were classified as anatomically significant or nonsignificant 

350 

fractures on the basis of eight criteria (e.g., extension to 

major vascular channels). The clinical features of skull base 

injury and any subsequent treatment were noted in all cases 

of skull base fracture. 

RESULTS 

Using the HRMPRs, there were 80 separate skull base fractures 

detected in 36/407 patients. Fifty-seven of 80 of all 

fractures were visible on the thick sections. In 8/36 patients 

the significant anatomical features were only evident on 

review of the HRMPRs. In 6/36 patients none of the skull 

base fractures were visible on the thick sections; however 

there were no associated clinical features and no therapeutic 

requirement. 

CONCLUSION 

Only 71% of skull base fractures and 78% of patients with 

skull base fractures demonstrating significant anatomical 

features were detected on the 5 mm axial sections using 

HRMPRs as a gold standard. There were no clinical sequelae 

at short term follow up of those fractures only evident on 

HRMPRs. 

KEY WORDS: Fracture, skull base 

Poster 120 

Single-Shot Fast Spin-Echo Diffusion-Weighted MR 

Imaging of Cholesteatoma 

Yano, T. · Kodama, T. · Miyata, Y. · Tamura, S. 

University of Miyazaki 

Miyazaki, JAPAN 

PURPOSE 

Conventional MR imaging including contrast-enhanced 

study can detect a cholesteatoma as a soft tissue mass without 

contrast enhancement or with peripheral enhancement. 

However, because of nonspecific signal intensities of the 

cholesteatoma, it is sometimes difficult to distinguish a 

cholesteatoma from other lesions such as high proteinous 

fluid collection or postoperative fibrosis. Some previous 

reports suggested the usefulness of the diffusion-weighted 

imaging using echo-planar sequence (EPI) to differentiate 

cholesteatomas from complicated inflammatory process or 

postoperative change. Since diffusion-weighted imaging 

using EPI has disadvantages of prominent susceptibility artifacts 

and image distortion, this pulse sequence seems to be 

less suitable for the imaging of the temporal bone. The purpose 

of this study was to evaluate the usefulness of singleshot 

fast spin-echo (SSFSE) diffusion-weighted imaging in 

the detection of temporal bone cholesteatoma. 


We retrospectively reviewed 23 patients (17 males and 6 

females, 7-71 years old) with suspected temporal bone 

cholesteatoma who underwent MR imaging, including conventional 

spin echo (SE) images and SSFSE diffusionweighted 

imaging. Five of 23 patients had an operation of 

cholesteatoma before. Two patients showed bilateral lesions. 

MR imaging was performed at a 1.5 T using a quadrature 

detection head coil or a surface coil. Diffusion-weighted 

images were obtained with a SSFSE pulse sequence (TR/TE: 

12000/108 ms, B = 0, 900 sec/mm 2 , slice thickness: 6 mm,

gap: 1.2 mm, FOV: 22 x 22 cm). In all patients, high-resolution 

CT showed soft tissue density mass in the temporal 

bone. 

RESULTS 

Operation was performed in 22 of the 25 ears and 

cholesteatomas were found in 20 ears. In 18 of the 20 

cholesteatomas, diffusion-weighted imaging showed 

markedly hyperintense signal compared with that of brain 

tissue (sensitivity of 90%). The diameter of two 

cholesteatomas with negative diffusion-weighted imaging 

was 4 mm and 8 mm, respectively. Three patients (chronic 

otitis media: 2, cholesterol granuloma: 1) were followed 

without surgery depending on conventional MR and negative 

diffusion-weighted findings. 

CONCLUSION 

Single-shot fast spin-echo diffusion-weighted imaging is a 

useful technique to detect cholesteatomas as markedly 

hyperintense lesions because of less susceptibility artifact. 

However, small cholesteatomas may be missed because of 

low signal-to-noise ratio and low spatial resolution. 

KEY WORDS: Cholesteatoma, diffusion-weighted imaging, 

MR imaging 


121-148 

Poster 121 

Current State of Aneurysm Treatment 

Cramer, D. · Miller, B. · Siddiqi, J. · Muilli, D. 

Arrowhead Regional Medical Center 

Colton, CA 

PURPOSE 

Many view the pinnacle of aneurysm surgery occurred in the 

1980s when leaders such as Yasargil, Smith and Sundt 

reported mortality rates of only 4-6% and morbidity rates of 

under 10%. However, since the 1980s, the world has seen an 

exponential growth in endovascular technology for the treatment 

of vascular disease and today, many believe endovascular 

treatment of aneurysms may be the safest and most 

appropriate therapy. The results from the International 

Subarachnoid Aneurysm Trial demonstrated increased likelihood 

of a good outcome for patients treated by endovascular 

methods as compared with clipping. The US Food and Drug 

Administration recently has approved an intracranial stent 

(Neuroform Stent) to treat aneurysms that were once regarded 

as “uncoilable” because of their wide neck. We hypothesize 

that the current literature (2002-Sept. 2004) will reveal 

modern endovascular treatment of aneurysms with coiling 

(with or without stenting) is associated with improved neurologic 

outcomes and less perioperative complications compared 

to microsurgical clipping of anterior or posterior circulation 

aneurysms. 


A Medline search for “cerebral aneurysm treatment” was 

done between Jan. 2002-Sept. 2004 yielding 760 articles 

(coiling = 179, surgery = 581). Only articles reporting a 

351 

measurement of outcome were included in our meta-analysis 

of 29,997 patients. Besides the data collected seen in Table 

1, we included: incomplete coiling and clipping, regrowth 

after treatment, percentage of definite surgery, procedure 

complications, location, intraoperative rupture rate, length of 

hospital stay, and in-patient hospital costs. 

RESULTS 

After a review of the literature, a significant difference was 

found demonstrating improved neurologic outcome (GOS 4 

or 5) for patients undergoing coiling compared with clipping 

(81.20% vs 74.67%). No age difference was seen between 

the two treatment options; however, published reports 

demonstrated statistical significance between Hunt and Hess 

(H&H) grades and treatment option. Neurologic complication 

rates from surgery were almost twice as frequent vs coiling 

(11.61% vs 7.05%). More patients underwent surgery if 

presenting with an unruptured aneurysm (79.61% vs 

54.76%) or H&H grade 1 (60.04% vs 37.64%). The majority 

of patients were treated with endovascular surgery if the 

aneurysm was located in the posterior circulation (32.46% vs 

7.06%). Coiling was associated with a shortened length of 

length of hospital stay and lower cost; 2 vs 5 days, and the 

total average cost for clipping was $21,800 vs $13,200. In 

terms of a treatment requiring no further intervention, surgical 

clipping completes treatment 94.62% of the time; whereas 

coiling is definitive only 85.40%. Surgical incomplete 

clipping is far less common than incomplete endovascular 

coiling, but neurologic outcomes after retreatment were similar. 

Patient characteristics & presentation 

Patients Age Hunt Hunt Hunt Hunt Good Moderate Poor Dead Death or 

(yrs) &Hess &Hess &Hess &Hess (GOS Outcome Outcome Outcome Dependence 

Grade Grade Grade Grade 4 or5 (Gos 3) (GOS 2) (GOS 1) % 

0 (%) 1 (%) 3 (%) 5 (%) (%) (%) (%) (%) 

Surgery 23,342 52.75 79.61 60.04 9.31 4.46 74.67 11.95 6.22 7.91 15.94 

std deviation 2.39 49.51 26.36 13.08 8.40 7.96 2.95 4.16 5.01 10.22 

Number 21,573 6,029 5,276 4,557 4,726 9,069 6,818 7,888 8,164 9,157 

Coiling 6,655 53.49 54.76 37.64 17.71 5.29 81.20 7.55 9.14 5.11 15.31 

std deviation 1.59 4.274 26.91 12.85 3.26 11.95 5.86 4.48 4.62 6.33 

Number 1,824 3,273 3,951 3,93 3,212 5,411 2,390 4,837 5,023 5,374 

CONCLUSION 

Our meta-analysis of current research clearly shows 

improved outcomes with endovascular techniques compared 

with clipping. Despite patients with worse H&H grades and 

posterior circulation locations designated to coiling, morbidity 

and mortality decreased. Microsurgical clipping was 

associated with a higher likelihood of definitive treatment 

but outcomes were similar after adjunctive endovascular 

treatment. 

KEY WORDS: Clipping, coiling, outcomes 

Posters

Posters 

352 

Poster 123 

Poster 125 

Pretreatment Planning and Virtual Stent Visualization in Giant Aneurysm at the Junction of the Internal Carotid 

the Parent Artery of Intracranial Aneurysms 

and Persistent Primitive Trigeminal Artery Treated with 

Endovascular Guglielmi Detachable Coiling: A Case 

Report of Experience in Ramathibodi Hospital 

Karmonik, C. 1 · Strother, C. M. 1 · Chen, X. 2 · Deinzer, F. 2 · 

Klucznick, R. 3 · Mawad, M. E. 1,3 

1 2 Baylor College of Medicine, Houston, TX, Siemens 

Medical Systems, Erlangen, GERMANY, 3The Methodist 

Hospital, Houston, TX 

PURPOSE 

To demonstrate the value of the virtual reconstruction of the 

parent artery segment across the neck of intracranial 

aneurysms for stent-assisted coil placement. 


Angiographic data were obtained with bi-plane C-arm systems 

(Axiom Artis; Siemens Medical System, Erlangen, 

Germany). The algorithm for creating a virtual image of the 

reconstructed parent artery across the aneurysm ostium (virtual 

stent) was implemented on the Siemens Leonardo workstation 

as a research prototype. All images were analyzed 

retrospectively after treatment had been completed. After 

interpolating the centerline of the parent artery across the 

aneurysm ostium, contiguous 2D cross sections perpendicular 

to the interpolated centerline were obtained. The radius of 

the virtual stent in each cross section was interpolated linearly 

using the radii measured at the normal proximal and 

distal segments of the parent artery. The morphologies of 

two aneurysms treated with stent-assisted coiling were 

assessed in this way. One was a paraophthalmic aneurysm 

having a sidewall geometry (case1), the other was a carotid 

bifurcation aneurysm (case 2). 

RESULTS 

For both cases, pretreatment 2D DSA projection images 

show the course and size of the parent artery, the aneurysm 

size, and the neck length. The standard 3D DSA surface volume 

reconstruction emphasizes the expansion and irregularity 

of the artery from which the aneurysm arises. However, 

only the virtual stent superimposed on the 3D DSA surface 

reconstruction and in the 2D cross sections demonstrates that 

in case 1 the aneurysm ostium involves at least 180 degrees 

of the parent artery circumference. For both cases, the virtual 

stent also demonstrates the existence of “pockets” of the 

aneurysm that lie outside of the boundaries of the virtual 

stent. On the posttreatment 2D DSA projection images, coil 

loops in these “pockets” appear to lie within the parent 

artery. Comparing these with the cut-plane section and surface 

volume reconstruction shows that these are, in fact, outside 

of the boundaries of the stent and are not compromising 

the parent artery. 

CONCLUSION 

The technique described allows the visualization of a virtual 

stent in 3D DSA volume reconstructions. As demonstrated in 

two cases, this provides information that can be potentially 

valuable in pretreatment planning and should also facilitate 

verification that coils are not herniating through nonvisible 

portions of stents used for stent-assisted coiling. 

KEY WORDS: Aneurysm, virtual stent, stent-assisted coiling 

Jiarakongmun, P. · Pongpech, S. 

Ramathibodi Hospital 


PURPOSE 

The persistent trigeminal artery (PPTA) is the most common 

persistent carotid-basilar anastomotic channel observed in 

adult life. It can be associated with other congenital abnormalities 

such as cerebral aneurysms, but only rarely do 

aneurysms of the primitive trigeminal artery itself arise. We 

retrospectively review a rare case of an aneurysm arising 

from PPTA and discuss its significance and treatment. 


A 65-year-old woman present with chronic headache for 2 

years with progressive worsening in 1 week, and CT scan 

shows large aneurysm in right cavernous sinus with partial 

wall calcification. There is no evidence of intracranial hemorrhage 

or cerebral emboli or cortical infarction. Cerebral 

angiogram on right internal carotid injection shows an 

unruptured giant aneurysm at the junction of the right internal 

carotid artery (RICA) and the persistent primitive trigeminal 

artery with distention of the right cavernous sinus. There 

is no intramural clot or stagnation of the contrast medium. 

The large PPTA provides main blood supply of the posterior 

fossa; however the smaller vertebral arteries still contributing 

supply to the posterior fossa. Due to intracavernous location 

of the PPTA, hemodynamic pulsatility and dural distention 

from stretching of the dural wall of cavernous sinus 

from large aneurysm are possible cause of chronic headache. 

The patient underwent successful serial partial transarterial 

embolization with Guglielmi detachable coiling in the 

attempt not to sacrifice the PPTA and her symptoms of 

severe headache became much improvement on the clinical 

follow up. 

RESULTS 

Persistent trigeminal artery is rare condition and by far more 

common presenting unilaterally. The prevalence of intracranial 

aneurysms in patients with PPTA is approximately 3% 

and not greater than in general population, and only rarely do 

aneurysms of the PPTA itself or its junction arise. The presenting 

symptoms of the aneurysm arise from PPTA in previous 

reports are varied and most common symptoms are 

progressive ipsilateral abducen nerve palsy and less common 

sensory disturbance around orbit and progressive headache. 

However, due to intracavernous location of the PPTA 

aneurysm, the risk of intradural hemorrhage is possibly very 

low. For therapeutic options, most of the reported cases were 

treated by surgical clipping or ligation of the parent internal 

carotid artery, with or without arterial bypass anastomosis. A 

common postsurgical complication is diplopia due to cramial 

nerve injury in the attempt to approach the cavernous 

sinus.

CONCLUSION 

To the best of our knowledge, there is no previous report in 

the literature of giant aneurysm arising from junction of 

PPTA and carotid artery treated with endovascular GDC 

coiling. 

REFERENCES 

1. Cloft HJ, Razack N, Kallmes DF. Craniofacial pain and 

incomplete oculomotor palsy associated with ipsilateral 

primitive trigeminal artery. Case report. J Neurosurg Sci 

1993;37:251-255 

2. Freitas PE, Aquini MG, Chemale I. Persistent primitive 

trigeminal artery aneurysm. Surg Neurol 1986;26:373-374 

KEY WORDS: Trigeminal artery, aneurysm 

Poster 126 

Therapeutic Management of Unruptured Intracranial 

Aneurysms in Japan 

Ishibashi, T. · Murayama, Y. · Saguchi, T. · Ebara, M. · Irie, 

K. · Takao, H. · Onoue, H. · Ogawa, T. · Abe, T. 

The Jikei University School of Medicine 

Tokyo, JAPAN 

PURPOSE 

The management of unruptured intracranial aneurysms is 

controversial. The most important task for an effective way 

of SAH prevention is to estimate the rupture risk of unruptured 

intracranial aneurysms (UIAs) and to reduce the operative 

risk for surgical or interventional treatment. We recommend 

treatment to an aneurysm with a size of 5 mm or more 

and our first recommendation of therapeutic option for UIAs 

is endovascular coil embolization. The purpose of this study 

is to verify the validity of therapeutic management of UIAs 

in a single center’s experience in Japan. 


From March 2003 through November 2004, a total of 229 

patients with 286 UIAs were referred to our institution. All 

aneurysm sizes were measured by CTA or MR imaging in an 

outpatient clinic. The aneurysms were classified according 

to their size as follows; small: (0 to 4 mm diameter), medium/small 

(m/s): (5 to 10 mm diameter) with small neck (0 to 

3 mm), medium/wide (m/w): (5 to 10 mm diameter) with 

wide neck (> 4 mm), large size (large): (> 10 mm), giant: (> 

25 mm). We explained to patients the overall risk of 

endovascular surgery to be 5%; rupture during the procedure: 

1%, and thromboembolic events: 3-4%. Our first recommendation 

is endovascular coil embolization. The 

aneurysms without surgical treatment were followed and 

their size measured by CTA every 6 months. 

RESULTS 

Total therapeutic rate for UIAs was 25.5% (64 aneurysms) 

and almost 75% patients were selected for follow up without 

treatment. Attempts at embolization were performed in three 

patients (5%). Endovascular treatment was technically feasible 

for 54 of 56 UIAs. Only one patient had neurologic 

deficit according to a thromboembolic event, resulting in a 

morbidity rate of 1.8% and a mortality rate of 0%. On the 

other hand, three aneurysms were ruptured in patients without 

surgical treatment. Annual rupture rates and complica- 

353 

tion rates in each size of the aneurysm are shown in the 

Table. Complication rates in each size classification of 

aneurysm were lower than the annual rupture rate in our 

institute. 

Complication and annual rupture rates in each size of the aneurysms 

complication rate (%) n annual rupture rate (%) n 

small 0 0/13 1.3 1/137 

m/s 0 0/10 0 0/27 

m/w 0 0/20 5.8 1/47 

large 6.3 1/17 22.3 1/24 

CONCLUSION 

Our therapeutic management of UIAs confirmed the validity 

of endovascular coil embolization within our historical 

annual rupture rate and complication rate in each size classification 

of the aneurysm. We verified the validity of our therapeutic 

management of UIAs with endovascular surgery. 

KEY WORDS: Cerebral aneruysm, incidental, embolization 

Poster 127 

Intracranial Aneurysm Growth after Coil Embolization 

Bae Ju, K. 1 · Moon Hee, H. 1 · Hyun Seung, K. 1 · Oh Kee, K. 2 

· Sung Hyun, K. 2 

1Seoul National University College of Medicine, Seoul, 

REPUBLIC OF KOREA, 2Seoul National University 

College of Medicine, Bundang, REPUBLIC OF KOREA 

PURPOSE 

It is mandatory for the postembolization growing aneurysm 

to be treated urgently. The purpose is to show the clinical and 

radiologic features in patients with a postembolization growing 

aneurysm and to suggest possible explanations for this 

phenomenon. 


A total of 455 patients harboring 511 intracranial aneurysms 

were treated by endovascular coil embolization during last 8 

years. We experienced eight cases of a postembolization 

growing aneurysm. Radiologic and clinical data were 

reviewed to find the possible explanations for enlargement 

of aneurysms embolized. 

RESULTS 

Among eight growing aneurysms after embolization, four 

were ruptured at preembolization. Four aneurysms were 

present at the anterior circulation (2 ACom, 1 ICA, and 1 

PCom), and four at the posterior circulation (3 basilar tip and 

1 basilar-SCA branching). All the basilar tip and ACom 

aneurysms were the bifurcation aneurysms (63 %). Initial 

diameters of aneurysms were 5 to 10 mm in four and larger 

than 10 mm in four. The aneurysm neck was broad in five (> 

4 mm) and dome to neck ratio (DNR) was less than two in 

six (67 %). Immediate postembolization results were total or 

near total in four, subtotal in two, and incomplete in two. The 

growing aneurysm was confirmed on the plain radiography 

and conventional angiography performed at 1.5 to 89 months 

postembolization (mean, 24.6 months.) Two aneurysms presented 

with rebleeding (22%) at the time of growth (1.5 and 

75 months postembolization each). Radiographically, coil 

meshes of the growing aneurysms were dispersed in three, 

displaced in five, and these combined in two. 

Posters

Posters 

354 

CONCLUSION 

Poster 129 

Adequate coil packing is required to avoid the postembolization 

aneurysm growth, and the importance of followup 

imaging cannot be overemphasized especially in the 

broad-neck, low DNR aneurysm that was even totally 

Safety and Efficacy of the Trufill Detachable Coil 

System: Initial Experience and Follow Up in 23 

Intracranial Aneurysm Treatments 

occluded on angiography. 

Ken, L. · Burger, I. M. · Gailloud, P. 

KEY WORDS: Cerebral aneurysm, embolization, recanalization 


Baltimore, MD 

Poster 128 

Clinical Presentation, Treatment, and Outcome of Tiny 

Aneurysms 

Liu, H. · Wang, Y. · Tsai, Y. 

National Taiwan University Hospital 

Taipei, TAIWAN REPUBLIC OF CHINA 

PURPOSE 

How often is the rupture rate of tiny aneurysms (< 3 mm in 

longest diameter) and what is the outcome of such ruptured 

tiny aneurysms. 


From 2001 to June 2004, a total of 210 cases of aneurysms 

diagnosed in our institution. Forty patients had tiny 

aneurysms and a total of 50 aneurysms were found (6 

patients had multiple tiny aneurysms). They were 27 females 

and 13 males. Their age ranged from 25 years old to 93 years 

old, and 22 of them were less than 50 years old. 

RESULTS 

In the 50 aneurysms, 38 were in anterior circulation and 12 

in posterior circulation. Subarachnoid hemorrhage was the 

clinical presentation in 34 patients. Of the 34 patients, 44 

aneurysms were found. The distribution of the 44 ruptured 

tiny aneurysms: AComA : 25%, PComA: 16%, ICA: 20%. 

Outcome of 40 patients with tiny aneurysms 

Good Moderate disability Poor 

No treatment 8 5 

Clipping 13 4 1 

Coiling 8 1 

CONCLUSION 

Tiny aneurysms do rupture. Every patient with an unruptured 

aneurysm should be considered carefully and individually. 

The management takes not just the size of the aneurysm but 

also the other very important factors such as age, health, psychological 

factors, and most importantly, the risk of treatment 

of that particular aneurysm in that particular patient. 

KEY WORDS: Aneurysm, surgery, coiling 

PURPOSE 

Detachable microcoils were introduced in the early 1990s as 

a minimally invasive alternative to surgical clipping for the 

treatment of intracranial aneurysms. Since then, the design 

of microcoils has improved constantly. This study reports 

our preliminary results with the Trufill orbit system, a new 

detachable microcoil designed for the treatment of ruptured 

and unruptured intracranial aneurysms (Trufill orbit detachable 

coil system (DCS), Cordis Neurovascular, Miami, FL). 


This IRB-approved study is based on the retrospective evaluation 

of 23 coiling procedures performed in 20 patients at a 

single institution (10 ruptured and 13 unruptured 

aneurysms). One hundred and sixty-four coils were deployed 

(mean: 7, range: 1 to 27 coils per aneurysm). Monitored end 

points included incidence of device-related adverse events, 

incidence of procedure-related events, achieved aneurysm 

occlusion rate (primary in all patients, and at follow-up 

angiography in 9 patients so far). Balloon-assisted and stentassisted 

coiling was performed in 2 and 4 patients, respectively. 

Coiling success was rated as follows: complete, partial 

(> 90% aneurysm occlusion), and incomplete (< 90%). 

RESULTS 

No device-related adverse event was observed. One procedure-related 

adverse event was noted (4%) in a patient presenting 

with a ruptured MCA aneurysm after failed surgical 

treatment, resulting in an MCA stroke despite successful clot 

thrombolysis. Two patients with Hunt & Hess grade V subarachnoid 

hemorrhage died from causes unrelated to the procedure. 

Achieved occlusion rates (23 cases) were as follows: 

complete (15 cases, 65.2 %), partial > 90% (6 cases, 26.1%), 

incomplete < 90% (2 cases, 8.7%). One partial (> 90%) 

aneurysm was left partially occluded on purpose in order to 

spare a normal superior cerebellar artery arising from the 

aneurysm neck. Follow-up occlusion rates (9 cases so far): 

complete (7 cases, 77.8%), partial > 90% (1 case, 11.1%), 

incomplete < 90% (1 case, 11.1%). One aneurysm had progressed 

from partial to complete, and one from incomplete to 

partial on follow-up angiography (Fig.). No aneurysm 

recanalized or enlarged after treatment.

A 53-year-old woman with unruptured distal internal carotid 

artery aneurysm. The immediate postcoiling appearance is 

shown on the left (rated incomplete, < 90%). The follow-up 

angiogram obtained 10 months later is shown on the right, 

with coil compaction and aneurysm obliteration now 

upgraded to partial (> 90%). 

CONCLUSION 

The Trufill orbit system appears to be a safe and efficient 

new microcoil for the treatment of intracranial aneurysms. 

The combined primary complete or partial (> 90%) occlusion 

rate was 91.3%, with no observed recanalization at 

available follow up. 

KEY WORDS: Cerebral aneurysm, endovascular treatment, 

microcoil 

Poster 130 

Initial Experience with Hydrocoil 10 System for 

Endovascular Aneurysm Treatment 

White, P. M. · Horribine, L. · Keston, P. · Sellar, R. J. 

Centre for Interventional Neuroradiology Edinburgh 

Edinburgh, UNITED KINGDOM 

PURPOSE 

Hydrocoils have potential advantages of improved aneurysm 

packing and reduced recurrence risk. Initially hydrocoils 

were only available in 14/18 diameters, with a relatively stiff 

coil pusher and were not stretch resistant. Their role in treating 

small acutely ruptured aneurysms was therefore limited. 

The stretch resistant Hydrocoil 10 system with softer, more 

flexible pusher overcomes these problems. The first clinical 

use of Hydrocoil 10 in Europe was in Edinburgh on third of 

March 2004. We report initial experience with the Hydrocoil 

10 system over the ensuing 7 months. 


Forty-seven aneurysms (36 ruptured, 3 unruptured, 4 recoils) 

were treated (in part) using hydrocoil 10 in 43 patients (26 

female). Mean age 51. Mean maximum aneurysm dimension 

6.2 mm. WFNS grade: 7 patients were grade 0, 31 were 

grade 1/2, 3 were grade 3, 2 were grade 4/5. Over the same 

7-month period an additional 12 aneurysms were treated 

using hydrocoil 14/18 (but no HES 10) and 51 using bare 

platinum alone. So Hydrocoil 10s were used in 43% of all 

355 

cases. Aneurysm volume and packing were assessed on a 

purpose-designed program. Aneurysm occlusion was 

assessed using Raymond-Roy criteria. 

RESULTS 

Twenty-seven (57%) aneurysms were Raymond-Roy class 1 

(occluded), 16 (34%) were class 2 (neck remnant) and 4 

(9%) were class 3 (incomplete occlusion). Median aneurysm 

packing density was 68% (interquartile range 53-87.5). No 

rebleeding has occurred to date (longest follow up 8 

months). 

Clinical outcomes at latest follow up were: Glasgow 

Outcome Score (GOS) 5 (good) in 30, GOS 4 in 3, GOS 3 in 

5, GOS 2 in 2 and GOS 1 (death) in 3 patients respectively. 

Complications with HES 10 coils: 1 aneurysm rupture, 1 

delayed parent vessel (MCA) occlusion causing major stroke 

and subsequent death, 2 asymptomatic coil tails, 1 failure to 

retrieve a coil leading to asymptomatic vessel occlusion and 

1 thromboembolic event (clinically asymptomatic small 

occipital stroke). 

CONCLUSION 

The Hydrocoil 10 system extends the use of hydrogel-coated 

coils to most aneurysms including small, acutely ruptured 

aneurysms. Few technical complications were experienced 

with HES 10 coils. High packing density and angiographic 

results achieved are encouraging but a randomized controlled 

trial is required to definitively confirm the value of 

hydrocoils. 

KEY WORDS: Aneurysm, hydrocoil, packing 

Poster 131 

Resolution of Brain Edema following Coiling of Giant 

Partly Thrombosed Basilar Tip Aneurysm 

Tampieri, D. 1 · Melanson, D. 1 · Mohr, G. 2 · Lechter, M. 2 

1 Montreal Neurological Hospital, McGill University, 

Montreal, PQ, CANADA, 2 Jewish General Hospital, McGill 

University, Montreal, PQ, CANADA 

PURPOSE 

The aim of this presentation is to describe a case of giant 

partly thrombosed basilar tip aneurysm causing severe vasogenic 

edema which resolved completely following endovacular 

treatment using coils. 

Posters

Posters 


This is a 65-year-old female patient presenting with gait difficulty 

and confusion progressing over several months. The 

neurologic exam revealed a very confused and disoriented 

patient who was extremely ataxic. The condition rapidly 

deteriorated and she became completely bedridden. 

RESULTS 

The MR imaging and MRA revealed a 3 cm in diameter partly 

thrombosed giant aneurysm arising from the basilar apex. 

The lesion was responsible for severe mass effect on the 

midbrain and thalamus leading to severe vasogenic edema 

involving mainly the right thalamus, genu and posterior limb 

of the right internal capsule.There was no evidence of acute 

or chronic bleed or hydrocephalus. The angiogram confirmed 

the diagnosis and revealed the patent portion of the 

aneurysm and its neck, relatively narrow. In consideration of 

the clinical and imaging findings it was decided to proceed 

with coiling of the aneurysm. The procedure was performed 

in two separate sessions achieving total occlusion of the 

lesion. The MR imaging, MRA, and angiographic controls at 

5 months demonstrated a stable total occlusion of the 

aneurysm and the complete resolution of the edema in spite 

of some persistent mass effect. Clinically the patient had a 

total resolution of the symptomatology and she regained her 

regular life activities. 

CONCLUSION 

This case illustrates the changes caused by the pounding 

effect of circulating pulsatile blood in a partly thrombosed 

aneurysm leading to vasogenic edema. The use of coils 

achieves obliteration of the lesion and resolution of the transmitted 

pulsatile effect. Although this type of lesions carries a 

high recurrence rate the endovascular approach represents 

the treatment of choise leading to the complete clinical resolution 

of the symptoms. 

KEY WORDS: Giant aneurysm, coiling, edema 

Poster 132 

Delayed Contrast Washout in Aneurysms during 

Angiography: Flow Stagnation or Contrast Pooling? 

Lieber, B. B. 1 · Gounis, M. J. 1 · Bahar, N. 1 · Wakhloo, A. K. 2 

1 2 Univerity of Miami, Coral Gables, FL, Univerity of Miami, 

Miami, FL 

PURPOSE 

Delayed contrast washout form parts of cerebral aneurysms 

during angiography gave rise to speculations that higherthan-blood 

contrast material density may be responsible for 

356 

a contrast pooling. Gravity was proposed to be the acting 

force for this observation. We investigated the dispersion of 

contrast material descending in various fluids under gravity. 


Three blood sample vials (6.5 ml) were positioned vertically 

in a specially designed holder permitting rapid inversion. 

Each tube was filled with either saline (density = 1.009), 

canine blood (density = 1.056), or a glycerin-distilled water 

mixture (60-40 mixture, % by volume, density = 1.153). 

Subsequently, using a long spinal needle, 0.25 ml of a mixture 

containing 50% by volume of iodinated contrast agent 

(Omnipaque-300, Amersham Health, Buckinghamshire, 

UK) and 50% by volume of saline was added to the bottom 

of the tubes (density = 1.179). Using high-speed angiography 

(Siemens Angiostar Plus, Forcheim, Germany) operating 

at 15 fps, the tubes were inverted rapidly and simultaneously 

while acquiring a 20 second-long angiographic 

sequence. In a second experiment, the three vials containing 

the aforementioned fluids and the contrast material were 

shaken rigorously for 2 minutes. Angiographic images were 

taken of these vials every 5 minutes for a period of 1 hour. 

Image processing was employed to characterize the spatial 

and temporal distributions of the contrast throughout the different 

fluids. In addition, the speed of decent of the contrast 

in the vials was evaluated. 

RESULTS 

In the premixed vials the contrast remained homogeneously 

dispersed within the three fluids for the entire observation 

period of 1 hour. In the inversion experiment, the contrast 

propagated down the vials with velocities of 15, 4.6 and 2.6 

cm/sec in the saline, blood and the glycerol mixture, respectively. 

A high rate of descent (saline) created extensive flow 

disturbances and the contrast more readily dispersed in the 

vial. Conversely, the slow rate of descent (aqueous glycerol 

mixture) did no generate enough agitation in the vial and 

most of the contrast descended to the bottom. The rate of 

contrast descent in blood generated some agitation and the 

contrast partially dispersed in the blood and partially 

descended to the bottom. 

CONCLUSION 

The rate contrast dispersion is strongly influenced by the 

amount of agitation during the process, which is related to 

the magnitude of the velocity mismatch between the solvent 

and solute. A velocity mismatch of 15 cm/sec in the vial containing 

saline was sufficient to produce a nearly homogeneous 

dispersion in less than 1 sec. If one considers a normal 

injection of 3 cc of contrast in 1 second into the internal 

carotid artery in which blood flows at a rate of about 3 

cc/sec, the velocity mismatch is as high at 150 cm/sec for a 

4F catheter. This high velocity mismatch guarantees homogeneous 

dispersion of the contrast with the blood. Delayed 

contrast washout from aneurysms corresponds to regions of 

flow stagnation of blood rich in contrast and not the "pooling" 

of contrast alone. This may be important for aneurysm 

growth, inflammation, rupture, and thrombus formation. 

KEY WORDS: Contrast, stagnation, pooling

357 

Poster 133 

Poster 134 

Multiple Intracranial Dural Arteriovenous Fistulae: New MR Sequences: 3D FIESTA and Tricks: 

Synchronous and Metachronous Presentations 

Applications in Vascular Malformations of Spinal Cord 

Kim, D. · Kim, D. · Lee, Y. · Suh, S. · Kim, J. 

Yonsei University College of Medicine 


PURPOSE 

To describe the findings of synchronous and metachronous 

presentations of multiple intracranial dural arteriovenous fistulae 

(DAVF). 


From a database including 150 cases of intracranial DAVFs, 

five cases with multiple DAVFs were collected and the 

angiograms and medical records were reviewed. The clinical 

presentation, location of the lesions, Borden types and treatment 

were assessed. 

RESULTS 

Multiple DAVFs were discovered in two males and three 

females [incidence: 3.3% (5/150), age: 53-68 years, mean: 

60 years]. Neurologic deficit (n = 1), ocular symptoms (n = 

3) and intracranial hemorrhage (n = 1) were the initial presenting 

symptoms. Hemorrhage developed during the follow-up 

period in one case with metachronously developed 

DAVF in the sigmoid sinus (SS). Synchronous presentations 

(n = 3, M:F = 1:2) were found in the left cavernous sinus 

(CS, type II) + right marginal sinus (type I), superior sagittal 

sinus (SSS, type II) + right SS (type I) and left CS (type I) + 

right SS (type II). Transvenous embolizations for the aggressive 

lesions failed in the first two cases due to sinus occlusion, 

thus transarterial embolizations were performed with 

good outcome. The third patient developed signs of cerebellar 

venous congestion after combined transarterial and transvenous 

embolization. The initial location of the fistulae in 

the metachronous cases (n = 2, M:F = 1:1) were left CS and 

the SSS (type II). Subsequent DAVF newly developed in the 

left SS (type II, 17-month interval, hemorrhage) and the right 

transverse sinus (type I, 20-month interval), respectively. In 

both of these patients, transarterial embolizations were performed 

for the initial lesions due to failure of transvenous 

approach through the occluded sinuses. The initial lesion had 

occluded completely in the first case but the fistula was 

patent in the second case at the time of the development of 

the metachronous lesion. The aggressive lesions subsequently 

were treated by direct burr hole puncture and sinus 

embolization with good outcome. 

CONCLUSION 

Multiple DAVFs are relatively rare with variable patterns of 

presentation in terms of temporal development and location. 

Multiplicity seem to be associated with sinus occlusions and 

aggressive types. 

KEY WORDS: Dural arteriovenous fistula 

Jiménez de la Peña, M. · Carrascoso Arranz, J. · Recio 

Rodriguez, M. · Montoya Bordón, J. · López Pino, M. · 

Martínez de Vega, V. · Viaño López, J. 

Ntra Sra Rosario 

Madrid, SPAIN 

PURPOSE 

Preliminary experience of the new sequences 3D FIESTA 

and TRICKS in the evaluation of the spinal cord vascular 

malformations is shown. 


Three-dimensional FIESTA (fast imaging employing steadystate 

acquisition) and TRICKS (time-resolved imaging of 

contrast kinetics) were used. 

RESULTS 

Three-dimensional FIESTA is a noncontrast sequence using 

an extremely short repetition time between radiofrequency 

pulses; such high-resolution 3D volume images can be 

acquired rapidly (less than 4 minutes), providing a great contrast 

between CSF and the spinal cord. Multiplanar reconstructions 

are possible. Three-dimensional FIESTA benefits 

conventional MR images identifiying: 1) normal vascular 

and nervous structures, 2) small vascular malformations, 3) 

size, number and tortuosity of the vessels, 4) level of the fistula, 

helping posterior digital spinal arteriography (DSA). 

Feeding arteries are evaluated incompletely without contrast, 

making this sequence complementary to the AngioMR imaging. 

Time-resolved imaging of contrast kinetics is a technique 

that allows acceleration of the temporal resolution of 

the 3D imaging without sacrificing spatial resolution. Timeresolved 

imaging of contrast kinetics provides a pure arterial 

phase, with the acquisition of images each 4 seconds. 

Provides: 1) Clear visualization of the filling of the vascular 

malformation, depicting feeding arteries, 2) Contribution or 

not of the anterior spinal artery, 3) Possible role in the assessment 

of treated malformations to reduce the number of follow-up 

DSA examinations. Small feeding arteries can be 

missed, keeping DSA the gold standard examination. 

Posters

Posters 

CONCLUSION 

Three-dimensional FIESTA benefits noncontrast MR studies 

in the evaluation of the vascular malformations of the spinal 

cord, specially in the small ones. It is a noncontrast 

sequence, being useful in pregnant women or in patients 

with a history of adverse reactions to gadolinium. Timeresolved 

imaging of contrast kinetics clearly improves MR 

angiography studies, where a pure arterial phase is difficult 

to obtain, depicting the feeding pedicle. This sequence is a 

potential application in the follow up of treated vascular malformations. 

KEY WORDS: Vascular malformation, 3D FIESTA, tricks 

Poster 135 

Validation of the Adamkiewicz Artery as Imaged by MR 

Angiography 

Nijenhuis, R. J. 1 · Mull, M. 2 · Wilmink, J. T. 1 · Thron, A. 2 · 

Backes, W. H. 1 

1Maastricht University Hospital, Maastricht, THE 

NETHERLANDS, 2Aachen University Hospital, Aachen, 

GERMANY 

PURPOSE 

To validate the location and morphology of the intradural 

arteries of the spinal cord as imaged by contrast-enhanced 

MR angiography (CE MRA). The anterior spinal artery 

(ASA) and Adamkiewicz artery (AKA), which are considered 

the main arteries supplying blood to the thoracolumbar 

spinal cord, are millimeter-sized vessels. To image these 

small arteries, and particularly to separate them from closely 

adjacent thicker veins with a similar configuration 

requires both a high spatial and temporal resolution. 

358 


Sixteen patients with suspected spinal vascular abnormalities 

underwent a CE MRA and a digital subtraction angiography 

(DSA) study. The 1.5 T MR imaging protocol included a bolus 

timing sequence in which the arrival time of contrast in the 

distal aorta was determined. Subsequently, a 40 s CE MRA 

exam was performed using a 3D spoiled gradient-echo 

sequence. A centric k-space filling scheme synchronized to the 

arrival of the contrast agent (0.3 mmol/kg) was used to suppress 

venous blood. Voxel sizes were 0.8 × 0.8 × 1.2 mm at 

acquisition. Level and side of the AKA was determined by the 

creator of (curved) multiplanar reformatted (MPR) images, 

who was blinded for the DSA findings. Three readers jointly 

scored the level and side of the AKA on DSA projection 

images and MRA MPR images and compared the image quality 

in terms of vessel conspicuity, contrast, continuity, sharpness, 

and background homogeneity on a 5-point scale. 

Vascular pathology was evaluated in another study. 

RESULTS 

The localization and morphology of the AKA by MRA was 

in agreement with the DSA result in 15 out of 16 cases. One 

mismatch of one vertebral level (not side) appeared to be due 

to interrupted vessel continuity. All intradural arteries identified 

by MRA were confirmed by DSA. Comparison of image 

quality revealed that DSA remains superior (p < 0.001) to 

MRA concerning vessel sharpness and continuity, and background 

homogeneity. Overall vessel conspicuity was judged 

to be somewhat better for DSA (p < 0.05), while contrast was 

better for CE MRA (p < 0.05). In cases where the arterialized 

veins were close to the cord (8 out of 16), the intradural 

arteries appeared well resolved on CE MRA. 

CONCLUSION 

Contrat-enhanced MRA is able to visualize intradural arteries 

and to localize the vertebral level of the AKA correctly. 

Image quality of CE MRA is sufficient for reliable detection 

of the ASA and AKA, but is still inferior to DSA due to the 

lower spatial resolution (partial volume effect) and the nonselective 

way of contrast agent administration (venous contamination). 

KEY WORDS: Adamkiewicz artery, CE MRA, validation

Poster 136 

Endovascular Treatment of Wide Neck Intracranial 

Aneurysms with Neuroform Stent and Coil 

Embolization: Assesment of the First 66 Cases 

Myers, M. E. · Madison, M. T. · Goddard, J. 


St. Paul, MN 

PURPOSE 

Wide neck intracranial aneurysms have been treated traditionally 

with surgical clipping or coiled with balloon remodeling 

technique. We wish to report our experience with using 

the neuroform stent in treating these difficult aneurysms. 


Since 2002, we have treated 66 wide neck aneurysms using 

a combination of neuroform stent placement followed by 

coil embolization. Six-month angiographic follow up was 

available in 44 cases. Angiograms were accessed for 

aneurysm occlusion and patency of the artery containing the 

stent. Operative and postoperative complications were 

recorded. 

RESULTS 

Stents were placed successfully across the neck of the 

aneurysm in all cases. Two cases required placement of a 

second stent after premature detachment of the stent proximal 

to the aneurysm. Available 6-month studies showed 

complete obliteration of the aneurysm in 34 of 44 cases with 

mild (10% or less residual filling) recanalization in 7 of the 

remaining 10 cases. Three cases demonstrated significant 

recanalization requiring additional treatment. One major 

stroke occurred within 12 hours of treatment. One patient 

had FLAIR-positive MR imaging and TIA symptoms that 

cleared 1 week posttreatment. No angiographic stenosis was 

present within or adjacent to the stent. 

CONCLUSION 

This study demonstrates stent-assisted coiling of wide neck 

aneurysms to be safe and effective in the short term. 

Complications are within the expected range and comparable 

to coiling procedure alone. The main immediate advantage 

of stent-assisted coiling is being able technically to safely 

and more efficiently pack coils in wide neck aneurysms. 

The long-term trend appears to be significantly less recanalization. 

KEY WORDS: Neuroform, stent 

Poster 137 

Detachable Fiducial Metal Marker for Treatment 

Planning of Lesions in the Bone or Vicinity 

Krol, G. S. · Lis, E. 


New York City, NY 

The conventional treatment planning of lesions of the spine 

generally rely on natural bone landmarks. More recently, a 

technique of percutaneous implantation of gold fiducial 

markings has been utilized in an effort to achieve greater 

359 

precision and reproducibility. The lengthy implantation procedure 

is performed under CT guidance and consists of several 

steps, including penetration of the bone with biopsy needles 

at minimum six sites and coaxial deposition of markers 

in subcortical bone. We designed a one-step approach, where 

the nonmagnetic, threaded tip of the needle is placed in the 

desired position and it can be detached from the shaft, thus 

eliminating all in-between steps of coaxial technique. This is 

in experimental stage and has not been used on patients, but 

it has a potential of making placement more efficient and 

shortening of operating time. 

KEY WORDS: Fiducial marker, treatment planning 

Poster 138 

Multipurpose Endovascular Operative Suite for the 

Treatment of Neurovascular Disease 

Murayama, Y. · Saguchi, T. · Ishibashi, T. · Ebara, M. · Irie, 

K. · Takao, H. · Onoue, H. · Ogawa, T. · Abe, T. 

Jikei University School of Medicine 

Tokyo, JAPAN 

PURPOSE 

In the last decade endovascular treatment of cerebrovascular 

disease has been established dramatically. Coil embolization 

for brain aneurysms has been recognized as a safe and effective 

treatment option when compared to open surgery. 

However this procedure requires general anesthesia and 

sometimes open surgery is necessary in case of hemorrhagic 

complications. For this reason we established a new concept 

of “endovascular operative suite (OR)” for the treatment of 

cerebrovascular disease. We report newly designed state-ofthe-art 

endovascular suite in the surgical operating room that 

offers integrated neurosurgical and radiologic capabilities. 


The new suite, which has 3D digital subtraction angiography 

(DSA) imaging and microsurgery capabilities, allows physicians 

to perform a wide array of neurosurgical and endovascular 

procedures. Specially designed bi-plane DSA system 

(Siemens AX; Germany) has been installed in the neurosurgery 

OR at the Jikei University Hospital. This DSA system 

consists of bi-plane C-arm and multipurpose surgical 

table with radiolucent head cramp system. The surgical table 

can rotate 120 degrees with capability of head tilt angle up to 

30 degrees and lateral tilt 15 degrees. In addition to conventional 

2D and 3D angiographic imaging, newly developed 

imaging system, “soft tissue visualization” also was 

installed. This advanced “CT-like imaging” provides bone 

and soft tissue images using the DSA C-arm system. 

Therefore postprocedural 3D CT images can be obtained by 

DSA C-arm in the operating room without moving patient. 

RESULTS 

From November 2003 to October 2004, 226 procedures were 

performed in the endovascular OR. When stroke patients 

arrive at the emergency room, they were transferred immediately 

to the endovascular OR without delay. Diagnostic 

angiography followed by endovascular procedures were performed 

by the neuro-endovascular team. When aneurysm 

shape was not appropriate for embolization, surgical procedure 

was performed without moving the patient. After surgi- 

Posters

Posters 

cal clipping, angiography is immediately performed to confirm 

location of the clips. Eight combined endovascular and 

open-surgery procedures were performed. 

CONCLUSION 

The newly designed endovascular OR provides safe and systemic 

treatment for cerebrovascular disease. 

KEY WORDS: Endovascular operative suite, embolization, 

clipping 

Poster 139 

In Vitro Validation Studies of Straight-on 

Intraaneurysmal Deployment of a Neuroform 

Microstent: A Novel Technical Application for Stent- 

Assisted Coiling of Terminus Aneurysms 

Hsu, S. 1,2 · Chaloupka, J. C. 1 · Suzuki, Y. 1 · Fujitsuka, M. 1 

1 University of Iowa Hospitals and Clinics, Iowa City, IA, 

2 Chang Gung Memorial Hospital, Kaohsiung, TAIWAN 

REPUBLIC OF CHINA 

PURPOSE 

Deployment of the Neuroform microstent (NFM) within 

acutely angled geometry, such as that encountered at the 

basilar, ICA and MCA terminuses, may result in substantial 

deformations, owing to larger than expected intersegmental 

and/or intracell gaps. This in turn can produce poor neckbridging 

scaffolding of the stent. Also, excessive rebounding 

forces may be encountered, causing difficulty in achieving 

precise positioning or stability of the stent. This led us to 

consider developing a novel alternative technique, consisting 

of intentional deployment of the NFM straight on into a terminus 

aneurysm. The purpose of this in vitro study was to 

validate this theoretical application by assessing the mechanics, 

feasibility, and efficacy of such a technique using standard 

in vitro modeling. 


Silicon models of basilar submit wide-neck aneurysms were 

constructed of variable sizes and shapes. Under an established 

closed pulsated circulation system a NFM was 

deployed straight from the proximal bifurcating vessel into 

the terminal aneurysm with at least half to two thirds of the 

stent remaining deployed within the basilar trunk. 

Simulation of coiling of the aneurysm was performed exactly 

as in a clinical setting. Digital video recording and still 

360 

photography were obtained during the procedure for analysis. 

Stability of the coil mass was tested by gross inspection, 

as well as use of a compression deformity test (comparison 

of the radial force of the stent side-wall and supporting force 

of the stent side-end). 

RESULTS 

Straight-on intraaneurysmal deployment of the NFM at basilar 

submit was technically much easier than deployment 

across basilar and PCA continuum. Very good stability of the 

coil mass could be achieved consistently by the fixation of 

the NFM end crown(s), which incarcerated the coil loops 

within the aneurysm. Digital photography of NFM-coil complex 

showed coil loops crossing the stent cells, and sitting at 

the saddle part or the turning point of the crown of NFM. 

The forces that stabilized the NFM coil complex appeared to 

derived mainly from the proximal attachment of the NFM to 

parent artery (via radial force), the longitudinal forces of the 

stent intersegmental connections (struts), and the incarcerated 

coil mass embedded into the NFM end crown(s). 

Compression deformity tests revealed a linear correlation of 

NFM side-wall radial force with the degree of compression 

deformity, but the radial force was much lower compared to 

the end-crown “en face” compression which was nonlinear. 

A potential problem with the technique is the possible locking/ensnaring 

of a coil loop into the end crown, which could 

result in inability to reposition or retrieve a partially 

deployed coil. 

CONCLUSION 

Intentionally straight on intraaneurysmal deployment of the 

NFM in wide-neck terminus aneurysms, could consistently 

achieve a very stable NFM-coil complex. Some technical 

difficulties were encountered initially during subsequent 

coiling, such as difficulty retrieving or repositioning coil 

loops. However, with some added care in proper size selection 

and slow delivery, this is a minor risk. This alternative 

technique to conventional neck bridging appears to be a 

promising strategy for endovascular surgical management of 

wide-neck terminus aneurysms. 

KEY WORDS: Neuroform microstent, in vitro, aneurysm 

Poster 140 

Device-Related Problems in Extracranial Carotid Artery 

Stenting with the Filter-Type Protection Device 

Bae Ju, K. 1 · Moon Hee, H. 1 · Oh Kee, K. 2 · Hyun Seung, K. 1 

· Sung Hyun, K. 2 

1 Seoul National University College of Medicine, Seoul, 

REPUBLIC OF KOREA, 2 Seoul National University 

College of Medicine, Bundang, REPUBLIC OF KOREA 

PURPOSE 

The filter-type protection device is being applied to carotid 

artery stenting (CAS) with an advantage of keeping the cerebral 

flow during procedure. The purpose is to show the 

device-related problems and intraprocedural findings in the 

stent-era using the filter-type protection device (PD).


Twenty-six patients with extracranial atherosclerotic carotid 

stenosis underwent CAS with PD during last 6 months. PDrelated 

technical difficulties or fluoroscopic/angiographic 

abnormalities occurred in six patients. Procedural steps, 

radiologic findings, neurologic changes, and rescue procedures 

were reviewed retrospectively. 

RESULTS 

Protection device delivery across the carotid stenosis was not 

possible due to severe carotid tortuosity in one patient. This 

was overcome with a stiff buddy wire (0.018”), which 

straightened the arterial course and made the PD cross the 

stenosis. In the remaining five, all the problems occurred 

immediately after the stent deployment. The contrast flow 

through the carotid artery was compromised in two, a filling 

defect within PD was seen in two, and the stent delivery 

catheter tip adhered to the proximal PD segment without 

separation. The first two problems were solved just by 

removing PD and the last one by a steady pullback of the 

delivery catheter together with a steady push of PD 

guidewire. Neurologic change happened only in one patient 

with the flow compromised but recovered immediately after 

PD removal. 

CONCLUSION 

A significant PD-related event in CAS may occur with the 

stent deployment. Neurointerventionists should be aware of 

and be alert to the possibility of the flow compromise, and 

should be ready to withdraw PD with care concerning the 

neurologic change. 

KEY WORDS: Carotid arteries—interventional procedures 

Poster 141 

Brain Perfusion Imaging during Temporary Balloon Test 

Occlusion with Hypotensive Challenge 

Gkogkas, C. · Baker, J. · Chen, P. · Day, A. · Frerichs, K. 

Brigham & Women’s Hospital 

Boston, MA 

PURPOSE 

Patients considered for carotid or vertebral artery sacrifice 

may still suffer neurologic injury despite successfully passing 

temporary preoperative balloon test occlusion (BTO) 

even in combination with hypotensive challenge. Our goal 

was to increase the sensitivity of BTO to predict the risk of 

perfusion-related neurologic problems following therapeutic 

vessel sacrifice. 


Conventional balloon test occlusion (BTO) under normotension 

and with hypotensive challenge was performed in 4 

patients; two patients with head and neck tumors and two 

patients with aneurysms. The balloon was inflated for a total 

of 30 minutes while continuous neurologic testing was performed. 

The test was aborted if a neurologic change compared 

with the baseline examination occurred at any time. 

After the patient passed 15 minutes of occlusion test the systolic 

blood pressure was decreased by 25%. This pressure 

was maintained for 15 minutes. Tc-99m ECD was then 

injected intravenously during balloon inflation and 5 min- 

361 

utes into the hypotensive challenge. A single-photon emission 

CT (SPECT) study then was obtained within 3 hours 

after completion of the test. 

RESULTS 

In two patients without clinical symptoms during BTO the 

brain SPECT showed symmetric cerebral perfusion. In one 

patient who clinically tolerated the balloon occlusion testing 

with the hypotensive challenge the brain SPECT revealed 

unilaterally diminished perfusion. Vessel sacrifice was not 

carried out in this patient. In one patient who developed 

aphasia towards the end of the hypotensive period the brain 

SPECT confirmed asymmetric perfusion. 

CONCLUSION 

Cerebral perfusion measured by Tc-99m ECD SPECT during 

BTO with hypotensive challenge may be able to increase 

the sensitivity of BTO. This may reduce the likelihood of 

neurologic injury following carotid or vertebral artery sacrifice. 

KEY WORDS: Balloon test occlusion, hypotensive challenge, 

brain SPECT 

Poster 142 

Modeling Endovascular Neurointervention Using 

Combined in Vitro Silicone and Ex Vivo Swine 

Vasculature 

Suzuki, Y. · Fujitsuka, M. · Chaloupka, J. C. 


Iowa City, IA 

PURPOSE 

With the rapid development and application of neuroendovascular 

intervention, there is a growing need for more 

realistic simulation of clinical conditions for the dual purpose 

of training and further research/development of 

endovascular techniques and/or technology. Unfortunately, 

mechanical-tactile simulation of device performance in artificial 

(in vitro) models is often substantially different from 

that in vivo, which has led to significant limitations in the 

utility of such models for evaluating critical performance 

characteristics of devices and techniques. This motivated us 

to consider developing a “hybrid” in vitro/ex vivo model to 

improve such simulations. 


Hybrid in vitro/ex vivo models were constructed by incorporating 

extirpated swine arteries and veins into silicone vascular 

casts. The pathologies selected for a given model were 

derived from biplanar or 3D catheter angiography, MRA, or 

CTA of actual human cases. All hybrid models then were 

placed into a standard pulsatile flow circuit and imaged by 

conventional X-ray fluoroscopy and DSA. 

RESULTS 

Aneurysm models consisting of globular, hemispherical, and 

cylinder geometries of varying size were possible form constructions 

using extirpated veins. Side branches and various 

realistic curvatures were satisfactorily reproduced as well. 

Other pathologies, such as a middle cerebral artery stenosis, 

and carotid cavernous fistula also could be created with

Posters 

excellent reproduction. A carotid embolic occlusion model 

also was created, which used tantalum impregnated thrombus 

for visualization, and some delicate adjustment of flow 

velocity to produce a good simulation. The embolus containing 

tantalum powder was delineated clearly in the image. 

Simulations of microcatheterizations, coil deployments, coil 

repositioning, stent deployment, balloon angioplasty, and 

mechanical thrombectomy all could be performed with a surprisingly 

high degree of realistic simulation compared to 

clinical experience. Carotid cavernous fistula models were 

treated either transarterially or transvenously by detachable 

balloon or coil. The transarterial route simulation was quite 

satisfactory; however the transvenous route was inadequate 

because of the friction from silicone (swine vessels were not 

set into the sinus and draining vein). 

CONCLUSION 

Incorporation of harvested arteries into silicone vascular 

models resulted in subjectively more realistic simulations of 

mechanical-tactile behavior of endovascular devices. Such a 

construct maximizes the individual advantages of in vitro 

and in vivo models, allowing for better assessments of biologic 

compatibility and performance, while still taking 

advantage of the relative strengths of in vitro models (e.g., 

lower expense, higher reproducibility, better repeatability, 

better control of experimental conditions, etc.). This hybrid 

model system holds promise for enhancing endovascular 

device technologic and/or technical evaluation and development, 

as well as possibly providing a superior platform for 

preclinical technical training. 

KEY WORDS: Animal model, silicone model, training 

Poster 143 

Radiologic Management of Brachiocephalic Arterial 

Injuries with Wallgraft Endoprostheses 

Tisnado, J. 1 · Maroney, T. P. 1 · Cirillo, R. L. 2 · Klisch, G. 1 

1 Medical College of Virginia Hospitals/Virginia 

Commonwealth University Medical Center, Richmond, VA, 

2 Chesapeake General Hospital, Chesapeake, VA 

PURPOSE 

Pseudoaneurysms and penetrating injuries to the subclavian 

arteries are uncommon lesions difficult to manage surgically. 

With the advent of metallic and covered stents, the interventional 

radiologic management of these arterial injuries is 

being accepted, thereby avoiding additional trauma and 

operative risks, and allowing a shortened hospital stay, rehabilitation, 

and the patient’s return to work. 


We present six patients (16 to 87 years old) with subclavian 

artery pseudoaneurysms secondary to blunt and penetrating 

trauma. The patients were poor surgical risks, and the trauma 

surgeons referred them to us for treatment. 

RESULTS 

The interventional radiologic treatment was successful in all 

patients. No major complications were recorded. The procedures 

were completed in less than 30 minutes. At follow up 

after several months no adverse sequela was found. Blood 

362 

pressure was equal in both arms, with no evidence of 

microemboli. One patient died of unrelated cause, 4 months 

later. Five patients are doing well 2, 6, and 18 months later. 

CONCLUSION 

This novel approach to managing traumatic injuries to the 

subclavian arteries must be considered the first choice in 

arterial (or venous) injuries, seen more often in this era of 

violence and social decay. 

KEY WORDS: Wallgraft endoprostheses, brachiocephalic, 

managing traumatic injuries 

Poster 144 

Inferior Petrosal Sinus Sampling: Clinical, Pathologic, 

and Imaging Correlation 

Fink, J. R. · Duckwiler, G. R. 

David Geffen School of Medicine at University of 

California Los Angeles 


PURPOSE 

Inferior petrosal sinus sampling (IPSS) can distinguish central 

from peripheral sources of ACTH in patients with 

Cushing’s syndrome. We reviewed all IPSS studies performed 

at our institution to assess the safety and technical 

success profile of the procedure. Inferior petrosal sinus sampling 

results were correlated with available clinical, pathologic, 

and imaging data to assess the diagnostic accuracy of 

IPSS, as well as the rate of IPSS tumor lateralization in cases 

of Cushing’s disease. 


All IPSS studies were performed under general anesthesia. 

Catheterization of the bilateral inferior petrosal sinuses was 

performed via a bilateral common femoral vein approach, 

with a larger sheath placed on one side to allow for peripheral 

blood sampling. Venography of both inferior petrosal 

sinuses was performed in all cases in order to identify variant 

anatomy, vessel size, and drainage patterns. Baseline 

sampling of cortisol and ACTH was performed, with multiple 

timed ACTH samplings following intravenous administration 

of corticotropin-releasing hormone (CRH). Centralto-peripheral 

ACTH gradients of 2:1 or greater at baseline 

and 3:1 or greater following CRH administration indicated a 

central source of ACTH production. A lateral gradient of 

greater than 1.4:1.0 was considered predictive of tumor lateralization 

in cases where a central source of ACTH was 

indicated, taking into account the pattern of venous drainage. 

Correlation of IPSS results with clinical, pathologic, and 

imaging data was performed through retrospective review of 

hospital records. 

RESULTS 

Thirty-one IPSS studies were performed between January 

1998 and November 2004. Inferior petrosal sinus sampling 

was technically successful in all cases, without complications. 

Clinical, pathologic, and imaging data were available 

for correlation with IPSS results in seventeen cases, wherein 

IPSS had a sensitivity of 92%, specificity of 60%, and overall 

accuracy of 82% for the diagnosis of Cushing’s disease, 

as compared to 42%, 80%, and 53%, respectively, for MR

imaging. In cases where the pre-IPSS diagnosis of Cushing’s 

syndrome was unequivocal, the specificity of IPSS was 

100%. In 11 cases of Cushing’s disease, IPSS correctly lateralized 

the tumor in seven cases (64%), as compared to five 

(45%) cases correctly localized by MR imaging. 

CONCLUSION 

Inferior petrosal sinus sampling is a safe, relatively simple, 

and clinically useful procedure with high diagnostic accuracy 

in the diagnosis of central and peripheral sources of 

ACTH elevation in patients with Cushing’s syndrome. In 

cases of Cushing’s disease, IPSS may often be useful in 

presurgical planning for preoperative tumor lateralization. 

The diagnostic specificity of IPSS is highly dependent upon 

appropriate patient referral. 

KEY WORDS: Inferior petrosal sinus, pituitary, Cushing’s 

disease 

Poster 145 

Noninvasive Optimal Vessel Analysis of a Superficial 

Temporal Artery to Middle Cerebral Artery Bypass 

Graft 

Miyakoshi, A. 1 · Lefton, D. R. 2 · Langer, D. J. 2 

1 Beth Israel Medical Center, New York, NY, 2 St. Luke’s 

Roosevelt Hospital Center, New York, NY 

PURPOSE 

Noninvasive optimal vessel analysis (NOVA) is an emerging 

technology which uses phase imaging to calculate blood 

flow. We assessed this technique’s ability to measure blood 

flow in milliliter per minute through a superficial temporal 

artery to middle cerebral artery bypass graft. 


A 60-year-old male with right middle cerebral artery (MCA) 

occlusion underwent right superficial temporal artery to 

MCA bypass. Three dimensional time of flight MR angiography 

was performed postoperatively and NOVA was utilized 

to assess quantitative flow in mL/min through the 

extracranial and intracranial portions of the graft. 

363 

RESULTS 

Extracranial and intracranial flow through the graft measured 

84 mL/min and 85 mL/min, respectively. Measured 

flow through the left MCA was 99 mL/min with normal 

MCA flow reported to be in the range of 110-210 mL/min. 

CONCLUSION 

Noninvasive optimal vessel analysis is a viable method to 

assess blood flow through a superficial temporal artery to 

MCA bypass graft. 

KEY WORDS: NOVA, bypass graft, MR angiography 

Poster 146 

Intraarterial Thrombolysis for Acute Embolic Stroke: 

Analysis of 3-Year Experience at a Regional Stroke 

Center 

Levy, R. A. 1 · Herm, R. 2 · Clark, H. 2 

1 Saint Mary’s Hospital, Field Neurosciences Institute, 

Saginaw, MI, 2 Field Neurosciences Institute, Saginaw, MI 

PURPOSE 

To assess parameters related to neurologic outcomes in 

patients treated with intraarterial thrombolysis for acute 

embolic stroke. 


Medical records from 36 patients enrolled in the INSTORE 

(Interventional Stroke Outcomes Registry) at our institution 

from November 2001 to November 2004 were reviewed retrospectively 

by a CAQ neuroradiologist and two research 

nurses. Twenty-six patients (72%) had 3-month postthrombolysis 

follow up and were included in this study. 

Unenhanced head CT to assess ischemic changes and 

exclude intracranial hemorrhage was obtained prior to 

angiography on all patients. All patients had transfemoral 

diagnostic and therapeutic catheter angiography performed 

by an experienced neuroradiologist using institutional guidelines 

and informed consent. A small number of early participants 

received intraprocedural heparinization. An average 

intraarterial TPA (tissue plasminogen activator) dose of 7 

mg/patient was used. Later in the study, mechanical thrombolysis 

(including Tracker catheter and guidewire manipulation, 

thromboaspiration, and balloon angioplasty) was utilized 

with TPA. Parameters assessed were 3-month postthrombolysis 

Rankin score, time to treat following symptom 

onset, arterial segment occluded and extent of arterial 

recanalization. Rankin scores were grouped from 0-3 and 4- 

6 with 6 assigned to patients who died. Time to treat was 

divided into 0-3 and 3-6 hour groups. Arterial distributions 

were ICA (internal carotid artery), MCA (middle cerebral 

Posters

Posters 

artery) M1, and MCA M2. Postprocedural arterial recanalization 

was described as complete/near complete, partial, 

and none/minimal. These categories were used in statistical 

analysis of variance with 3-month Rankin score as the 

dependent variable. 

RESULTS 

There were no results achieving statistical significance. Of 

the 26 patients, 61.5% had 3-month Rankin scores 0-3. Of 

these 16 patients, 25% were in the 0-3 hour time to treat 

group. Of the remaining 10 patients, 60% were in the 0-3 

hour time to treat group. Forty-two percent of patients 

achieved complete/near complete recanalization and 38% 

achieved partial recanalization. Eighty-two percent of 

patients with complete/near complete recanalization and 

30% of patients with partial recanalization had 3-month 

Rankin scores 0-3. Thirty-one percent of patients had ICA, 

42% had MCA M1, and 27% had MCA M2 occlusions. 

Thirty-eight percent of ICA, 64% of MCA M1, and 86% of 

MCA M2 occlusions had 3-month Rankin scores 0-3. 

Thirteen percent of ICA occlusions achieved complete/near 

complete and 50% achieved partial recanalization. 

Corresponding values for the MCA M1 occlusions were 45% 

and 45%, and for the MCA M2 occlusions, 71% and 14%. 

The single ICA occlusion with complete/near complete 

recanalization had a 3-month Rankin score of 0-3. Eighty 

percent of MCA M1 and 80% of MCA M2 occlusions with 

complete/near complete recanalization had 3-month Rankin 

score 0-3. No patient suffered neurologically significant 

intracranial hemorrhage. 

CONCLUSION 

The completeness of arterial recanalization postthrombolysis 

and location of arterial occlusion are more important determinants 

of neurologic outcome than time to treat. Middle 

cerebral M2 occlusions have the best, M1 occlusions have 

intermediate and ICA occlusions have the worst neurologic 

outcome. Further validation is anticipated with ongoing 

investigation. 

KEY WORDS: Stroke, thrombolysis, intraarterial 

Poster 147 

Relationship between Severity of Arterial Occlusion and 

Early Ischemic Changes on CT Scan 

Mohammad, Y. M. · Christoforidis, G. A. · Bourekas, E. C. · 

Slivka, A. P. 

Ohio State University Medical Center 

Columbus, OH 

PURPOSE 

To determine the relationship between early ischemic 

changes on initial CT scan and severity of arterial occlusion 

in patients undergoing intraarterial thrombolysis within 6 

hours of symptom onset. Ischemic changes identified on CT 

scans represent early cytotoxic edema and development of 

irreversible injury. Patients with more proximal occlusions 

without collaterals may be more prone to ischemic injury. 

The Qureshi grading scheme has been proposed to evaluate 

the severity of arterial occlusion in acute ischemic stroke 

after accounting for occlusion location or collateral circulation. 

364 


The Qureshi grading scheme assigns a score from 0 to 5 on 

the basis of occlusion site and collateral supply. We determined 

the relationship between early ischemic changes on 

CT scan and the Qureshi grading scale assessed from initial 

angiography (by a neuroradiologist blinded to the clinical 

examination) patients who underwent intraarterial therapy 

for acute ischemic stroke within 6 hours of symptom onset. 

Early CT changes were identified based on presence of tissue 

hypodensity, effacement of sulci, hypodensity of 

lentiform, loss of insular ribbon sign, and middle cerebral 

artery hyperdense sign. 

RESULTS 

We evaluated the relationship in 35 patients (mean age 65 ± 

14 years; 19 were men). The patients were treated with prourokinase 

(n = 2), urokinase (n = 3), and tissue plasminogen 

activator (n = 30). The mean time interval between symptom 

onset and treatment was 263 ± 71 minutes. The initial occlusion 

was categorized as grade I (n = 2), grade II (n = 2), 

grade IIIA (n = 13), grade IIIB (n = 6), grade 4A (n = 2), and 

grade 4B (n = 2). Early ischemic changes on initial CT scan 

were observed in 10 (29%) of the 35 patients. There was a 

positive correlation with initial severity of arterial occlusion 

and early ischemic changes (p = 0.001). Among patients with 

middle cerebral artery M1 occlusion, early ischemic changes 

were observed in none of the 13 patients with adequate collaterals 

(grade 3A) and 5 of the 6 patients without adequate 

collaterals (grade 3B) (p < 0.026). 

CONCLUSION 

The results support the concept of variable time window for 

onset of ischemic injury among patients with ischemic stroke 

that can be predicted based on initial site of occlusion and 

presence or absence of collaterals. 

KEY WORDS: Collaterals, thrombolysis, Qureshi grading 

scheme 

Poster 148 

Early Loss of Cortical Density on Admission CT Scan Is 

Associated with Poor Pial Collaterals 

Mohammad, Y. M. 1 · Christoforidis, G. A. 1 · Atieh, J. M. 2 · 

Bourekas, E. C. 1 · Slivka, A. P. 1 

1 Ohio State University Medical Center, Columbus, OH, 

2 Columbus State University Community College, Columbus, 

OH 

PURPOSE 

Patients with good pial collaterals on admission angiogram 

have better functional outcome when treated with intraarterial 

thrombolysis within 6 hours from stroke onset. We conducted 

this study to evaluate the extent of pial collaterals in 

the patients presenting with early loss of cortical density on 

admission CT scan. 


This is a retrospective study conducted on the patients who 

received intraarterial thrombolysis within 6 hours of stroke 

onset at Ohio State University between November 1993 and 

April 2004. Loss of cortical density was graded on a 0 to 2 

scale. 0 indicated no evidence of early loss of cortical densi-

ty. 1 indicated subtle evidence of early loss of cortical density 

whereas 2 indicated obvious evidence of early loss of cortical 

density on the admission CT scan. Pial collaterals were 

graded on a 1 to 5 scale depending on the extent of reconstitution 

of occluded vasculature on delayed angiogram. We 

further categorized the pial collaterals to either poor or good 

pial collaterals. 

RESULTS 

The admission CT scan and angiogram were evaluated in 50 

patients. Seventy-eight percent of those patients with no evidence 

of early loss of cortical density have good pial collaterals. 

Only 18% of the patients with subtle evidence of early 

loss of cortical density and 8% with obvious evidence of 

early loss of cortical density have good pial collaterals (P = 

0.0390). 

CONCLUSION 

Early loss of cortical density on admission CT scan is associated 

with poor pial collaterals. A simple admission CT scan 

is extremely important in selecting the acute stroke patients 

whose ischemia can be potentially reversible by urgent 

recanalization. 

KEY WORDS: Cortical density, pial collaterals, CT scan 

Pediatrics 

149-172 

Poster 149 

Cerebral Venous Thrombosis: A Mimic of Nonaccidental 

Injury 

Barnes, P. D. 

Lucile Salter Packard Children’s Hospital 

Palo Alto, CA 

PURPOSE 

To describe the findings in children with cerebral venous 

thrombosis (CVT) who present as a mimic of alleged nonaccidental 

injury (NAI). 


The clinical, imaging, and pathologic findings of 7 children 

with confirmed CVT (4 females, 3 males; ages 6 weeks - 23 

months) who were alleged victims of NAI were reviewed. 

CT was done in 7 and MR imaging in 5. 

RESULTS 

Clinically, there was prematurity in 2, infectious illness in 7, 

recent vaccination in 2, acute live threatening event in 7, 

seizures in 3, retinal hemorrhages in 7, and abnormal hematology 

in 7. By CT and MR imaging there was subarachnoid 

or subdural hemorrhages/thromboses in 7; cerebral hemorrhages 

/ infarctions in 7; ventriculomegaly in 3; and, intraventricular 

hemorrhage in 3. The skeletal surverys were negative 

in 7. There were 5 deaths and autopsy showed venous 

thrombosis in 7, hypoxia-ischemia in 6, inflammation in 2, 

and trauma in 0. 

365 

CONCLUSION 

The clinical and imaging findings of CVT may mimic the 

findings otherwise associated with NAI. These and other 

reports question the specificity of the traditional clinical and 

imaging criteria for NAI, including the shaken baby syndrome. 

KEY WORDS: Venous thrombosis, nonaccidental injury, 

child abuse 

Poster 150 

Leptomeningeal Ivy Sign in Childhood Moyamoya 

Disease: Findings on Susceptibility-Weighted Imaging 

Kim, E. 1,2 · Kim, S. 3 · Yoon, H. 1 · Roh, H. 1 · Na, D. 4 · Ryoo, 

J. 5 · Kim, H. 1 

1 Samsung Medical Center, Seoul, REPUBLIC OF KOREA, 

2 Yonsei University College of Medicine, Seoul, REPUBLIC 

OF KOREA, 3 Kangwon National University Hospital, 

Chunchon, REPUBLIC OF KOREA, 4 Seoul National 

University Hospital, Seoul, REPUBLIC OF KOREA, 

5 Gyeongsang National University Hospital, Jinju, 

REPUBLIC OF KOREA 

PURPOSE 

To describe the findings of leptomeningeal ivy sign in children 

with moyamoya disease on susceptibility-weighted MR 

imaging. 


Seven patients with angiographically confirmed moyamoya 

disease were included in our study. Fast FLAIR, susceptibility-weighted 

imaging, and gadolinium-enhanced T1-weighted 

images were obtained by using a 1.5 T system. SPECT 

was also available in all patients. The parameters of susceptibility-weighted 

imaging were as follows: repetition time, 

100 ms; echo time, 60 ms; flip angle, 20°; matrix, 512 × 512; 

slice thickness, 2 mm (1 mm after interpolation). 

Susceptibility-weighted images obtained by using both magnitude 

and phase images were transformed into minimumintensity 

projection images of 5 mm thickness and compared 

with fast FLAIR, gadolinium T1-weighted imaging, and 

SPECT after coregistration. 

RESULTS 

All patients showed ivy sign on both fast FLAIR and T1weighted 

imaging. Part of ivy sign on both fast FLAIR and 

gadolinium T1-weighted imaging showed dark-signal intensity 

on susceptibility-weighted imaging, representing an 

increased concentration of deoxyhemoglobin in the leptomeningeal 

collateral vessels. The rest of ivy sign appeared 

as iso- or high-signal intensity, suggesting relatively higher 

oxygen tension. The areas of decrease in perfusion on 

SPECT partly corresponded to the areas near the dark-signal 

intensity vessels on susceptibility-weighted imaging. 

CONCLUSION 

Ivy sign on gadolinium T1-weighted imaging and fast 

FLAIR was depicted as both dark signal and iso- or high-signal 

intensity vessels on susceptibility-weighted imaging, 

suggesting variable oxygen tension in the leptomeningeal 

collateral vessels. Partial correlation was noted between the 

Posters

Posters 

areas near the dark-signal intensity vessels on susceptibilityweighted 

imaging and areas of decreased perfusion on 

SPECT. 

KEY WORDS: Moyamoya disease, Cerebral blood vessels— 

MR imaging, Cerebral blood vessels—stenosis or obstruction 

Poster 151 

Occult Brain Microhemorrhages Associated with 

Cardiopulmonary Bypass in Some Infants and Children 

Thomas, A. K. · Jacobson, J. P. · Holshouser, B. · Bailey, L. 

L. · Kido, D. K. 

Loma Linda University School of Medicine 

Loma Linda, CA 

PURPOSE 

To report the detection of otherwise occult brain microhemorrhages 

in some infants and children following cardiopulmonary 

bypass using a relatively novel, high spatial resolution, 

susceptibility-weighted MR imaging technique (SWI). 


The SWI technique uses 3D gradient-echo acquisitions covering 

the entire brain to create both magnitude and phase 

images, which are postprocessed to produce strongly susceptibility-weighted 

images (1). The underlying contrast 

mechanism relates to the magnetic susceptibility difference 

between oxygenated and deoxygenated hemoglobin. 

Susceptibility-weighted imaging has been shown to detect a 

larger number and volume of hemorrhagic lesions than conventional 

2D gradient-echo images in children and adolescents 

with posttraumatic diffuse axonal injury (2, 3). An 

IRB-approved review was performed of routine SWI in pediatric 

patients with suspected brain injury from January 2001 

to mid-November 2004. A subset of atraumatic patients with 

brain microhemorrhages on SWI was identified. 

Susceptibility-weighted images were compared to conventional 

MR and CT images and the patients’ medical records 

were reviewed retrospectively. 

RESULTS 

Brain microhemorrhages were detected in 4 patients (mean 

age 43 months, range 2 months -10 years), especially near 

the gray-white junction. These were not visible on conventional 

MR sequences or on CT. All of the patients previously 

had undergone cardiopulmonary bypass (mean elapsed 

time 10.5 months, range 1 month - 29 months) (Fig.). 

366 

CONCLUSION 

Susceptibility-weighted imaging detects otherwise occult 

brain microhemorrhages, a rare finding in the atraumatic 

pediatric population. Cardiopulmonary bypass appears to be 

associated with brain microhemorrhages in some infants and 

children. 

REFERENCES 

1. Reichenbach JR, Venkatesan R, Schillinger DJ, Kido DK, 

Haacke EM. Small vessels in the human brain: MR venography 

with deoxyhemoglobin as an intrinsic contrast agent. 

Radiology 1997;204:272-277 

2. Tong KA, Ashwal SA, Holshouser BA, et al. Hemorrhagic 

shearing lesions in children and adolescents with posttraumatic 

diffuse axonal injury: Improved detection and initial 

results. Radiology 2003;227:332-339 

3. Tong KA, Ashwal SA, Holshouser BA, et al. Diffuse axonal 

injury in children: clinical correlation with MRI hemorrhagic 

lesions. Ann Neurol 2004;56:36-50 

KEY WORDS: Brain microhemorrhage, cardiopulmonary 

bypass, susceptibility-weighted imaging 

Poster 152 

Cognitive Impairment in Sickle Cell Disease: A Diffusion 

Tensor Imaging Study 

Moinuddin, A. 1 · Snyder, A. Z. 1 · McKinstry, R. C. 1 · White, 

D. A. 2 · DeBaun, M. R. 2 

1 Mallinckrodt Institute of Radiology, St. Louis, MO, 

2 Washington University, St. Louis, MO 

PURPOSE 

Children with Sickle cell disease (SCD) often experience 

difficulties in school compared with healthy siblings. Some 

of these problems can be attributed to overt and/or “silent” 

cerebral infarction. However, some SCD patients have no 

MR signs of infarction yet they have IQs below 85. We

hypothesize that diffusion tensor imaging (DTI) will detect 

microstructural changes in the white matter in SCD patients 

with low IQ and normal conventional MR scans. 


At our institution, patients with SCD are screened with MR 

imaging and IQ testing. Permission to review their data retrospectively 

was granted by our IRB. Seven patients with 

SCD complicated by infarction, and 17 patients with SCD 

and normal MR imaging were studied. Standard 6-direction 

DTI encoding was performed with 2 acquisitions. The diffusion 

tensor was computed after motion correction and transformation 

to the standard atlas space. Mean diffusivity (MD) 

and relative anisotropy (RA) were computed for regions of 

normal-appearing white matter (NAWM) and for the regions 

of infarction. The values of MD and RA in patients with and 

without infarction were compared using a t test. The relationship 

between IQ and the DTI metrics was tested using a 

t test and a Mann-Whitney U test. 

RESULTS 

The figure shows a FLAIR image of a patient with silent 

infarction. The arrow shows a cluster of signal abnormalities 

in the right frontal lobe. This patient was experiencing difficulties 

in school but had no evidence for overt infarction. 

Difference in MD between regions of infarction and NAWM 

were statistically significant for all regions. Differences in 

RA between NAWM and regions of infarction reached statistical 

significance in only 7 of 12 regions measured. DTI 

was not statistically different in patients with normal IQ and 

for IQ below 85. 

CONCLUSION 

Diffusion tensor imaging is no more predictive of low IQ in 

SCD than routine MR imaging. Mean diffusivity performs as 

well as FLAIR and T2-weighted MR imaging for detection 

of abnormal white matter. Relative anisotropy, because of its 

367 

inherent noise sensitivity, is less sensitive than FLAIR and 

MD for the detection of such brain injury. In summary, the 

combination of FLAIR, T2-weighted and DTI scans 

obtained in a typical clinical MR protocol cannot explain the 

neurocognitive impairment that affect some patients with 

SCD. Mean diffusivity performs as well as FLAIR and T2weighted 

scans for the detection of brain injury in SCD. 

Relative anisotropy is not suitable for routine evaluation of 

NAWM in SCD with low IQ. 

KEY WORDS: Diffusion, anisotropy, IQ 

Poster 153 

Leukoencephalopathy: An Unusual Injury Pattern in 

Infantile Hypoxia-Ischemia 

Barnes, P. D. · Lertvananurak, R. · Hahn, J. · DiDomenico, P. 

Lucile Salter Packard Children’s Hospital 

Palo Alto, CA 

PURPOSE 

To present the imaging findings in two infants with 

acute/subacute leukoencephalopathy due to hypoxiaischemia. 


The medical records, EEGs, and brain imaging (CT, MR 

imaging) were reviewed in two infants with hypoxicischemic 

leukoencephalopathy, a 6-month-old male with 

head trauma complicated by suffocation and a 15-month-old 

female with near-drowning. 

RESULTS 

For the first child, the initial CT and MR imaging within the 

first 24 hours were normal. For the second child, the CTs at 

8 and 36 hours were normal. The followup MR images at 

day four (child 1) and day five (child 2), showed extensive 

cerebral white matter injury with restricted diffusion. The 

EEGs in both patients showed diffuse slowing but no seizure 

activity. Both children had poor neurodevelopmental outcomes. 

CONCLUSION 

Predominant, or exclusive, white matter injury is an unusual 

manifestation of hypoxic-ischemic brain injury in the acuteto-subacute 

phase. This implies a primary axonal injury 

(necrosis, apoptosis) rather than a secondary axonal injury 

(i.e., wallerian), or indicates a primary glial injury (e.g., 

oligodendroglial injury or myelinopathy). Such injury may 

be the result of hypoxia rather than hypoperfusion. 

KEY WORDS: Leukoencephalopathy, hypoxia-ischemia, 

pediatric 

Posters

Posters 

368 

Poster 154 

Poster 155 

Neuroradiologic Correlates of Clinical Disability and Menkes’ Disease: A Neuroradiologic Essay from Onset to 

Progression in Pelizaeus-Merzbacher Disease 

Death 

Laukka, J. · Makki, M. · Lewis, R. · Shy, M. · Kamholz, J. · 

Bernardi, B. · Garbern, J. Y. 


Detroit, MI 

PURPOSE 

To determine whether quantitative measures of MR imaging 

data from patients with the inherited leukodystrophy 

Pelizaeus-Merzbacher disease (PMD) correlate with clinical 

disease severity or progression. Reliable, noninvasive quantitative 

biomarkers of disease activity will be useful not only 

for following the natural history of disease, but also potentially 

for evaluating future therapies. 


Pelizaeus-Merzbacher disease is caused by mutations of the 

proteolipid protein 1 (PLP1) gene. The initial severity of 

PLP1 mutations appears to correlate with the degree to 

which they cause apoptosis of oligodendrocytes, the myelinating 

cells of the central nervous system (CNS). We speculate 

that late disease progression correlates with the degree 

of axonal injury, which we have demonstrated in patients 

with PLP1 null mutations. To extend these studies to patients 

with other PLP1 mutations, we analyzed the brains of 34 

PMD patients by MR imaging. Twenty-five of these patients 

had a variety of PLP1 point mutations, mostly missense substitutions, 

and 9 had PLP1 duplications. For each patient we 

measured, white matter volume (WMV), sagittal corpus callosum 

area, fractional anisotropy in several white matter 

areas, white matter N-acetyl aspartic (NAA) acid levels (a 

biomarker for axonal integrity), and intercaudate distance 

(ICD). The MR data were correlated with functional disability 

scores (FDS) using a system we developed for clinical 

evaluation of PMD patients. 

RESULTS 

Comparison of the MR measurements and the FDS demonstrates 

that white matter volume inversely correlates with 

functional disability, suggesting that the initial disability 

does correlate with the extent of myelination. The intercaudate 

distance and sagittal corpus callosum area also correlate 

with both WMV and the FDS, and may usefully substitute 

when gray-white matter segmentation is not possible. Our 

preliminary results support the prediction that disease progression 

rates correlate with the degree of axonal loss, as 

inferred from reduction in white matter NAA decline and 

reduction in white matter tract anisotropy. 

CONCLUSION 

Our data demonstrate that white matter atrophy is a significant 

early determinant of neurologic disability in PMD 

patients while axonal dysfunction or loss correlates with late 

clinical progression. These observations appear to hold 

regardless of mutation type. These data support the use of 

MR imaging as a clinical tool to follow the natural history of 

PMD, and potentially for its application in evaluation of 

future therapies. 

KEY WORDS: Pelizaeus-Merzbacher disease, segmentation, 

MR spectroscopy 

Mortilla, M. 1 · Procopio, E. 1 · Donati, M. A. 1 · Pasquini, E. 1 · 

Domenici, R. 2 · Fonda, C. 1 

1 2 Children’s Hospital A. Meyer, Florence, ITALY, Lucca 

Hospital, Lucca, ITALY 

PURPOSE 

To evaluate radiologic findings in children affected with 

Menkes’ disease from diagnosis to terminal stages of the disorder. 


The brains of 2 affected children (both male) were evaluated 

serially with a 1.5 T scanner (27 mT/m) with conventional 

MR imaging sequences, SS-EPI diffusion imaging, and proton 

MR spectroscopy. X-rays of skeletal segments and serial 

abdominal ultrasounds also were performed. Case 1: G.B. 

came to our observation after diagnosis in another hospital. 

He presented typical facies, kinky hair, developmental delay, 

and seizures. Molecular analysis revealed a mutation 

g.3013G>A. He was treated with copper histidinate 3.5 

mg/die. At the age of 5 months he underwent a skeletal Xray 

that showed diffuse demineralization and abnormal long 

bones metaphisis. Serial MR imaging (at 5, 12, and 30 

months) showed areas of parenchymal damage predominantly 

localized at temporal lobes and progressive atrophy confirmed 

by a progressive decrease of NAA/Cr ratio. 

Abdominal ultrasound showed an increased volume of both 

kidneys associated to medullary hyperechogenicity (nephrocalcinosis-like) 

and progressive decrease of cortical thickness. 

Case 2: G.N.D. came to our observation because of 

hypotonia, feeding problems, and weight loss. Diagnosis 

was made because of typical fenotype and laboratory tests. 

Molecular analysis revealed a mutation c.2956C>T. He was 

treated with copper histidinate (3.5 mg/die). The MR imaging 

performed at the age of 3 months showed a right basal 

ganglia alteration with increased subarachnoid spaces, while 

at the age of 8 months a diffuse hyperintensity of white matter 

and atrophy were evident. Ultrasound showed a 

hypoplastic right kidney and a hypertrophic left kidney. At 1 

year he came to the emergency unit for a microematuria and 

underwent a chest X-ray that showed a diffuse interstitial 

thickening. The day after a sudden death occurred after a 

brief cyanosis. 

RESULTS 

Although clinical features like hair kinking and laboratory 

findings (low ceruloplasmin and copper serum levels) lead to 

the diagnosis of Menkes’ disease, radiologic techniques 

show characteristic patterns of this disorder. 

CONCLUSION 

Despite the availability of a pharmaceutical treatment that 

allows a normalization of laboratory tests, the disease is progressive 

and it is possible to monitor cerebral deterioration 

with serial neuroimaging. 

KEY WORDS: Menkes’ disease, MR imaging, pediatric brain

Poster 156 

Evaluation of Anterior Pituitary Gland Volume in 

Childhood Using 3D MR Imaging 

Marziali, S. · Garaci, F. G. · Gaudiello, F. · Ludovici, A. · 

Floris, R. · Simonetti, G. 


Rome, ITALY 

PURPOSE 

Inaccurate volumetric measurement of normal anterior pituitary 

gland in childhood has been performed in the past by 

2D MR imaging. Three-dimensional MR imaging is a reliable 

tool in the evaluation of anatomical volumes. The aim 

of this study was to assess an accurate normal anterior volume 

of pituitary gland in childhood using 3D MR imaging. 


Ninety-five prepubertal patients (range 2 months - 10 years) 

with clinically normal pituitary function and no pituitary or 

brain abnormalities were recruited. Normal volume of the 

anterior pituitary gland was measured using coronal T1weighted 

3D gradient-echo sequences (section thickness: 

0.75 mm). A measurement error of 0.2-0.4% was assessed by 

using a phantom study. 

RESULTS 

Volumetric data and progressive growth of normal anterior 

pituitary gland are reported: a) 131 ± 24 mm 3 at 2-12 months; 

b) 249 ± 25 mm 3 at 1-4 years; c) 271 ± 29 mm 3 at 5-10 years. 

CONCLUSION 

Normal anterior pituitary volume was assessed by 3D MR 

imaging coronal sections. This data may be of value for pediatricians 

in the evaluation of patients with neuroendocrine 

diseases in particular growth hormone deficiency. 

KEY WORDS: Anterior pituitary gland volume 

Poster 157 

Abnormalities of the Fornix in Septo-Optic Dysplasia: 

Spectrum of Findings 

Decarie, J. 

Hopital Ste-Justine 


PURPOSE 

Septo-optic dysplasia is a congenital anomaly associating 

optic nerve hypoplasia and absence of septum pellucidum. It 

can be associated also with hypothalamic-pituitary anomalies 

and abnormal cortical development leading sometimes 

to schizencephaly. The normal septum pellucidum is a membrane 

stretched from the inferior aspect of the corpus callosum 

to the fornix. An abnormal development of the fornix is 

probably the basic malformation is this entity. The purpose 

of this study is to illustrate the spectrum of forniceal abnormalities 

in septo-optic dysplasia. 

369 


The imaging findings on sagittal T1-weighted images were 

analyzed retrospectively in patients with known septo-optic 

dysplasia. The fornix was identified on parasagittal images. 

RESULTS 

The normal fornix makes an arch with an upper convexity to 

join the inferior surface of the corpus callosum. In septooptic 

dysplasia, the fornix showed several patterns: 1. 

Horizontal; 2. Inferior convexity; 3. The so-called fornicealcallosal 

continuity; 4. Incomplete upper convexity. 

CONCLUSION 

When making a diagnosis of septo-optic dysplasia, one must 

analyze the fornix with care. In secondary and acquired 

defects of the septum pellucidum, the fornix will be normal 

enabling one to distinguish these types of defect from septooptic 

dysplasia. 

KEY WORDS: Septo-optic dysplasia, fornix, congenital 

Poster 158 

An Unusual Case of Split Cord Malformation 

Moriya, J. · Kakeda, S. · Ohnari, N. · Hatakeyama, Y. · 

Korogi, Y. · Hashimoto, M. · Yokota, A. 

University of Occupational and Environmental Health 


Kitakyushu, JAPAN 

PURPOSE 

Split cord malformations (SCMs) are rare spinal anomalies, 

and are classified as one of two types according to unified 

theory; in type I, the hemicords are invested with individual 

dural sacs and medial walls of the sacs ensheath a rigid 

(bony or cartilaginous) midline spur, while in type II, the 

hemicords are within a single dural sac and the midline septum 

is composed of nonrigid fibrous or fibrovascular tissues. 

We will report radiologic and surgical features of an unusual 

case of SCM; it was not able to classify this case according 

to the unified theory. 


A 1.2 kg genetically female newborn was delivered by 

cesarean section at 36 weeks of gestation, and motor dysfunction 

of her small left lower extremity was present at 

birth. Pregnancy and family history were unremarkable. She 

was diagnosed with SCM and presented with a 5-year history 

of worsening neurologic dysfunction such as spastic 

hemiplegia. MR imaging revealed a vertebral anomaly at 

levels of T12-L5, a SCM below the level of T12, and a tethered 

cord at levels of L4-L5. The left hemicord had no apparent 

intradual connection to the upper cord, and an extradural 

neural tract through a canal in the deformed vertebra was 

suspected. 

RESULTS 

The operation was done for relief of the tethered cord, and it 

was confirmed that left hemicord preserved electrophysiologic 

function. 

Posters

Posters 

CONCLUSION 

To our knowledge, such a case has not been reported in the 

literature, and a new classification of SCM may be necessary. 

KEY WORDS: Split cord malformation 

Poster 159 

Hippocampal Malrotation in Patients with Epilepsy 

Kesavadas, C. · Thomas, B. · Gupta, A. · Kapilamoorthy, T. 

· Krishnamoorthy, T. 

Sree Chitra Tirunal Institute for Medical Sciences and 

Technology 

Trivandrum, INDIA 

PURPOSE 

Hippocampal malrotation recently has been reported to 

occur in patients with medically refractory epilepsy (MRE), 

comprising up to 6% of patients. Better understanding of this 

entity may help in identifying structural abnormalities in 

more patients, who were previously labeled as “MR imaging 

negative” cases of MRE. 


Review of 204 MR studies of patients with MRE revealed 12 

patients with hippocampal malrotation. The clinical history 

(family history, pre and perinatal history, age at onset of 

epilepsy, type of seizures, and epilepsy syndrome) were correlated 

with the imaging findings (side of abnormality, size 

of hippocampus, T2 signal change, internal architecture, 

angle of collateral sulcus, position of fornix, sizes of fornix, 

370 

temporal horn and temporal lobe, and presence of other 

pathology such as malformations of cortical development). 

The histopathologic findings were available in 2 patients. 

RESULTS 

The mean age of the 12 patients (8 male, 3 female) was 15 

years. Mean age at onset of seizures was 4 years. Two 

patients had developmental delay. Hippocampal malformations 

were noted bilaterally in 7 and unilaterally in 5 

patients. Of the 19 hippocampi studied, the abnormal configuration 

of hippocampus was located in the whole of the 

hippocampus in 19, hippocampal head in 4, body in 6. 

Atrophy of temporal lobe and cerebral hemisphere was noted 

in 9 and 5 sides, respectively. Associated anomalies included 

bluring of gray-white matter junction (9), gyral abnormalities 

(4), subependymal heterotopics (2), focal hypoplasia 

of corpus callosum (1) and cerebellar hypoplasia (1). 

Good correlation was seen between severity of hippocampal 

abnormalities and degree of temporal lobe and hemispheric 

white matter changes. In the histopathologic diagnosis was 

cortical dysgenesis in one of the patient. 

CONCLUSION 

Hippocampal malrotation may represent a recognizable 

manifestation of a more generalized developmental disorder 

that cannot be identified by current MR imaging sequences. 

Further studies on imaging - EEG correlation and results of 

surgery are needed for better understanding of this entity. 

Take Home Message: Closer observation of the hippocampal 

formation, with attention towards abnormal orientation may 

reveal underlying temporal lobe developmental malformations. 

KEY WORDS: Epilepsy, hippocampus 

Poster 160 

Focal Cortical Dysplasias: MR Imaging and 

Histopathologic Correlation 

Thomas, B. · Radhakrishnan, V. · Kesavadas, C. · Gupta, A. 

· Mathew, A. · Radhakrishnan, K. 

Sree Chitra Tirunal Institute for Medical Sciences and 

Technology 

Trivandrum, INDIA 

PURPOSE 

MR imaging is a well recognized technique for detecting 

focal cortical dysplasias (FCD). A large majority of patients 

with FCD present with focal seizures. This study aims at 

analyzing and describing the various MR findings in FCD 

and correlating them with histopathology of the resected 

specimen. 


Retrospective analysis of MR findings in 14 seizure patients 

with FCD was performed. In patients with focal cortical dysplasias 

who were operated radiologic/pathologic correlation 

was attempted. MR findings looked for included cortical 

thickening, cortical and/or subcortical hyperintensity, abnormal 

widened/deep sulci, and radial hyperintense band 

extending to the ventricle. Each of these findings was correlated 

with histopathologic features.

RESULTS 

A correlation was attempted in both FCD-Taylor’s balloon 

cell type and the nonballoon cell type. Histopathologically, 

the subcortical zone of the FCD-balloon cell displayed 

hypomyelinated white matter with radially oriented balloon 

cells and gliosis. Dysplastic neurons were found in the adjacent, 

disorganized cortex. 

CONCLUSION 

MR imaging recognition of focal cortical dysplasias is 

important since the surgical outcome following resective 

surgery is favorable. Knowledge of MR characteristics of 

FCD and the optimal techniques used in picking up these 

lesions is essential in the diagnosis. 

KEY WORDS: Cortical dysplasia, epilepsy, MR imaging 

Poster 161 

MR Imaging Features of Central Nervous System 

Involvement of Hemophagocytic Lymphohistiocytosis in 

Children: 15 Consecutive Cases 

Weon, Y. 1 · Goo, H. 2 · Choi, S. 3 

1Samsung Medical Center, Seoul, REPUBLIC OF KOREA, 

2Asan Medical Center, Seoul, REPUBLIC OF KOREA, 

3Ulsan University Hospital, Ulsan, REPUBLIC OF KOREA 

PURPOSE 

Hemophagocytic lymphohistiocytosis (HLH) are rare systemic 

diseases and could frequently involve the central nervous 

system (CNS). The purpose of this study was to describe 

the CNS involvement of HLH, focusing primarily on the 

neuroradiologic aspects, and correlate the imaging findings 

with the clinical manifestations, laboratory findings, and 

outcome. 


Between 1996 and 2003, 15 consecutive children with HLH 

were diagnosed by BM biopsy. Patient ages ranged from 1 

day to 7 years, with mean age at diagnosis of 19 months and 

median age of 12 months. Seven patients were boys, 8 girls. 

Brain MR imaging was performed in all of them, brain CT 

in four cases, and proton MR spectroscopy in three cases. 

Findings of the brain CT and MR imaging as well as the clinical 

records reviewed by two radiologists retrospectively. 

RESULTS 

Ten of 15 (67%) patients had CNS involvement at the time 

of the diagnosis and during the course of the disease: Four of 

10 (40%) at diagnosis and 6/10 (60%) during follow up. The 

initial CNS manifestations consisted of meningitis in 2 

patients and focal neurologic symptoms or seizure in 6. Two 

patients complained headache. Two patients with meningitis 

showed normal on CT scan and MR imaging. Two patients 

present mild ventriculomegaly on MR imaging: a patient 

manifest headache with normal CSF study and a patient with 

high fever with mild mental change. Five patients showed 

multiple enhancing nodules or confluent lesions in white 

matter and deep gray matter with or without leptomeningeal 

enhancement, white matter abnormalities, and laminar 

necrosis. Two patients showed subdural hematoma at the 

cerebral convexity. Proton MR spectroscopy showed small 

lactate peak and elevated glutamine/glutamate complex. 

371 

CONCLUSION 

Most common findings of CNS involvement of HLH were 

multiple enhancing lesions in white matter and deep gray 

matter with or without leptomeningeal enhancement, white 

matter abnormalities, and laminar necrosis. These radiologic 

features might help for early awareness of CNS involvement 

of this disease. 

KEY WORDS: Hemophagocytic lymphohistiocytosis, MR 

imaging 

Poster 162 

Cerebellar Hemorrhage in Fetal Parvovirus Infection 

Glenn, O. A. · Callen, P. W. · Parer, J. T. · Barkovich, A. J. 



PURPOSE 

To describe cerebellar involvement in human fetal parvovirus 

infection. 


We reviewed two recent cases referred for fetal MR imaging 

because of suspected cerebellar abnormalities in the setting 

of fetal parvovirus infection. Both fetal MR exams were performed 

on 1.5 T strength magnets, and fast T2-weighted 

images as well as T1-weighted images of the fetal brain were 

obtained. This study was approved by our institutional 

review board. 

RESULTS 

Two cases of fetal parvovirus infection were diagnosed in 

the second trimester because of complications of nonimmune 

hydrops fetalis. In both cases, an umbilical cord blood 

transfusion was performed successfully at 18-20 weeks gestation 

for treatment of the hydrops fetalis. Subsequent ultrasound 

exam demonstrated an abnormal-appearing cerebellum 

in both cases. In one case, there was asymmetry of both 

size and echogenicity of the cerebellar hemispheres, and in 

the second case the cerebellar hemispheres were small. 

Patients were referred for a fetal MR exam for further evaluation 

of the suspected sonographic abnormality at 20 and 

23 weeks, respectively. In the first case, there was evidence 

of hemorrhage in the cerebellar hemisphere characterized by 

high signal on the T1- and T2-weighted images. In the second 

case, there was increased signal on the T1-weighted 

images in both hemispheres consistent with hemorrhage. 

Both pregnancies were terminated and autopsies were not 

performed. 

CONCLUSION 

Human (B19) parvovirus is a common cause of infection and 

is the cause of erythema infectiosum (fifth disease) or 

slapped cheek disease (1). In utero exposure to B19 parvovirus 

can occur, and the outcome is usually good. 

However, fetal parvovirus infection can be associated with 

nonimmune hydrops fetalis and fetal demise, particularly 

when exposure occurs during the first 20 weeks of gestation 

(2). B19 parvovirus infects erythroid precursors, leading to 

severe anemia in the fetus (1). It also infects fetal myocardial 

cells, and this likely also contributes to the development of 

hydrops fetalis (1). Review of the literature reveals that these 

Posters

Posters 

two cases are the first known cases of cerebellar involvement 

in human fetal parvovirus infection. The etiology of the cerebellar 

hemorrhages in our two cases is unknown; however 

one possible cause might be hemorrhage in the cerebellar 

germinal matrix associated with the hydrops fetalis. 

Interestingly, parvovirus infection in animals can result in 

cerebellar hemorrhage and hypoplasia from both a combination 

of direct infection and vascular damage (3). In conclusion, 

we report two cases of cerebellar hemorrhage in the setting 

of fetal parvovirus complicated by hydrops fetalis 

requiring fetal blood transfusion. The cerebellum should be 

evaluated carefully on ultrasound in fetuses with suspected 

parvovirus infection, and fetal MR exam should be performed 

if any abnormalities are suspected. 

REFERENCES 

1. Koch WC. Fifth (human parvovirus) and sixth (herpesvirus 

6) diseases. Curr Opin Infect Dis 2001;14:343-356 

2. Hall SM. Prospective study of human parvovirus (B19) infection 

in pregnancy. BMJ 1990;300:1166-1170 

3. Jordan EK, Sever JL. Fetal damage caused by parvoviral 

infections. Repro Toxicol 1994;8(2):161-189 

KEY WORDS: Fetal, parvovirus, cerebellum 

Poster 163 

Histopathologic Correlates of Proton MR Spectroscopy 

in a Neonatal Piglet Model of Hypoxic-Ischemic 

Encephalopathy 

Wang, X. M. · Guo, Q. Y. 

Second Hospital of China Medical University 

Shenyang, CHINA 

PURPOSE 

To evaluate proton MR spectroscopy ( 1 H-MRS) in diagnosis 

of hypoxic ischemic brain damage (HIBD) in hyperacute 

period. 


Twenty-five term piglets at 3 to 7 days old were subjected 

and divided into one control group (n = 5) and two model 

experiment groups. Model 1: 11 piglets were subjected to 

unilateral common carotid artery ligation and exposure to 

8% oxygen. Model 2: 9 piglets were subjected to bilateral 

common carotid artery transient block and hypoxia. 1 H spectrum 

curve was measured continuously in all cases 0-6, 20- 

24, 44-48, 68-72 hours after HI in frontoparietal region, 

basal ganglia, and hippocampus, Lac/Cr was calculated. 

Histopathologic examination included hematoxylin and 

eosin (HE) stain, gliofibric acid protein (GFAP) stain, terminal 

transferase mediated dUTP-biotin nick-end labeling 

(TUNEL) stain and transmission electron microscope. 

RESULTS 

Both Model 1 and Model 2 showed Lac/Cr increase 0~6 

hours after HI. Lac/Cr in hippocampus region was 0.95 ± 

0.88 in control group compared with 5.65 ± 1.93 in Model 1 

and 8.95 ± 6.59 in Model 2. Model 1 showed significantly 

glial cells swelling in hippocampus region on histopathologic 

examination; Model 2 showed neurons and glial cells 

swelling significantly in hippocampus, the peripheral neurons 

and glial cells were seen prominent apoptosis; further- 

372 

more Lac/Cr stood high within 72 hours. Lac/Cr was 0.41 ± 

0.03 in basal ganglia compared with no significant elevation 

in Model 1 and 13.59 ± 10.23 in Model 2. Model 1 did not 

show neurons and glial cells significantly pathologic 

changes in basal ganglia; Model 2 showed glial cells swelled 

obviously, while neurons changed mildly, Lac/Cr was high 

within 48 hours, and then declined. Lac/Cr in frontoparietal 

region was also seen increasing, but the value was lower 

than the former two regions. 

CONCLUSION 

Proton MR spectroscopy may reflect neurons metabolic 

changes in early HI. Lac/Cr change is not consistent with 

glial cells damage extent, neurons have a acute energy consumption 

after HI. 

KEY WORDS: Hypoxia-ischemia encephalopathy, MR spectroscopy, 

piglet 

Poster 164 

Apparent Diffusion Coefficients in Neonatal Brain 

Injury: Patterns of Injury and Outcome 

Matsuda, K. M. · Lopez, C. J. · Pectasides, M. · Pienaar, R. · 

Krishnamoorthy, K. S. · Grant, P. 


Charlestown, MA 

PURPOSE 

To determine if the extent of brain involvement on neonatal 

diffusion-weighted imaging is more than expected by standard 

ROI analysis and to determine if apparent diffusion 

coefficient (ADC) values correlate clinical outcome. 


Term neonates presenting with seizure and MR imagingI 

within 7 days of life were included. Metabolic disorders 

were excluded. All patients were scanned on a 1.5 T system. 

Imaging included axial DTI (TR/TE 6000/119 or 

7500/102ms, b = 3 and 1221s/mm 2 or b = 0 and 1000 s/mm 2 , 

single shot, FOV 400 x 200 or 200 x 200 mm, matrix 256 x 

128 or 128 x 128, 4 to 6 mm thick, slice gap 0-1 mm, NEX 

= 3). Raw data were transferred to a Linux workstation 

where eddy current corrections were performed and then

DWI and ADC maps calculated. Diffusion-weighted imaging 

and ADC maps were coregistered to a normal reference 

case using FSL. Areas of bright DWI signal were outlined 

using Analyze 5.0 and defined as lesion ROI. Neonates with 

DWI abnormalities were classified as having vascular territory 

(focal) lesions, primarily white and cortical (peripheral) 

lesions or primarily VL thalamus, corticospinal tract and 

perirolandic cortex (central) lesions. Apparent diffusion 

coefficient values for lesion ROI and remaining supratentorial 

brain for each pattern were calculated and compared to 

the mean supratentorial ADC in neonates with no DWI 

abnormalities. Clinical chart reviews determined clinical 

outcome. 

RESULTS 

Mean ADC for supratentorial brain in neonates with normal 

DWI (N = 14) was 1.36 +/- 0.08 x 10 -3 mm 2 /s. In focal lesions 

(N = 11), mean ADC for lesion was 0.91 +/- 0.13 x 10 -3 

mm 2 /s and mean ADC for nonlesional brain was 1.35 +/- 

0.09 x 10 -3 mm 2 /s. In peripheral lesions (N = 9), mean ADC 

for lesion was 0.91 +/- 0.11 x 10 -3 mm 2 /s and mean ADC for 

nonlesional brain was 1.22 +/- 0.11 x 10 -3 mm 2 /s. All focal 

and peripheral cases had good clinical outcomes up to 2.5 

years. In central lesions with good outcomes (N = 1) mean 

ADC of lesion was 1.20 x 10 -3 mm 2 /s and mean ADC for nonlesional 

brain was 1.52 x 10 -3 mm 2 /s. In central lesions with 

poor outcomes (N = 3) the mean ADC for lesion was 0.89 +/- 

0.12 x 10 -3 mm 2 /s and mean ADC for nonlesional brain was 

1.23 +/- 0.05 x 10 -3 mm 2 /s. 

CONCLUSION 

Single vascular territory lesions may occur in otherwise normal 

brains and have good short term outcomes. Both central 

and peripheral patterns having diffuse brain involvement as 

evidenced by decreased ADC in normal-appearing (nonlesional) 

brain but similar ADC reductions, may portend a 

poorer prognosis in the central pattern. 

KEY WORDS: Diffusion-weighted imaging, hypoxic brain 

injury, neonate 

373 

Poster 165 

Elevated Fractional Anisotropy in Neonatal Vascular 

Territory Lesions and Outcome 

Matsuda, K. M. · Lopez, C. J. · Pectasides, M. · Pienaar, R. · 

Krishnamoorthy, K. S. · Grant, P. 


Charlestown, MA 

PURPOSE 

The purpose of this study was to determine if there are 

changes in the directional bias of diffusion as measured by 

fractional anisotropy (FA) in terms of neonatal vascular territory 

lesions. 


Term neonates presenting with focal vascular territory diffusion-weighted 

imaging (DWI) abnormalities on MR imaging 

within 8 days of life were included. Metabolic disorders 

were excluded. All patients were scanned on a 1.5 T system. 

Imaging included axial diffusion tensor imaging (TR/TE 

6000/119 or 7500/102 ms, b = 3 and 1221 s/mm 2 or b = 0 and 

1000 s/mm 2 , single shot, FOV 400 x 200 or 200 x 200 mm, 

matrix 256 x 128 or 128 x 128, 4- 6 mm thick, slice gap 0-1 

mm, NEX = 3). Raw data was transferred to a workstation 

where eddy current corrections were performed and then 

DWI, ADC and FA maps calculated. Diffusion-weighted 

imaging, ADC and FA maps were coregistered to a normal 

reference case. Areas of bright DWI signal were outlined and 

defined as lesion regions of interest (ROIs). Lesion ROIs 

were flipped about the midline and adjusted so that a similar 

region was outlined in the contralateral hemisphere. Mean 

FA was measured in the lesion, and in the similar area as 

lesion on contralateral normal side and a histogram analysis 

performed. 

RESULTS 

Eleven cases with focal vascular territory lesions on DWI 

were studied. Eight had peripheral vascular lesions and 3 had 

deep gray lesions. An additional 4 cases were excluded due 

to artifacts. On average, DTI studies were performed on day 

2.8 (range day 1 - 8). Results are shown in Fig. 1. The eight 

peripheral lesions had a mean FA value of 0.25 +/- 0.04 compared 

to 0.16 +/- 0.02 in the contralateral normal region with 

FA values shifted to the right and decreased in kurtosis on 

the histograms (One patient’s data are shown in Fig. 2). The 

four deep gray lesions had a mean FA of 0.19 +/- 0.07 compared 

to 0.21 +/- 0.06 in the contralateral normal region and 

similar histograms. 

Posters

Posters 

CONCLUSION 

Fractional anisotropy was elevated in focal peripheral vascular 

territory lesions within 8 days of birth compared to contralateral 

normal brain. The presence of elevated FA suggests 

that there is still some structural integrity to the tissue at this 

time. The cause of the increased FA is unclear and further 

studies are required to determine if this represents salvageable 

tissue. Deep gray vascular territory lesions showed no 

significant FA change. These findings may reflect differences 

in pathophysiology of white and gray matter ischemic 

injuries. 

KEY WORDS: Fractional anisotropy (FA), hypoxic brain 

injury, histogram analysis 

Poster 166 

Prospective, Longitudinal Study of the Predictive Value 

of Early Detection of Intracranial Hemorrhage in 

Premature Infants Using T2*(Susceptibility)-Weighted 

MR Imaging 

Benzinger, T. L. S. 1 · Mathur, A. 2 · Moinuddin, A. 1 · Almli, 

R. 3 · Neil, J. J. 1,2 · McKinstry, R. C. 1 

1Mallinckrodt Institute of Radiology, Washington University, 

St. Louis, MO, 2St. Louis Children’s Hospital, Washington 

University, St. Louis, MO, 3Washington University, St. 

Louis, MO 

PURPOSE 

To evaluate the sensitivity of T2*(susceptibility)-weighted 

MR imaging for early detection of intracranial hemorrhage 

in premature infants and its utility in predicting clinical outcomes. 


Eighteen premature infants (gestational age 26-32 weeks) 

underwent FLAIR, T1-weighted imaging, T2-weighted 

imaging and T2* MR imaging at 1.5 T. T2*-weighted imaging 

were obtained using single-shot gradient-echo EPI. 

Initial MR imaging occurred during the first 48 hours of life. 

Follow-up scanning was targeted to term equivalent age. 

Blinded readers evaluated initial and follow-up MR images 

for severity of injury and confidence in diagnostic interpretation, 

both including and excluding T2*-weighted imaging. 

The conventional MR imaging obtained at term equivalence 

was used as the gold standard to identify areas of injury. 

From ages 6 to 48 months, children were enrolled in a developmental 

testing protocol with standardized assessments of 

cognitive, motor, and social function. 

RESULTS 

Fifty-six percent (10/18 infants) of these otherwise healthy 

premature infants had abnormal MR exams at birth. Of 

these, three with abnormal findings had subtle lesions at 

birth, confirmed at the 40-week scans. In these cases, T2*weighted 

imaging was necessary for identification of the 

abnormality on the early scan. In an additional 4 infants, 

early hemorrhage was identifiable on both T2- and T2*weighted 

imaging, but T2*-weighted imaging added significantly 

better conspicuity. Lesions consisted of both intraparenchymal 

and intraventricular hemorrhages. Sixteen 

infants completed clinical follow up. Standardized testing 

demonstrated 14/16 (87.5%) to be at least one standard devi- 

374 

ation (SD) outside of normal controls on cognitive and/or 

behavioral testing and 6/16 (37.5%) to be more than two SD 

beyond the norm. However, of the seven infants with hemorrhagic 

lesions on MR imaging, two were normal at follow 

up, three were between 1 and 2 SD, and only two were more 

than 2 SD beyond normal subjects. Furthermore, of the six 

infants with completely normal MR images, four were more 

than 2 SD beyond normal subjects at clinical follow up. 

CONCLUSION 

T2* MR imaging is highly sensitive for the early diagnosis 

of intracranial hemorrhage in premature infants. In this 

study, 56% of otherwise “normal” preterm infants had 

intracranial hemorrhages detectable within 48 hours of birth 

using T2*(susceptibility)-weighted imaging. However, this 

early detection was not predicative of early childhood disability 

(p > 0.05). The children remain enrolled in the developmental 

testing protocol and predictive value of the MR 

scans will be reassessed once the children reach school age. 

KEY WORDS: T2*-weighted MR imaging, susceptibility, 

prematurity 

Poster 167 

Immature Brains: Volume and Cortical Surface Analysis 

Pienaar, R. 1,2 · Fischl, B. 1 · Nishida, M. 2 · Busa, E. 1 · Caviness, 

V. S. 2 · Makris, N. 2 · Kennedy, D. N. 2 · Grant, P. E. 1 

1Massachusetts General Hospital/Massachusetts Institute of 

Technology/Harvard Medical School Athinoula A. Martinos 

Center for Biomedical Imaging, Charlestown, MA, 2Center for Morphometric Analysis, Charlestown, MA 

PURPOSE 

Most available tools in immature brain analysis allow only 

volumetric analysis. Software tools capable of performing 

more sophisticated analysis are designed to process adult 

brains, which have significantly different MR contrast. The 

purpose of this study is to develop a technique to import 

manually segmented immature brains into an existing adult 

software processing stream to allow more complex measurements. 

We also report some preliminary quantitative results 

for the immature brain.


Volumetric MR imaging data were collected on three normal 

newborn subjects with gestational ages 31 - 39 weeks and 

one normal adult control. T1-weighted axial 3D SPGR 

images were collected on a 1.5 T (GE, Milwaukee, WI)) 

scanner, with TR/TE = 30/8, flip angle = 25 to 30, matrix = 

256 x 192, FOV = 220 x 165 mm or 200 x 150 mm with a 

slice thickness of 1.2 to 1.4 mm. These images were segmented 

manually, and the resultant surfaces then were converted 

into a volume format compatible with a cortical 

analysis and reconstruction package tool. The converted volumes 

did require some minor postconversion manual editing 

to correct some topological defects—primarily topological 

“bridges” and some missed cortical gyri. Once these defects 

were corrected, we reconstructed cortical surfaces and performed 

cortical-based analyses. 

RESULTS 

In the following tables, [L] = left; [R] = right; WMV = white 

matter volume; cGMV = cortical gray matter volume; IGC = 

Integrated Gaussian Curvature 

Right Hemisphere 

Subject Total WMV [R] Surface [R] cGMV [R] Cortical [R] Average 

(mm^3) area (mm^2) (mm^3) Thickness (mm) IGC 

31 gestational 79.973 x 10^3 14.956 x 10^3 30.923 x 10^3 2.088 +- 0.536 0.066 

37 gestational 129.140 x 103 31.883 x 10^3 72.487 x 10^3 2.280 +- 0.593 0.076 

39 gestational 152.998 x 10^3 48.883 x 10^3 113.923 x 10^3 2.319 +- 0.663 0.086 

Adult control 722.140 x 10^3 90.773 x 10^3 260.189 x 10^3 2.868 +- 0.979 0.179 

Subject [L] Surface area 

Left Hemisphere 

[L] cGMV [L] Cortical [L] Average 

(mm^2) (mm^3) Thickness (mm) IGC 

31 gestational 14.762 x 10^3 31.750 x 10^3 2.169 +- 0.514 0.07 



Adult control 87.791 x 10^3 255.154 x 10^3 2.909 +- 0.944 0.179 

CONCLUSION 

The total white matter cortical gray matter volume measurements 

for the immature brains show a very good correspondence 

with existing data. New data are provided for cortical 

thickness and the increase in the standard deviation as well 

as the mean cortical thickness with age is consistent with 

concepts of cortical specialization. New data are provided 

also for cortical surface area and average IGC, which provides 

a measure of cortical folding, providing quantitative 

measures of the increasing surface area and folding with age. 

KEY WORDS: Immature brain, cortical analysis, automated 

tools 

Poster 168 

Brain Metabolite Changes in Normal-Appearing White 

Matter in Tuberous Sclerosis Complex 

Garaci, F. G. · Ludovici, A. · Marziali, S. · Melis, M. · 

Gaudiello, F. · Floris, R. · Simonetti, G. 


Rome, ITALY 

PURPOSE 

Significant advances have been made in understanding the 

molecular and clinical features of tuberous sclerosis complex 

(TSC). Microscopic lesions of white matter are not 

depicted by conventional MR imaging. Statistically significant 

changes of normal-appearing white matter (NAWM) 

have been demonstrated by diffusion MR imaging. The present 

study was designed to test the hypothesis that MR proton 

375 

spectra of NAWM is abnormal because of clusters of dysplastic 

cells invariably present in the white matter of TSC 

patients. 


Nine patients with clinically established TSC and nine agematched 

control subjects underwent proton MR spectroscopy 

with a single-voxel point-resolved spectroscopy 

pulse sequence (1.5 T scanner). Voxels (8 cm 3 ) were placed 

in the NAWM of the centrum semiovale bilaterally. N-acetylaspartate 

(NAA)/creatine (Cr) and choline (Cho)/Cr ratios 

were assessed. Generalized linear regression analysis test 

was used for statistical comparisons. 

RESULTS 

Compared with control subjects, TSC patients showed 

increased NA/Cr (1.87 vs 2.23) and NA/(Cho+Cr) ratio 

(0.94 vs 1.03). Increase in Cho/Cr ratio was observed also 

between the controls and TSC group (0.98 vs 1.16). 

CONCLUSION 

In TSC proton MR spectroscopy depicted abnormalities in 

the NAA/Cr and NACho ratios in the normal-appearing 

white matter suggesting that brain changes may be more diffuse 

than expected. Further studies are needed to determine 

if abnormalities detected by proton MR spectroscopy might 

represent a neuroimaging marker to assess functional outcome. 

KEY WORDS: Tuberous sclerosis complex 

Poster 169 

Novel Artificial Intelligence Approach to Knowledge 

Representation in Radiologic Disease Definition 

Govindarajan, S. 1 · Ranganathan, L. 2 · Krishnamoorthy, A. 1 · 

Govindarajulu, S. 3 · Srinivasan, A. 1 

1 2 Madras Medical College, Chennai, INDIA, Institute of 

Mental Health, Chennai, INDIA, 3Bharat Institute of 

Imaging, Chennai, INDIA 

PURPOSE 

Knowledge representation is a vital initial step in the development 

of artificial intelligence expert systems in radiologic 

decision-making. Acquisition from known experts in the 

speciality is fraught with several interpretative pitfalls. Selfgeneration 

of rules combined with supervised learning is an 

optimal method. Aim: To evaluate self-generation of disease 

defining rules combined with supervised learning in the radiologic 

evaluation of leucodystrophies. 


Software employing artificial intelligence in the form of neural 

networking was developed in visual basic. Knowledge 

representation was through self-generation of rules from a 

supplied set of known cases. These rules were reviewed by a 

set of radiologists and then incorporated in a supervised 

learning scenario. Twenty-three known cases of adrenoleucodystrophy 

were analyzed. The imaging signs were tabulated 

in a binary format. The eventual rules generated were 

compared with standard disease-defining criteria in the literature. 

The overall accuracy of the generated rules were evaluated. 

The logical stability was inferred by reevaluating all 

Posters

Posters 

previous data for consistency, at every new learning episode. 

The minimum number of cases required for satisfactory 

accuracy was related to the number of radiologic signs used 

in a relation plotted. 

RESULTS 

The accuracy of the generated rules showed a significant 

increase from 67% to 89% using the current number of data. 

Logical stability, as an index of consistency was maintained. 

A graph relating accuracy to size of the dataset and number 

of radiologic signs used was generated. 

CONCLUSION 

Self-generation of rules with incorporation under supervised 

learning is a reliable method of knowledge representation. 

The minimum size of the dataset can be related to the number 

of radiologic signs used for interpretation. A larger 

dataset is expected to improve the accuracy. 

KEY WORDS: Neural networking, artificial intelligence, leucodystrophy 

Poster 170 

Susceptibility-Weighted Imaging: A New MR Imaging 

Technique with Application to Pediatric Neurologic 

Disorders 

Tong, K. A. 1 · Ashwal, S. 1 · Kido, D. 1 · Haacke, M. 2,3 

1 Loma Linda University Medical Center, Loma Linda, CA, 

2 Wayne State University, Detroit, MI, 3 MRI Institute of 

Biomedical Research, Detroit, MI 

PURPOSE 

Susceptibility-weighted imaging (SWI) is a high spatial resolution 

3D gradient-cho MR imaging technique with phasesubtraction 

postprocessing that accentuates the paramagnetic 

properties of blood products and is very sensitive for 

detection of intravascular venous deoxygenated blood as 

well as extravascular blood products. This technique has 

potential to enhance the visibility of various pathologic conditions. 

We present several cases to illustrate the application 

of this technique. 


The SWI sequence consists of a strongly susceptibilityweighted, 

low-bandwidth (78 Hz/pixel) 3D-FLASH 

sequence (TR/TE = 57/40 ms, FA = 20°) first-order flow 

compensated in all 3 orthogonal directions. Thirty-two to 64 

partitions of 2 mm are acquired using a rectangular FOV (5/8 

of 256 mm) and a matrix size of 160 x 512, resulting in a 

voxel size of 1 x 0.5 x 2 mm 3 . The acquisition time varies 

from approximately 5-10 minutes, depending on the desired 

coverage. After the data is acquired, there is additional postprocessing 

that accentuates the signal loss caused by any 

susceptibility effects. This primarily involves the creation of 

a phase mask to enhance the phase differences between paramagnetic 

substances and surrounding tissue. The sequence 

and automatic postprocessing is available for Siemens 

Vision and Sonata 1.5 T MR scanner platforms. 

376 

RESULTS 

Susceptibility-weighted imaging has been used at our institution 

since 2001. Six cases are presented to demonstrate the 

clinical usefulness of SWI, specifically the (1) improved 

detection of the number and volume of hemorrhagic lesions 

seen with posttraumatic diffuse axonal injury (DAI); (2) 

improved detection of subtle cryptic vascular malformations; 

(3) improved delineation of unusual vascular malformations 

with insidious parenchymal steal phenomenon such 

as Sturge-Weber syndrome; (4) improved demonstration of 

increased oxygen extraction in the setting of hypoxic, 

ischemic, or infarctive injury; (5) improved delineation of 

internal architecture of neoplasms; and (6) improved detection 

of mineral deposition in metabolic disorders such as 

Hallovorden-Spatz disease. 

CONCLUSION 

Susceptibility-weighted imaging is a highly sensitive technique 

that is particularly useful in delineating the extent of 

hemorrhagic lesions, such as in the setting of trauma. It is 

also better at detecting venous vascular lesions, possibly due 

to physiologic changes in brain oxygen extraction that may 

occur in some of these malformations as well as in areas of 

hypoxia, ischemia, or infarction. This imaging has provided 

new understanding of some of these disease processes. It is 

hoped that the information provided by susceptibilityweighted 

imaging may help provide prognostic information 

in children with a variety of neurologic disorders. 

REFERENCES 

1. Haacke EM, Xu Y, Cheng YC, Reichenbach JR. Susceptibility 

weighted imaging (SWI). Magn Reson Med 2004;52:612-618 

2. Tong K, Ashwal S, Holshouser B, et al. Improved detection of 

hemorrhagic shearing lesions in children with post-traumatic 

diffuse axonal injury - initial results. Radiology 

227;2003:332-339 

3. Lee BC, Vo KD, Kido DK, et al. MR high-resolution blood 

oxygenation level-dependent venography of occult (lowflow) 

vascular lesions. AJNR Am J Neuroradiol 1999;20:1239- 

1242 

KEY WORDS: Susceptibility-weighted imaging, hemorrhage, 

vascular lesions 

Poster 171 

Pontine Tegmentum of the Human Brain: Is It Normally 

Hyperintense on T2-Weighted MR Imaging? 

Asao, C. 1 · Hirai, T. 1 · Imuta, M. 1 · Okuda, T. 1 · Goto, Y. 2 · 

Korogi, Y. 3 · Yamashita, Y. 1 

1 Kumamoto University Graduate School of Medical 

Sciences, Kumamoto, JAPAN, 2 Kumamoto Regional 

Medical Center, Kumamoto, JAPAN, 3 University of 



PURPOSE 

We have often encountered high signal intensities of the pontine 

tegmentum on T2-weighted MR images in neurologically 

healthy individuals. However, the signal intensity in the 

pontine tegmentum has not been reported. The purpose of

this study was to determine whether signal intensity of the 

pontine tegmentum differs from that of the pontine base on 

T2-weighted MR images. 


Signal intensity of the pontine tegmentum and base on T2weighted 

images in 38 neurologically healthy subjects were 

evaluated. There were 29 adults (16 males and 13 females; 

age range, 23-48 years; mean age, 39.5 years) and 9 children 

(4 boys and 5 girls; age range, 4-9 years; mean age, 6.5 

years). Contrast-to-noise ratios (CNRs) were compared 

between the tegmentum and the base in the upper pons, midpons, 

and lower pons. Signal intensity ratios of the tegmentum 

to the base also were evaluated. 

RESULTS 

Contrast-to-noise ratios of the tegmentum were significantly 

higher than those of the base in each level of the pons (p < 

.0001), and signal intensity ratios of the tegmentum to the 

base in the upper pons were significantly higher in children 

than in adults (p < .005). 

CONCLUSION 

On T2-weighted images, high signal intensities of the pontine 

tegmentum are seen frequently in neurologically healthy 

subjects. These findings should not be considered abnormal, 

particularly in children. 

KEY WORDS: T2-weighted imaging, normal brain, anatomy 

Poster 172 

Imaging Manifestations of Whole Brain Radiation in 

Children 

Berkowitz, E. A. · Nedzi, L. A. · Zakaris, E. · Rosel, P. 

Tulane University Medical Center 

New Orleans, LA 

PURPOSE 

Evaluate MR and CT imaging characteristics in children 

who underwent whole brain radiation therapy 


Thirty children up to 12 years old were analyzed retrospectively 

via MR and CT imaging characteristics to evaluate 

radiation-induced changes within the brain. 

377 

RESULTS 

MR and CT postradiation changes were divided into acute, 

early-delayed, and late-delayed reactions. Acute and earlydelayed 

imaging findings were variable and included normal 

to hypodensity (CT) and T2 prolongation (MR imaging) 

consistent with edema with or without variable enhancement. 

Late-delayed injury secondary to vasculopathy was 

demonstrated as white matter infarction, ranging from single 

focal lesions to diffuse white matter abnormalities with or 

without contrast enhancement to mineralizing microangiopathy 

and telangiectasia. Chronic changes include 

encephalomalacia 

CONCLUSION 

A wide range of postradiation changes were seen in children 

including normal, edema/white matter changes with/without 

variable enhancement, microangiopathy, and encephalomalacia. 

KEY WORDS: Postradiation changes, whole brain radiation, 

children 

Socioeconomic 

173 

Poster 173 

Carotid Stenting: The Nationwide Inpatient Sample 

Database 1998-2002 

Bateman, B. T. · Pile-Spellman, J. 


New York, NY 

PURPOSE 

The purpose of this study is to examine carotid stenting from 

an epidemiological perspective. In part one of the study, we 

analyzed changes in frequency with which stents are being 

used. In part two, we analyzed the effect of hospital stenting 

volumes on outcomes following stenting. 

MATERIALS & METHOD 

This study used data from the Nationwide Inpatient Sample 

(NIS), the largest database of inpatient admissions available 

in the United States, for the years 1998-2002. The database 

represents a stratified sample of approximately 20% of all 

inpatient hospitalizations in the U.S. Patients with a primary 

diagnosis of carotid stenosis without infarction were identified 

(n = 157,672). Patients who underwent stenting of the 

carotid artery were then identified (n = 3478), as were 

patients who underwent carotid endarterectomy (n = 

132,520). For purposes of analyzing outcome following 

stenting, the cohort was then limited to those patients admitted 

from home, so that discharge to a facility other than 

home could serve as a marker for poor outcome (n = 3390). 

Hospitals averaging 10 or more stenting cases a year were 

designated as high-volume hospitals (corresponding to ≥ 

88th percentile for hospital case volume). Patients were designated 

as treated at either a high or nonhigh volume hospital. 

The effect of stenting hospital volume on three primary 

outcomes was examined: 1) in-hospital mortality, 2) discharge 

to a facility other than home, and 3) prolonged length 

Posters

Posters 

of stay (LOS) (> 7 days, corresponding to ≥ 90th percentile 

for LOS). Logistic regression was used to adjust for several 

patient characteristics, including co-morbid medical conditions. 

RESULTS 

In NIS sample for 1998, 230 patients were treated with stents 

at 11 high volume hospitals and 151 patients were treated at 

68 non-high volume centers, for an overall ratio of 1.5 

carotid stents per 100 endarterectomies. Stenting volumes 

steadily rose during the study period such that in 2002, 675 

patients were treated at 23 high-volume hospitals and 342 

patients were treated at 122 nonhigh-volume hospitals, for 

an overall ratio 3.6 stents per 100 endarterectomies. The inhospital 

mortality rate was 0.5% for high-volume centers 

and was 1.5% for non high-volume centers (p ≤. 01). The 

rate of discharge to facility other than home was 3.9% for 

high-volume centers and 5.8% for nonhigh-volume centers 

(p ≤. 025). The rate of prolonged LOS was 5.8% for high 

volume centers and 12.2% for nonhigh volume centers (p ≤. 

001). Differences in outcomes persisted after adjustment for 

other variables using logistic regression. For patients treated 

at nonhigh-volume centers (vs patients treated at high-volume 

centers) the odds ratio (OR) of in-hospital mortality was 

3.062 (95% CI, 1.436-6.526), the OR of discharge to facility 

other than home was 1.606 (95% CI, 1.151-2.240) and the 

OR of prolonged LOS was 2.319 (95% CI, 1.796-2.995). 

CONCLUSION 

Carotid stents are being used with increasing frequency for 

the treatment of carotid stenosis. As the procedure becomes 

more commonplace, it is important to determine the appropriate 

setting for this procedure. Our study suggests that centers 

with high levels of experience with the procedure have 

superior outcomes, arguing that high-volume referral centers 

are the preferred treatment setting. 

KEY WORDS: Carotid stenting, outcomes, hospital volume 

Poster 174 

MR Imaging Findings in Cauda Equina Gnathostomiasis 

Sawanyawisuth, K. 

Khon Kaen University 

Khon Kaen, THAILAND 

Spine 

174-184 

We report a patient with cauda equina syndrome caused by 

Gnathostoma spinigerum, comfirmed by immunoblotting 

test. The MR images of the lumbosacral spine showed long 

segmental (Fig. 1), increased signal intensity along the cauda 

equina with irregular enhancement on the postcontrast study 

(Fig. 2). The conus medullaris was slightly enlarged with 

abnormal enhancement (Figs. 2 and 3). The patient was 

treated with corticosteroids and her clinical conditions 

improved. The MR imaging, 9 months after treatment, completely 

resolved (Fig. 4). 

378 

KEY WORDS: Gnathostomiasis, cauda equina, MR imaging

Poster 175 

Let the Gravity Work to Pick Up the Stiff Ropes 

Stantoun, M. · Ho, Y. · Ting, G. · Midia, M. 

McMaster University 

Hamilton, ON, CANADA 

PURPOSE 

To assess the value of lateral decubitis MR imaging of lumbar 

spine in depicting stiff cauda equina that may be associated 

with disorders of the distal techal sac such as arachnoiditis, 

drop metastasis, and tethered cord. 


Prospective study of 60 patients referred to our department 

for lumbar spine MR imaging. The subjects were divided into 

two groups: Group A consist of 30 patients who had no prior 

history of spinal surgery, arachnoiditis, malignancy, or tethered 

cord. The remaining 30 patients with one or more of the 

above features in their medical history were assigned to 

group B. Patients with severe spinal stenosis were excluded 

from the study. Sixty consecutive patients underwent MR 

imaging of the lumbar spine. All of the patients had an axial 

T2-weighted imaging in either right or left lateral decubitus 

position (in patient with unilateral symptoms the exam was 

conducted having the symptomatic side in nondependent 

position and subject with no lateralizing symptoms where 

randomized to right or left) in addition to the standard exam 

was obtained in supine position. One neuroradiologist and 

one MR imaging radiologist evaluated the studies for presence 

or absence of gravitational position change of the cauda 

quina. 

RESULTS 

All subjects in group A showed gravitational movement of 

the cauda equina. Five patients out of 30 patients in group B 

had partial or complete restricted cauda equina movement 

with gravity. 

CONCLUSION 

Stiff cauda equina depicted on lateral decubitus MR imaging 

can be a useful sign in investigating cauda equina pathologies 

that sometimes can be difficult to assess on standard 

supine exam. 

KEY WORDS: Cauda equina, arachnoiditis, metastasis 

Poster 176 

MR Imaging of the Cervical Cord and Brachial Plexus in 

Patients with Multifocal Sensorimotor Demyelinating 

Neuropathies with Pathologic Correlation 

Laothamatas, J. 1 · Mitrabhakdi, E. 2 

1 2 Ramathibodi Hospital, Bangkok, THAILAND, Chulalongkorn 

University, Bangkok, THAILAND 

PURPOSE 

To study the neuroimaging with electrodiagnostic and pathologic 

features correlation in patients with multifocal sensorimotor 

demyelinating neuropathy. 

379 


Fifteen MR imagings of the cervical spine and brachial 

plexus of the patients with clinically diagnosed sensorimotor 

mononeuropathy multiplex were performed with a 1.5 T 

superconductive magnet at Ramathibodi Hospital (CVi/NVi, 

General Electric Medical Systems, Milwaukee, WI). 

Gadolinium, 0.1 mmol/kg, was adminstered in all patients. 

The cervical spine study composed of sagittal T1-weighted, 

FSEPS/T2-weighted, coronal GRET2, axial T1-weighted 

and GRET2 from C2-3 to C7-T1 and postgadolinium sagittal 

and axial T1-weighted. The brachial plexus study composed 

of coronal and axial T1-weighted, FSET2-weighted 

with fat suppression and T1-weighted with suppression 

before and after gadolinium from C3 to T4. 

Electrophysiologic diagnosis and sensory nerve biopsies 

with tease fiber technique were performed in all patients. 

Routine blood chemistries, spinal fluid examination, and 

antiGM1 Ab assay were tested. 

RESULTS 

Fifteen patients with sensorimotor demyelinating neuropathies 

with hypersignal T2 exchange of the cervical cord, 

mild intradural cervical nerve root, and varying degree of 

enhancement and hypersignal T2 along the course of the 

brachial plexus were found in all cases corresponding to the 

clinical and electrophysiologic assesement. None of these 

studies demonstrates compressive lesion nor tumor infiltration. 

Sensory nerve biopsies showed lymphocytic infiltration 

and segmental demyelination in all cases. Anti-GM1 Ab 

assay were negative in all. All patients responded dramatically 

to IVIG and steroid or azathioprine. 

Posters

Posters 

CONCLUSION 

Abnormal cervical cord signal, intradural cervical nerve root 

enhancement and abnormal enhancing hypersignal T2 along 

the brachial plexus can be demonstrated in patients with 

multifocal sensorimotor demyelinating neuropathies, a CIDP 

variant due to segmental demyelination with lymphocytic 

infiltration of the peripheral sensory nerves. 

KEY WORDS: Brachial plexus, myelopathy, MR imaging 

Poster 177 

Comparison of Spinal Diffusion-Weighted Imaging and 

Apparent Diffusion Coefficient Values to Differentiate 

Spinal Infections, Tumors, and Degenerative Changes 

Tali, T. E. · Celik, H. · Ucar, M. · Gultekin, S. · Tokgoz, N. 

Gazi University School of Medicine 


PURPOSE 

Diffusion-weighted imaging allows measurement of the random 

motion of water protons, and has been used widely as 

an important modality in the diagnostic workup of central 

nervous system abnormalities. It also provides us to study 

bone marrow and bone marrow alterations on the basis of 

altered water-proton mobility in various diseases. There are 

limited number of researches on spinal diffusion-weighted 

imaging, and it has been used especially for the distinction of 

acute benign osteoporotic and malignant vertebral compression 

fractures. The purpose of this study is to evaluate apparent 

diffusion coefficient (ADC) values with echo-planar diffusion-weighted 

MR imaging in patients with spondylodiskitis, 

degenerative changes, and vertebral metastases. 


Twenty patients (8 degenerative changes, 8 spondylodiskitis, 

4 metastases) were examined with echo-planar diffusionweighted 

MR imaging. The diffusion-weighted imaging was 

performed using a 1.5 T system, echo-planar pulse sequence 

with b = 600 s/mm 2 . Following the qualitative evaluation, the 

ADC value was measured from the normal-appearing and 

pathologic vertebrae and the ADC ratios 

(ADC Pathologic /ADC Normal ) were calculated. 

RESULTS 

Apparent diffusion coefficient maps of the spondylodiskitis, 

degenerative changes, and vertebral metastases showed 

hyperintense signal to adjacent normal vertebral bodies. The 

mean ADC values for normal bone marrow (n = 20), degenerative 

changes, spondylodiskitis, and metastases were 0.60 

± 0.39 × 10 −3 mm 2 /s, 1.14 ± 0.49 × 10 −3 mm 2 /s, 1.54 ± 0.51 × 

10 −3 mm 2 /s, 1.53 ± 0.35 × 10 −3 mm 2 /s, respectively. The mean 

ADC ratios for degenerative changes, spondylodiskitis, and 

metastases were 2.27 ± 1.12, 2.63 ± 1.05, 4.08 ± 1.89, 

respectively. The ADC ratios of vertebral metastases were 

found higher than spondylodiskitis, and degenerative 

changes. All of the vertebral end plate degenerations were 

hypo to isointense to adjacent normal vertebral body on the 

diffusion-weighted imaging. Five of the eight cases showed 

linear hyperintensity around the degeneration. All of the vertebral 

metastases were hyperintense to normal vertebral bod- 

380 

ies at diffusion-weighted imaging. Spondylodiskitis were 

either iso (three of eight) or hyperintense (five of eight) to 

adjacent normal vertebral bodies. 

CONCLUSION 

All the metastases showed hyperintense signal while none of 

the cases with degeneration showed hyperintensity on diffusion-weighted 

imaging. The quantitative assessment showed 

high ADC ratios of metastases, when compared with degenerative 

changes and spondylodiskitis. We assume that ADC 

values and ADC ratios and also diffusion-weighted imaging 

of the spine could be helpful in the differentiation of metastases, 

degenerative changes, and spondylodiskitis. More 

cases are required for the accurate statistical evaluation for 

this ongoing research. 

KEY WORDS: Spinal diffusion-weighted imaging 

Poster 178 

Diffusion-Weighted MR Imaging and Apparent 

Diffusion Coefficient Maps Aid in the Identification of 

Treatable Brain, Retropharyngeal, Posterior Paraspinal, 

Neck, and Spine Abscesses due to Nocardia Asteroides in 

a Human Immunodeficiency Virus Positive Patient 

Fasano, N. · Bhatia, R. G. · Falcone, S. · Post, M. D. 

University of Miami School of Medicine 

Miami, FL 

PURPOSE 

To present an unusual case of disseminated nocardial infection 

affecting the brain, neck, and spine and to demostrate 

the utility of diffusion-weighted imaging/apparent diffusion 

coefficient (ADC) in diagnosis of this treatable opportunistic 

infection. 


A 47-year-old male with human immunodeficiency 

virus/acquired immunodeficiency syndrome (HIV/AIDS) 

and a CD4 count of 10 presented with a painful neck mass 

which had been increasing in size over the past 2 months and 

headache. MR and CT imaging of the brain, neck, and cervical 

spine was performed, showing four sizeable ring-enhancing 

lesions in the brain and a large multiloculated lesion 

extending from the skull base to the thoracic inlet involving 

the retropharyngeal, prevertebral, and posterior paraspinal 

spaces with invasion of the C1-4 vertebral bodies, disk 

spaces, and epidural spaces. Diffusion-weighted sequences 

were performed in the brain, showing restricted diffusion 

within these lesions (Fig. 1A) including the retropharyngeal 

and posterior paraspinal lesion, which was partially visible 

on the brain diffusion-weighted imaging/ADC (Fig. 1B). 

Fine needle aspiration of the neck abscess then was performed, 

demonstrating growth of nocardia asteroides, and 

establishing the diagnosis of nocardial abscesses. 

RESULTS 

The patient, who was found also to have disseminated nocardia 

to the skeleton, was placed on bactrim therapy with some 

initial treatment response.

CONCLUSION 

In this severely immunocompromised patient presenting 

with an enlarging neck mass and headache, MR and CT 

imaging were able to establish the presence of ring-enhancing 

brain lesions and extensive spine and neck involvement. 

While contrast MR imaging and CT identified these focal 

mass lesions, it was only the diffusion imaging which established 

the diagnosis of bacterial abscesses and excluded 

tumor due to the finding of centrally restricted diffusion in 

these ring-enhancing lesions in the brain, neck, and 

paraspinal spaces. These diffusion-weighted imaging/ADC 

findings then prompted biopsy which found that nocardia 

asteriodes was the specific bacterial agent and appropriate 

antimicrobial therapy therefore was able to be initiated. 

Based on this case, we postulate that in addition to diffusion 

imaging being a useful tool for differentiating infectious 

from neoplastic lesions in the brain, as has been reported 

previously, it also can be a useful diagnostic tool in spine 

cases. We speculate that these paraspinal masses with 

restricted diffusion could indicate the presence of infection 

rather than tumor, a conclusion supported by prior reports in 

the literature dealing with similar cases in the liver and 

spinal epidural spaces. 

KEY WORDS: Spine abscess, diffusion imaging 

Poster 179 

Adhesive Spinal Arachnoiditis and Arachnoid Cyst after 

Aneurysmal Subarachnoid Hemorrhage: Report of Two 

Cases 

Tampieri, D. · Cortes, M. D. P. · Melancon, D. 

Montreal Neurological Institute, McGill University 


PURPOSE 

We wish to report the imaging findings of two patients who 

developed spinal adhesive arachnoiditis following subarachnoid 

hemorrhage (SAH), secondary to ruptured intracranial 

aneuryms. One of the aneurysms was located in the anterior 

circulation. 


Two middle-aged women were treated succesfully with coiling 

for ruptured aneurysms. Both of them recovered satisfactorily 

from multiple initial complications, and remained 

asymptomatic for 8 and 52 months, respectively. They developed 

progressive thoracic cord compression myelopathy that 

required surgical intervention with resolution of symptoms. 

381 

RESULTS 

The first patient had a small ruptured right vertebro-basilar 

junction aneurysm documented angiographically. Late spinal 

MR imaging demonstrated an anterior subarachnoidal cyst 

extending from C6 to T3, and cord compression. In the second 

patient, CT and angiography showed a large ruptured 

left carotid ophthalmic aneurysm. A late CT myelogram 

revealed arachnoidal adhesions to the posterior aspect of the 

cord from T8-T10 levels. 

CONCLUSION 

Two cases of spinal adhesive arachnoiditis after SAH are 

reported. They illustrate an entity of rare ocurrence, and one 

of the cases constitutes, to our knowledge, the first report of 

a documented internal carotid-ophthalmic artery aneurysm 

originating the problem. A brief discussion on the topic is 

provided in the context of the two cases and the literature. 

KEY WORDS: Arachnoiditis, subarachnoid hemorrhage, 

arachnoid cyst 

Poster 180 

Kyphoplasty Increases Vertebral Body Height More than 

Vertebroplasty: A Cadaveric Study 

Hiwatashi, A. · Sidhu, R. · Lee, R. K. · deGuzman, R. · 

Piekut, D. T. · Westesson, P. A. 

University of Rochester Medical Center 

Rochester, NY 

PURPOSE 

To compare height restoration with kyphoplasty and vertebroplasty 

in fresh compression fractures in cadavers. 


We harvested 37 vertebrae from four donated cadavers of 

elderly individuals. The vertebrae were dissected free of the 

surrounding muscles and imaged with multidetector CT. 

Compression fractures were induced and the vertebrae 

wereimaged again. The vertebrae were randomized to be 

treated with kyphoplasty (n = 19) or vertebroplasty (n = 18) 

followed by CT. The anterior, central, and posterior vertebral 

body heights and wedge angles were measured in the midsagittal 

plane of the reformatted images. The amount of 

cement injected was determined by weighing the vertebrae 

before and after treatment. 

RESULTS 

Kyphoplasty increased the vertebral height more than vertebroplasty 

(5.1 mm versus 2.3 mm, p < 0.05). Ninety-three 

percent of the original vertebral height was restored with 

kyphoplasty compared to 82% with vertebroplasty (p < 

0.05). Kyphoplasty decreased the wedge angles more than 

vertebroplasty (3.1 degree versus 1.6 degree) but this difference 

was not statistically significant (P > 0.05). There was no 

statistically significant difference in the amount of cement 

injected with kyphoplasty and vertebroplasty (p > 0.05). 

Posters

Posters 

CONCLUSION 

Kyphoplasty increased vertebral body height more than vertebroplasty 

in this experimental model with acute fractures 

in cadaveric specimens. The differences in height restoration 

between the two techniques were small and the clinical significance 

is uncertain. 

KEY WORDS: Spine, vertebroplasty, kyphoplasty 

Poster 181 

Comparison of Vertebroplasty Cement Delivery Devices 

Watanabe, A. T. 1,2 · Yuhan, J. 2 · Georgy, B. 3,4 

1Long Beach Memorial Medical Center, Long Beach, CA, 

2University of Southern California School of Medicine Los 

Angeles, Los Angeles, CA, 3Valley Radiology, Del Mar, CA, 

4University of San Diego School of Medicine, San Diego, 

CA 

PURPOSE 

Syringes and cement delivery devices are commonly used 

methods of applying cement into the spine during vertebroplasty. 

We attempt to qualitatively and quantitatively evaluate 

various methods of cement delivery through in vitro testing. 


Evaluation was performed on multiple commercial devices 

including those made by Cook, Arthrocare-Parallax, 

Cardinal Health, Spinal Specialties, as well as with 1 cc luer 

lock syringes (Merit Medical). Factors evaluated included 

radiation exposure reduction (working distance), ease of 

device use such as device filling, device turning, torque, 

cement and device working time, cost, and safety. Working 

distance was measured from tip of delivery device dispenser 

site to nearest operator hand. Qualitative measurements were 

obtained using a glycerin pressure meter, force gauge, and 

Dial torque wrench. Devices were tested using Simplex 

cement (Stryker), Secour cement and Tracers barium 

(Arthrocare), and Play-Doh (Hasbro). The Play-Doh testing 

was used for qualitative assessment of delivery devices using 

a stable viscous material. Working time was assessed using 

two different cement mixtures and each device was tested at 

least three times with each cement mixture. 

RESULTS 

The working distance of the delivery devices ranged from 

200-480 mm. The syringes offered the shortest working distance 

and highest radiation exposure risk. The Cardinal 

device offered the furthest distance from delivery site. 

Device filling was easiest with Stryker and most difficult 

with Cook device. Cardinal and Cook offered greates ease of 

delivery due to propeller design; however rapid turning may 

lead to failure of device or too rapid cement delivery. Lowest 

torque was seen with Cook due to short tubing plus large 

internal diameter. All of the delivery devices had longer 

working time than syringes. Use of syringes may be prolonged 

if left in chilled water prior to application. Much 

lower psi generated with a 3 cc syringe (400 psi) than 1cc 

syringe (800-1400 psi). The maximum psi of most devices 

ranged between 1500-2500 psi. Significant physical strength 

is needed to generate the higher end pressures of the 

devices.The working times of the delivery devices using 

382 

room temperature cements was between 8-12 minutes. 

Unscrewing and leaking of connections were the main causes 

of device failure at high pressures. Device bursting 

occurred at levels of extremely high applied force that would 

likely not occur within the usual clinical realm. 

CONCLUSION 

All of the methods currently used for cement application in 

the spine have been shown to have advantages and disadvantages. 

Although high pressure is generated using 1 cc 

syringes, the actual working time of the cements (as used in 

the study) is longer with the delivery devices. Syringes for 

cement delivery also require significant physical strength as 

compared to the amount of torque needed to use the delivery 

devices. Use of syringes is cost effective but results in higher 

radiation exposure to the operator. Improvements in 

reducing radiation exposure and in the delivery of viscous 

cement may be beneficial. 

KEY WORDS: Vertebroplasty, cement delivery devices, compression 

fractures 

Poster 182 

Pictorial Review of the Most Serious Causes of Back Pain 

in Children 

Gasparini, F. F. · Branson, H. · Shroff, M. · Ibrahim, M. · 

Drake, J. · Blaser, S. 



PURPOSE 

This exhibit outlines the radiologic findings of back pain in 

children including infection, inflammatory processes; trauma; 

autoimmune and metabolic disease; idiopathic; neoplastic 

and congenital disorders. 


All patients studies from 2001-2004 were evaluated and 

patients with imaging for back pain were identified. 

RESULTS 

The incidence of back pain in children is low, but when present 

is often caused by a serious pathology. Chronic back pain 

may be found in up to 13% of adolescent children, most 

commonly from sports-induced injuries. Although back pain 

is a common problem in adults, it is uncommon in children 

and is often a manifestation of a serious underlying problem. 

It may result from infectious, congenital, neoplastic, and 

traumatic causes. 

CONCLUSION 

Kids are different regarding spinal problems because they 

have special features: relative macrocrania,weak neck muscles, 

flexible ligaments/muscles, unfused synchondrosis, 

growing vertebral bodies, pediatric disk, preexisting anomalies/syndromes 

and imature immune system. 

KEY WORDS: Back pain, children, tumors

Poster 183 

Poster 184 

Atlanto-Occipital Dislocation Revisited: Atlanto-Axial- Symptomatic Congenital Defect of Posterior Arch of the 

Occipital Dislocation on High Resolution Reformatted Atlas 

CT 

Kan, S. 

Lee, C. · Green, E. B. · Fields, T. M. · Given, C. A. 

University of Kentucky 

Lexington, KY 

1 · Hayakawa, K. 1 · Sagiuchi, T. 2 

383 

1 2 Kitasato University, Sagamihara, JAPAN, Kitamoto 

Medical Center, Kitamoto, JAPAN 

PURPOSE 

Atlanto-occipital dislocation (AOD) is often fatal at the time 

of the motor vehicle accident, the most common cause. 

However, aggressive CPR on site and helicopters bring 

many such victims to the ER and CT scanner. Reformatted 

CT of the skull base and spine make the diagnosis obvious 

and easy to make. With CT it becomes more obvious that the 

C1 ring is involved also and is often dislocated as well. It is 

the purpose of this exhibit to revisit AOD and the methods of 

diagnosis emphasizing reformatted CT, and focusing on the 

axial component of the injury. 


Twenty-five AOD of which 22 were fatal from a 20-year 

period were reviewd with the last 5 studied by reformatted 

CT. The other patients had plain radiographs, tomograms, 

and conventional CT, and angiography in 5. 

RESULTS 

The diagnosis of AOD was made by direct visual inspection 

of the AO joint (radiographs, tomograms, CT scans), and 

then by characteristic displacement of craniovertebral structures 

(Power’s ratio, X lines). Eighty-four percent also 

showed abnormality of the C1 ring indicating that it was dislocated—widening 

of the predental space, tilting of the ring, 

superior distraction of the ring with marked C1/2 posterior 

interspinous space widening. The empty C1 ring CT sign is 

also typical of the distraction. 

CONCLUSION 

The violent probably severely twisting forces at the craniovertebral 

junction tears all of the ligaments including the 

tectorial and apical, and separates the occipital condyles of 

the skull from the condylar fossa of the spine at C1. There is 

usually marked to massive prevertebral swelling due to 

bleeding from the torn ligaments. At the same time C1 also 

is twisted loose and is separated from the C2 body. Probably 

the most severe and injurious injury is to the vertebral arteries 

that are injured by the dislocation crimping the arteries as 

they exit from the transverse foramen. This severe injury to 

the craniovertebral junction should be considered to have 3 

components, the AAOD, particularly if there is the rare 

patient survival requiring surgical stabilization as in two of 

our patients. 

KEY WORDS: AOD, atlanto-occipital dislocation, atlantoaxial 

dislocation 

PURPOSE 

To report a rare case of congenital defect of posterior arch of 

atlas with symptoms. 


A 26-year-old male had noticed pain of bilateral upper and 

lower extremities while the patient yawned. He also began to 

have dizziness and unsteady gait. First he had been treated as 

multiple sclerosis because cervical MR imaging showed 

high intensity in the cord on T2-weighted image and contrast 

enhancement at same region. Abnormality of cervical spine 

prompted further radiologic studies. 

RESULTS 

Abnormality of atlas was evaluated with multislice CT and 

this study showed bilateral defects of posterior arch with persistent 

posterior tubercule. CT myelography was performed 

with neutral position (Fig. 1) and extension with open mouth 

position (Fig. 2). The latter position produced the patient 

symptom. With this position, the posterior tubercle moved 

forward and more horizontally oriented. Then spinal cord 

was compressed from behind. After removal of the posterior 

tubercule, the patient’s symptoms have subsided. 

Fig. 1. CT myelography with neutral position. The space 

between the posterior tubercle and C2 posterior arch is narrow. 

Posters

Posters 

Fig. 2. CT myelography in extension with open mouth. The 

posterior tubercle moves forward and horizontally positioned. 

The spinal cord is compressed from behind. 

CONCLUSION 

Plain film of cervical spine is still needed in a case of unexplained 

myelopathy such as our case. To elicit the pathogenesis, 

dynamic study with multislice CT was quite useful. 

REFERENCES 

1. Curraino G, Rollins N, Diethl JT. Congenital defects of the posterior 

arch of the atlas: A report of seven cases including an 

affected mother and son. AJNR Am J Neuroradiol 

1994;15:249-254 

2. Hosalkar HS, Gerardi JA, Shaw BA. Combined asymptomatic 

congenital anterior and posterior deficiency of the atlas. 

Pediatr Radiol 2001;31:810-813 

KEY WORDS: Spine, anomaly, CT 

384

Scientific Exhibits 1-91 



12:00 PM – 9:00 PM 


6:30 AM - 9:00 PM 


6:30 AM – 11:45 AM 

A missing Scientific Exhibit number indicates an 

abstract has been withdrawn. 

Adult Brain 

1-39 

Scientific Exhibit 1 

Clinically Silent Cerebral Ischemic Events after Cardiac 

Surgery: Their Incidence and Regional Vascular 

Occurrence 

Shah, P. N. · Melhem, E. R. · Harris, F. · Acker, M. · Bavaria, 

J. · Pochettino, A. · Woo, J. · McGarvey, M. · Floyd, T. 



PURPOSE 

While clinical stroke occurs in 3-5% of all subjects after cardiac 

surgery, the incidence of silent perioperative cerebral 

ischemic events may approach 40-70% in high risk groups. 

This is especially common among the elderly with additional 

risk factors of previous stroke, carotid disease, hypertension, 

and diabetes mellitus.With diffusion-weighted MR 

imaging, there is increasing evidence that infarction is more 

common after cardiac surgery. 

APPROACH/METHODS 

Thirty-four subjects underwent diffusion-weighted imaging 

before and after cardiopulmonary bypass (CPB) surgery. 

Twenty-one of these subjects were studied on a 1.5 T GE 

system, 10 on a 3.0 T Siemens system, and 3 with both magnets. 

Preoperatively, the following risk factors were 

assessed: Age ≥65, previous stroke or transient ischemic 

attack (TIA), hypertension (HTN), diabetes mellitus (DM), 

atrial fibrillation (AF), smoking, and peripheral vascular disease 

(PVD). Medical records were reviewed for discharge 

diagnosis of stroke, duration of CPB, postoperative AF, and 

type of surgery. MR studies were reviewed by a neuroradiologist 

blinded to pre and postoperative state. Images were 

evaluated for number, size of lesions, and vascular distribution. 

385 

FINDINGS 

Age of the subjects was 67 ± 15 years. Average number of 

risk factors were 3 ± 1, with relative frequencies of: age ≥ 65 

years -25/34 (74%), previous stroke/TIA -8/34 (24%), HTN 

-27/34 (79%), DM -6/28 (21%), smoking -15/34 (44%), 

PVD -5/34 (15%), AF -9/34 (26%). Surgical procedures 

were as follows: Coronary artery bypass grafting (CABG) - 

12/34, aortic valve replacement( AVR) -6/34, mitral valve 

repair/replacement (MVR) -2/34, CABG/AVR -9/34, 

CABG/MVR -4/34, AVR with ascending aortic root replacement 

-1/34. Subjects were imaged postoperatively at 6 ± 2 

days after surgery. New acute infarcts were evident on diffusion-weighted 

imaging in 6/34 subjects (18%). In those with 

new infarcts, the average size of the infarcts was < 1 cm and 

the average number of infarcts was < 3. Infarcts occurred 

more frequently in the right hemisphere (right 11, left 7) with 

90% occurring supratentorially often in the watershed 

regions of the ACA/MCA territories. Less often infarcts 

occurred in defined ACA, MCA, and PCA territories. 

Infarcts > 1 cm occurred only on the left sid , with no infarcts 

> 9 mm on the right side. The occurrence of new infarcts by 

surgical procedure was as follows: AVR (2/6), CABG/AVR 

(3/6), or AVR-Root (1/6). Overall incidence of new infarction 

by diffusion-weighted imaging in procedures involving 

the aortic valve was 6/16 (38%). CBP time was not significantly 

different in those with (122 ± 33 min) than those without 

new infarctions (125 ± 15 min), p = .84. The number of 

clinical risk factors in the group suffering new infarcts was 

not different from the total group (3 ± 1). Fifty percent of 

those with new infarcts had suffered a previous stroke/TIA. 

CONCLUSION 

Clinically silent infarcts are not uncommon after cardiac surgery 

and are quite prevalent after AVR replacement surgery. 

Diffusion-weighted imaging reveals new infarcts in nearly 

40% of subjects after AVR while other types of open-heart 

procedures such as MVR appear relatively immune to this 

complication. Previous stroke/TIA is an important predictor 

of infarction after cardiac surgery, while postoperative AF is 

not as important. Clinical correlates of these lesions may not 

be diagnosed either from inadequate surveillance or from 

occurrence of these lesions in clinically silent “watershed” 

regions. 

KEY WORDS: Stroke, silent ischemia, cardiac surgery 


Cerebral Sinovenous Thrombosis: Imaging Diagnosis 

and Management 

Chen, C. C. C. 1 · Tsai, F. Y. 2 · Shy, C. G. 3 · Chang, P. C. T. 1 · 

Chiang, C. M. 1 · Chen, W. S. 1 · Hung, H. C. 1 · Lee, T. 1 

1Taichung Veteran’s General Hospital, Taichung, TAIWAN 

REPUBLIC OF CHINA, 2University of California Irvine 

Medical Center, Irvine, CA, 3Ping-Tung Christian Hospital, 

Ping-Tung, TAIWAN REPUBLIC OF CHINA 

PURPOSE 

Cerebral sinovenous thrombosis (CSVT) is an uncommon 

disorder that affects the dural venous sinus and cerebral vein. 

In this study, we present the imaging findings of CSVT in 

Scientific Exhibits


386 

CT, MR imaging, MR venography (MRV), and digital sub- Scientific Exhibit 3 

traction angiography (DSA) respectively, in correspondence 

to the clinical appearance and the outcome of management. 

Clinical 3 T Neuroimaging of Brain Infarction: A 

Comparison to 1.5 T 


Thirty-four patients, aged from 2 weeks to 54 years old, with 

clinical manifestations from headache to a progressive neurologic 

deficit and conscious loss, were enrolled in this 

study. Pre and/or postcontrast-enhanced CT were done in 28 

patients. MR study including spin-echo T1 and T2 weighted 

and FLAIR MR imaging were done in 24 patients, and 2D 

TOF and contrast-enhanced MRV (CEMRV) were performed 

in 19 cases. DSA were performed in 18 patients. 

Sixteen patients received anticoagulant therapy via systemic 

intravenous heparinization, and 12 patients received 

endovascular thrombolytic treatment with urokinase combining 

with anticoagulant therapy. 

FINDINGS 

The venous sinuses involved were superior sagittal sinus, 24 

cases; inferior sagittal sinus, 9 cases; straight sinus, 14 cases; 

transverse sinus, 21 cases; sigmoid sinus, 11 cases; cavernous 

sinus, 4 cases; internal cerebral vein and/or vein of 

Galen, 10 cases; and cortical vein, 7 cases. The CT findings 

of intravascular thrombus were presented as cord sign and/or 

dense triangular sign on noncontrast-enhanced CT in 16 

cases. The triangular defect from enhanced dura surrounding 

the thrombus was demonstrated on contrast-enhanced CT in 

10 cases. On MR imaging, the signal intensity of venous 

thrombosis was affected by the degree of residual flow and 

the age of the thrombus. In acute stage (first 3-5 days, 7 

cases), thrombus was isointense on T1-weighted imaging 

and hypointense on T2-weighted imaging. In the subacute 

stage (5-15 days, 12 cases), thrombus appeared hyperintense 

on T1 and T2-weighted imaging. In the later stage (>15 days, 

5 cases), there was progressive heterogeneous signal loss. 

Absence or incomplete filling of dural sinuses was detected 

in 17 cases in combination of 2D TOF with CEMRV and the 

source image of MRV. Digital subtraction angiography 

demonstrated dynamic changes of intracranial circulation in 

CSVT in 16 cases, including absence or incomplete filling of 

dural sinuses or cerebral vein, engorged or tortuous collateral 

veins, reversal of normal venous flow direction or delayed 

venous flow. The abnormal parenchymal change on CT 

and/or both T2-weighted imaging and FLAIR MR findings 

were staged as follows. Stage I: No parenchymal change (6 

cases), Stage II: Brain swelling, no signal change (5 cases), 

Stage III: Hyperintensity with mild to moderate edema (8 

cases), Stage IV: Severe edema, with or without hemorrhage 

(10 cases), Stage V: Massive edema and/or hemorrhage (5 

cases). Following the CSVT, the neurologic deficits were 

resolved in 21 cases, while partial remaining in 10 cases. 

Death was inevitable in 3 cases. Recurrence occurred in 3 

cases. 

CONCLUSION 

Clinical suspicion and excellent neuroimaging are crucial in 

making diagnosis of CSVT. Proper management with anticoagulant 

and/or endovascular thrombolytic therapy is 

mandatory in preventing propagation of thrombosis and 

improving the clinical outcome. 

KEY WORDS: Cerebral sinovenous thrombosis, thrombolysis, 

cerebral venous sinus 

Case, R. S. · Sonnier, H. L. · Runge, V. M. 

Texas A&M University/Scott and White Clinic and 

Hospital 

Temple, TX 

Using patient case material, this exhibit explores the advantages 

of clinical 3 T neuroimaging in acute and early subacute 

brain infarction when compared with 1.5 T. Results are 

based on inter and intrapatient comparisons performed on a 

3.0 T MR unit located immediately adjacent to a similarly 

equipped 1.5 T unit in a major university hospital. Both systems 

are from the same vendor, feature similar 8 channel 

coils for head imaging, and employ identical gradient coils. 

3 T MR imaging represents one of the major forefronts of 

diagnostic neuroradiology today. However, information 

regarding routine clinical application is generally lacking 

due to rapid changes in instrumentation and the relatively 

small installed base. The exhibit focuses on five major 

subtopics. (1) 3 T offers higher signal-to-noise ratio (SNR) 

which can be used either for improved spatial resolution or 

to scan faster, the latter being important for in-patient imaging. 

This has led to a divergence of protocols, with long 

high-resolution scans on one end and fast lower-resolution 

scans on the other. (2) A reduction in slice thickness for routine 

brain imaging is an option with 3 T (with 2D multislice 

imaging), raising the question of whether slice thickness will 

be reduced further in standard clinical practice (for example 

from 5 to 3 mm). (3) Parallel imaging plays an essential role 

at 3 T, in particular for improved image quality with singleshot 

echo-planar imaging (EPI) diffusion-weighted imaging. 

Increased field inhomogeneity at 3 T due to air-tissue interfaces 

is countered by using parallel imaging with IPAT factors 

≥ 2. Implemented in this fashion, diffusion-weighted 

imaging is superior at 3 T as compared to 1.5 T, in particular 

for detection of small acute and early subacute lacunar 

infarcts. (4) Increased susceptibility effects at 3 T lead to better 

susceptibility-based imaging, providing improved perfusion 

imaging in cerebral infarction. (5) Lastly, the increase in 

T1 with field strength leads to improved time of flight MRA. 

Keeping scan time constant (versus 1.5 T), 3 T offers with 

parallel imaging the possibility of either increased spatial 

resolution (with depiction of the lenticulostriate arteries, for 

example) or improved anatomical coverage (routinely 

including PICA in a circle of Willis study). A central theme 

throughout the exhibit is the decision tree one is confronted 

with in moving from 1.5 to 3.0 T. 3 T MR imaging does not 

offer one simple best way to image, but rather at the minimum 

a dichotomy between exam time and spatial resolution. 

KEY WORDS: 3 T, brain, infarction


Brainstem and/or Cerebellar Changes after 

Cerebrovascular Accidents: MR Imaging 

Uchino, A. · Takase, Y. · Nomiyama, K. · Egashira, R. · 

Kudo, S. 

Saga Medical School 

Saga, JAPAN 

PURPOSE 

To illustrate the various possible MR imaging features of the 

brainstem and/or cerebellum after a cerebrovascular accident. 


MR images obtained from many patients with cerebrovascular 

accidents examined in the chronic stage were reviewed 

retrospectively. Images were conventional spin-echo T1weighted, 

fast spin-echo T2-weighted, or fast fluid-attenuated 

inversion recovery images obtained by one of two 1.5 T, 

a 1.0 T or a 0.3 T scanner. 

FINDINGS 

Several types of secondary degeneration in the brainstem 

and/or cerebellum were apparent on MR images after the 

cerebrovascular accidents. These included: 1) ipsilateral 

nigral degeneration after widespread middle cerebral artery 

territory infarction, 2) atrophy of the ipsilateral superior 

cerebellar peduncle, contralateral rubral degeneration, and 

contralateral inferior olivary degeneration after dentate 

nucleus hemorrhage, 3) ipsilateral inferior olivary degeneration 

after pontine tegmentum hemorrhage, 4) contralateral 

cerebellar atrophy after extensive supratentorial infarctions, 

5) wallerian degeneration of the pyramidal tract and frontopontine 

tract in the brainstem after frontal lobe infarction, 6) 

wallerian degeneration of the bilateral pontocerebellar tracts 

after ventromedial pontine infarction, 7) wallerian degeneration 

of the ipsilateral pontocerebellar tract after ventrolateral 

pontine infarction, and 8) ipsilateral cerebellar atrophy after 

middle cerebellar peduncle hemorrhage. 

CONCLUSION 

These types of secondary degeneration observed in the 

brainstem and/or cerebellum after a cerebrovascular accident 

should not be misdiagnosed as other serious diseases. 

KEY WORDS: Cerebrovascular accident, secondary degeneration, 

MR imaging 


Review of Stroke Mimics in the Differential Diagnoses of 

Stroke 

Ederies, M. A. · Gada, V. S. · Aviv, R. I. 

Charing Cross Hospital 


PURPOSE 

We present a pictorial review of the array of conditions that 

mimic stroke, both from a clinical and a radiologic perspective. 

The aim of this presentation is to demonstrate to trainee 

387 

neuroradiologists signs visible on the presentation imaging 

enabling distinction of the underlying causes of the apparent 

focal neurologic deficit and providing a workable differential 

diagnosis of entities encountered. 


Although the clinical diagnosis of acute ischemic stroke is 

often straightforward, we have encountered several conditions 

that have presented as radiologic or clinical stroke 

mimics. We reviewed our radiologic database for lesions that 

presented either clinically or radiologically as stroke where 

an alternative diagnosis was made based on further radiologic 

imaging. We highlight key radiologic signs and discriminating 

sequences or modalities that may distinguish these 

conditions. 

FINDINGS 

The clinical diagnosis was often problematic because of the 

presence of several stroke subtypes and of nonvascular disorders 

that may have similar clinical and radiologic features 

to stroke. In agreement with others (Libman, et al), we 

encountered several mimics including unrecognized systemic 

infections, intracranial tumors, multiple sclerosis, posterior 

reversible encephalopathy syndrome, cerebritis, cerebral 

abscess, subdural hemorrhages, cavernous malformations, 

metabolic and toxic disturbances, and seizure disorders. 

CONCLUSION 

Thorough and systematic review of presentation imaging 

and awareness of key clinical signs and the temporal evolution 

of the conditions that form the differential diagnosis, 

enable accurate distinction of stroke mimics from true vascular 

events. Supplementary sequences and modalities further 

facilitate the diagnosis. 

REFERENCES 

1. Libman RB, Wirkowski E, Alvir J, et al. Conditions that mimic 

stroke in the emergency department. Implications for acute 

stroke trials. Arch Neurol 1995;52(11):1119-1122 

KEY WORDS: Stroke mimics, differential diagnosis 


Differential Diagnosis of “Holohemispheric” Pathology 

Moritani, T. · Smoker, W. R. K. · Sato, Y. · Lee, H. · Maley, 

J. · White, M. L. 


Iowa City, IA 

PURPOSE 

1. Review the various pathologic conditions associated with 

unilateral involvement of an entire cerebral hemisphere. 2. 

Present the pathology and pathophysiology of various “holohemispheric” 

processes. 


Various imaging modalities, including plain radiography, 

CT, MR imaging and MR angiography, were used to evaluate 

74 patients with a variety of pathologic conditions affecting 

an entire cerebral hemisphere. These included hemispheric 

involvement related to status epilepticus, hemicon- 

Scientific Exhibits


vulsion-hemiplegia epilepsy (HHE) syndrome, Rasmussen 

encephalitis, Sturge-Weber syndrome, herpes simplex 

encephalitis, human herpes virus 6 encephalitis, large cerebral 

infarction, sickle cell disease, cerebral hyperperfusion 

syndrome, hemimegalencephaly, and gliomatosis cerebri. 

FINDINGS 

Not surprisingly, MR imaging was most useful for limiting 

the differential diagnosis of holohemispheric pathology, 

especially diffusion-weighted imaging which showed 

decreased ADC values in the early phase of hemispheric 

involvement related to seizures, including those associated 

with HHE syndrome. Gadolinium enhancement also was 

useful in the differential diagnosis, showing various 

enhancement patterns (i.e., gyriform leptomeningeal 

enhancement posteriorly in Sturge Weber syndrome vs gyrus 

enhancement in periictatic encephalopathy). We discuss the 

definition and pathophysiology of these diverse entities, 

including “Dyke-Davidoff-Masson syndrome.” 

CONCLUSION 

The differential diagnosis of “holohemispheric” pathology is 

reviewed using various imaging modalities. We discuss the 

definition and pathophysiology of these various conditions. 

KEY WORDS: Holohemispheric, MR imaging, computed 

tomography 


Posterior Reversible Encephalopathy Syndrome: A 

Commom Entity 

Gostenik, K. · Singla, R. · Magalhaes, A. C. A. · Zak, I. · 

Kish, K. 


Detroit, MI 

PURPOSE 

Posterior reversible encephalopathy syndrome (PRES) is a 

clinicoradiologic syndrome that is seen in a variety of clinical 

conditions such as acute or chronic renal disease, 

preeclampsia and eclampsia, vasculitis, thrombotic thrombocytopenic 

purpura, hemolytic-uremic syndrome and with 

drug toxicity. It often occurs in association with hypertension 

and/or use of immunosuppressive drugs. Patients frequently 

present with headaches, mental status changes, visual 

disturbances and seizures. Characteristic findings on MR 

imaging include areas of bilateral hyperintensity on T2 in the 

parieto-occipital subcortical white matter as well as the overlying 

cortical region. Diffusion-weighted MR images are 

often hypointense or isointense with an increase in the apparent 

diffusion coefficient (ADC). The purpose of our study 

was to further characterize the clinical diagnoses as well as 

the MR findings in patients who were identified as having 

PRES. 


We performed a retrospective study of MR imaging examinations 

in 10 patients with PRES. MR brain imaging studies 

were performed during the period from July 2001- 

September 2004 in Harper Hospital/Detroit Medical Center. 

388 

To evaluate brain lesions MR exams were done in a 1.5 T 

system imager (GE, Milwaukee). T1, T2, FLAIR and diffusion-weighted 

images were obtained. 

FINDINGS 

All patients were female, age range 18-47 years, mean age 

27.8 years. Clinical presentation of the patients included 

hypertension/renal failure, postpartum hypertension, 

eclampsia and a history of cyclosporine use. MR studies 

showed increased signal intensity on T2 images in the gray 

and subcortical white matter of the occipito-temporo-parietal 

regions as well as the basal ganglia. 

CONCLUSION 

Posterior reversible encephalopathy syndrome is not a single 

disease entity. It is an increasingly recognized clinicoradiologic 

syndrome that is associated with a variety of clinical 

disorders often related to hypertension and immunosuppressive 

treatment. The clinical presentation in the five patients 

we studied as well as the MR findings of predominately posterior 

T2-weighted signal intensity abnormalities in the subcortical 

white and gray matter are characteristic of previously 

reported cases. 

KEY WORDS: Reversible encephalopathy, eclampsia, hypertension 


Neuroferritinopathy: Genetics, Clinical Features, MR 

Imaging, and Neuropathology 

Coulthard, A. 1 · Chinnery, P. 2 · Crompton, D. 3 · Curtis, A. 4 · 

Morris, C. 4 · Birchall, D. 2 · Bates, D. 3 · Burn, J. 4 

1 Royal Brisbane Hospital, University of Queensland, 

Brisbane, AUSTRALIA, 2 Newcastle General Hospital, 

Newcastle upon Tyne, UNITED KINGDOM, 3 Royal Victoria 

Infirmary, Newcastle upon Tyne, UNITED KINGDOM, 

4 Institute of Human Genetics, University of Newcastle upon 

Tyne, Newcastle upon Tyne, UNITED KINGDOM 

PURPOSE 

Neuroferritinopathy is a recently described progressive neurodegenerative 

disorder arising in adult patients secondary to 

a mutation in the ferritin light chain gene (FTL) on chromosome 

19. The mutation leads to abnormal brain iron deposition 

in those areas of the brain normally associated with 

increased iron content. This exhibit describes the genetics, 

clinical presentation and features, MR imaging appearances, 

and neuropathology of neuroferritinopathy. 


Patients positive for the FTL gene mutation were identified. 

Patients were examined and clinical charts were reviewed by 

three neurologists, who documented patterns of clinical presentation 

and progression over time. Cranial MR examinations 

were reviewed by two neuroradiologists. Structural 

abnormalities were documented. Signal intensity (SI) 

changes were evaluated using the SI of white and gray matter 

as a comparator. Histopathologic material was available 

in three cases and was reviewed by a neuropathologist.

FINDINGS 

Twenty-seven patients with neuroferritinopathy were identified. 

Clinical presentation was between fourth and sixth 

decade with Parkinson-like bradykinesis, choreiform movement 

disorder, or dystonia. Typically serum iron was normal 

and serum ferritin low. Nineteen patients (10 females, 9 

males) had 27 cranial MR examinations from 1992-2004 

(age 35-69 years, mean 49 years). Seven patients had two, 

and one three separate examinations. All examinations were 

abnormal, with appearances ranging from abnormal low SI 

of deep brain nuclei visualized on T2* images in relatively 

asymptomatic patients, to gross cystic degeneration of the 

corpus striatum in severely affected individuals. A “tigereye” 

appearance was seen in several cases, often involving 

red nucleus or dentate nucleus. Abnormal low SI was noted 

in relation to motor cortex, internal capsule, subthalamic 

nucleus, and thalamus in other cases. Patients with repeat 

scans showed MR imaging evidence of disease progression. 

Histopathologic material was available from four patients. 

Glial accumulation of ferritin and iron was prominent in 

oligodendrocytes, astrocytes and microglia, with increased 

cytoplasmic and nuclear iron and ferritin accumulation. 

Neuronal and glial cell loss followed the distribution of iron 

in the normal brain. A prominent axonopathy was noted, 

with ubiquitin-positive axonal spheroids, and neurons showing 

evidence of Lewy body formation and accumulation of 

hyperphosphorylated tau. 

CONCLUSION 

Neuroferritinopathy leads to progressive pathologic changes 

in those brain structures associated with increased concentration 

of iron in the normal brain. Histopathologic changes 

are mirrored by MR imaging changes. T2* images are recommended 

for assessment of early stage disease. Clinical 

features may mimic other neurodegenerative conditions. 

KEY WORDS: Neuroferritinopathy, iron, ferritin 


Radiologic Spectrum of Uncommon Inflammatory 

Demyelinating Diseases 

Rovira, A. 1 · Rovira-Gols, A. 2 · Grivé, E. 1 · Rio, J. 1 · Alonso, 

J. 1 · Schorlemmer, C. 1 

1 2 Vall d’Hebron, Barcelona, SPAIN, UDIAT, Parc Taulí, 

Sabadell, SPAIN 

PURPOSE 

To review the clinical and radiologic findings of uncommon 

idiopathic inflammatory demyelinating diseases. 


Idiopathic inflammatory demyelinating diseases (IDD) represent 

a broad spectrum of disorders that can be differentiated 

on the basis of clinical, radiologic and pathologic findings. 

Relapsing-remitting and secondary progressive multiple 

sclerosis (MS) represents the common forms of IDD. 

Uncommon forms of IDD can be classified according to 

severity, clinical findings and temporal course: fulminant 

IDD, such as the Marburg variant of MS, Balo´s concentric 

sclerosis, Schilder´s disease, and acute disseminated 

encephalomyelitis (ADEM); monosymptomatic IDD, such 

389 

as transverse myelitis, isolated optic neuritis or brainstem 

syndrome; and recurrent IDD, which has a restricted topographical 

distribution and includes Devic´s neuromyelitis 

optica (NMO), and relapsing myelitis. Other groups of IDD 

include progressive disorders such as primary progressive 

MS and progressive ataxia, the benign form of MS, and 

pseudotumoral IDD, which has different forms of presentation 

and temporal course (fulminant, monophasic, relapsingremitting). 

In this scientific exhibit we will present representative 

cases of the uncommon forms of IDD, collected from 

our large series of more than 1500 IDD cases, in which 

relapsing and progressive forms of MS were by far the most 

common diagnoses. 

FINDINGS 

Uncommon forms of IDD vary in their clinical findings, 

regional distribution, pathology, severity, and temporal 

course, and consequently in their radiologic findings. MR 

imaging of the brain and spine is the imaging technique of 

choice for diagnosing these disorders, and together with clinical 

findings, can properly classify IDD. 

CONCLUSION 

This classification is important for etiological, pathologic, 

immunopathogenetic, prognostic and treatment considerations. 

KEY WORDS: Multiple sclerosis, demyelinating diseases, 

MR imaging 


Lesions of the Corpus Callosum: A Review 

Mamourian, A. C. 1 · Pekala, J. S. 2 

1 2 Dartmouth-Hitchcock Medical Center, Lebanon, NH, Duke 

University Medical Center, Durham, NC 

The corpus callosum is the largest commissure in the brain 

and as such provides the primary pathway for interhemispheric 

communication. There are a variety of diseases that 

selectively may involve the corpus callosum and for this 

exhibit they will be divided into three categories: disorders 

that decrease the size of the corpus callosum, mass lesions 

(tumor, MS, and infarct), and signal abnormalities with no 

mass effect. The exhibit will review the blood supply to the 

corpus callosum and anatomical relationships with the falx 

which most likely influence the pattern of lesions with trauma 

and infrequent occurrence of infarcts. The mnemonic 

“DART” is introduced to serve as a reminder of the most 

common causes of signal abnormality without mass effects: 

demyelination/drugs, aging, radiation, and trauma. The 

exhibit will review some features of lymphoma and gliomas 

that may allow discrimination on the basis of MR appearance. 

KEY WORDS: Corpus callosum, aging, tumor 

Scientific Exhibits



Acquired Lesions of the Corpus Callosum: MR Imaging 

Uchino, A. · Takase, Y. · Nomiyama, K. · Egashira, R. · 

Kudo, S. 

Saga Medical School 

Saga, JAPAN 

PURPOSE 

To identify the various types of acquired lesions that can 

appear in the corpus callosum on MR images. 


We reviewed the images obtained from many patients with 

acquired corpus callosal lesions diagnosed by MR imaging. 

Images were conventional spin-echo T1-weighted, fast spinecho 

T2-weighted, or fast fluid-attenuated inversion recovery 

images obtained by one of two 1.5 T, a 1.0 T or a 0.3 T 

scanner. 

FINDINGS 

The acquired corpus callosal lesions resulted from various 

diseases or conditions: 1) infarction, 2) multiple sclerosis, 3) 

acute disseminated encephalomyelitis, 4) Marchiafava- 

Bignami disease, 5) transient splenial lesion, 6) hypertensive 

encephalopathy, 7) wallerian degeneration after hemispheric 

damage, 8) diffuse axonal injury, 9) iatrogenic corpus callosal 

injury, 10) treated hydrocephalus, 11) lymphoma, 12) 

metastasis, 13) glioblastoma, 14), gliomatosis cerebri, and 

15) dilated Virchow-Robin spaces. 

CONCLUSION 

Acquired corpus callosal lesions include vascular, demyelinating, 

traumatic, neoplastic, and miscellaneous lesions. MR 

imaging is useful for the detection and differential diagnosis 

of corpus callosal lesions. 

KEY WORDS: Corpus callosum, acquired lesion, MR imaging 


Transthyretin-Related Familial Amyloid Polyneuropathy: 

Prospective Evaluation of Cerebrospinal 

Fluid Enhancement on Serial T1-Weighted and Fluid- 

Attenuated Inversion Recovery Images following 

Intravenous Contrast Administration 

Hirai, T. 1 · Ando, Y. 1 · Kitajima, M. 1 · Hayashida, Y. 1 · Korogi, 

Y. 2 · Yamashita, T. 1 · Yamashita, Y. 1 




JAPAN 

PURPOSE 

Various meningeal lesions present with cerebrospinal fluid 

(CSF) enhancement on contrast-enhanced fluid-attenuated 

inversion recovery (FLAIR) MR images. Some types of 

transthyretin-related familial amyloid polyneuropathy (FAP) 

produce leptomeningeal lesions. The purpose of this study 

390 

was to evaluate CSF enhancement on serial T1-weighted and 

FLAIR images following intravenous contrast-material 

administration in patients with transthyretin-related FAP. 


We serially studied T1-weighted and FLAIR images of the 

brain before, immediately after, and 3, 6, and 24 hours after 

the intravenous administration of contrast material in six 

patients with genetically proven transthyretin-related FAP. 

By consensus, two radiologists assessed the presence, 

degree, and extent of enhancement of the CSF, leptomeninges, 

brain parenchyma, and other structures. 

FINDINGS 

In three of six patients, marked CSF enhancement was 

observed on the FLAIR images at 3 and 6 hours, and on T1weighted 

images at 3 hours after contrast administration 

(Fig. 1). In these three patients, leptomeningeal enhancement 

was evident only on FLAIR images immediately after contrast 

administration. The labyrinth and vitreous body was 

also enhanced on postcontrast delayed MR images of these 

three patients. No images demonstrated parenchymal 

enhancement of the brain in any of the six patients. 

CONCLUSION 

In patients with transthyretin-related FAP, contrast material 

may leak into the CSF. Overall, FLAIR images showed more 

definite enhancement than T1-weighted images and are 

therefore more useful for the evaluation of abnormal 

enhancements. 

KEY WORDS: Amyloidosis, contrast-enhanced FLAIR, CSF 

enhancement 


Imaging Features of Ganglioglioma of the Temporal 

Lobe 

Adachi, Y. 1 · Yagishita, A. 2 

1 2 University of Yamanashi, Tamaho, JAPAN, Tokyo 

Metropolitan Neurological Hospital, Fuchu, JAPAN 

PURPOSE 

Ganglioglioma is an uncommon neoplasm of the central 

nervous system, occasionally seen in the temporal lobe. 

Ganglioglioma usually is associated with chronic epilepsy in 

children and young adults and therefore, always should be 

considered as a diagnosis in epileptic patients. The purpose 

of this study was to characterize the radiologic features of 

temporal lobe ganglioglioma. 


Twenty-two consecutive patients with a histopathologic 

diagnosis of ganglioglioma in the temporal lobe were analyzed 

retrospectively for their CT and MR imaging. The following 

findings were evaluated: exact location of the tumor, 

density on CT, signal intensity on T1- and T2-weighted 

images, presence of cystic component, calcification, and 

mass effect. These MR and CT findings were compared with 

those of other temporal tumors in the same period, including 

3 DNTs, 2 astrocytomas, and 1 PXA.

FINDINGS 

Gangliogliomas were located in the uncal part of the hippocampus 

(n = 7), the parahippocampul gyrus (n = 5), amygdale 

(n = 3), hippocampus (n = 1) and other tumporal part (n 

= 5). CT scans showed isodense to surrounding brain 

parenchyma in 16 patients (73%). On MR images, 11 gangliogliomas 

were isointense lesion on T1-weighted images. 

In the other tumors, only one case showed isodense on CT 

scan and isointense on T1-weighted images. Cystic component 

appeared to be separate from the main solid lesion. On 

T2-weighted images the lesions appeared to be hyperintense 

in 20 cases, isointense in 2 cases. Eleven cases of gangliogliomas 

demonstrated calcification (50%). Tumors located 

in the uncal part showed calcification, with higher frequency 

(71%) than lesion in other region (8%). Ganglioglioma 

appeared as a cystic lesion in 45% cases. Cyst appearance 

was more common in younger patients, exclusively in 

patients20 years old or less except one case, who was 35 

years old. The mean age of patients with cystic component 

was 11.1 years old, without cystic component was 21.1 years 

old. Similarly, calcification and contrast enhancement was 

more common in the young patient. 

CONCLUSION 

A cystic lesion in young patients and calcification of uncal 

part are a characteristic feature of ganglioglioma. 

Ganglioglioma has a tendency to demonstrate isodensity on 

CT and isointensity on MR images, in contrast with other 

temporal tumors, which usually show low density or signal 

intensity. 

KEY WORDS: Ganglioglioma, epilepsy, temporal lobe tumor 


Transient Lesion in the Splenium of the Corpus 

Callosum with Restricted Diffusion: An Enigma 

Maeda, M. 1 · Tsukahara, H. 2 · Terada, H. 3 · Takeda, K. 1 

1 2 Mie University School of Medicine, Tsu, JAPAN, Fukui 

University School of Medicine, Matsuoka, JAPAN, 3Toho University School of Medicine, Tokyo, JAPAN 

PURPOSE 

Transient lesion in the splenium of the corpus callosum 

(SCC) has been reported to be found uncommonly in MR 

examinations. The purpose of this exhibit is to illustrate 

characteristic MR imaging findings and clinical features in a 

wide variety of diseases or conditions and to discuss possible 

mechanisms of this unique phenomenon. 


Transient lesion in the SCC with restricted diffusion was 

found in nine patients with a wide variety of central nervous 

system diseases, including 3 patients with clinically mild 

encephalitis/encephalopathy, 2 patients with hypoxic 

ischemic encephalopathy, 1 patient with epilepsy receiving 

antiepileptic drugs, 1 patient with depression receiving 

antiepileptic drugs and antidepressant drugs, and 2 patients 

with neoplasm (leukemia and spinal meningeal melanocytosis). 

The last 2 patients had never had antiepileptic drugs or 

epilepsy. MR imaging was performed in all patients, including 

T1-weighted, T2-weighted, FLAIR and diffusion- 

391 

weighted imaging. Contrast-enhanced images were obtained 

in 5 patients. The shape and size of the SCC lesion, presence 

or absence of additional brain lesions excluding a SCC 

lesion, presence or absence of neurologic symptoms at the 

time of initial MR examination, and prognosis (if neurologic 

symptoms were present) were reviewed in these patients. 

FINDINGS 

Transient lesion in the SCC was oval (less than 2 cm) in 7 

patients and extended in length more than 2 cm in 2 patients 

(2 with hypoxic ischemic encephalopathy). Brain lesions 

excluding a SCC lesion were found in 3 patients (1 with mild 

encephalopathy and 2 with hypoxic ischemic encephalopathy). 

Neurologic symptoms were seen in 5 patients (3 with 

mild encephalitis/encephalopathy and 2 with hypoxic 

ischemic encephalopathy). Prognosis was poor in 2 patients 

(2 with hypoxic ischemic encephalopathy) and excellent in 3 

patients (3 with mild encephalitis/encephalopathy). Possible 

causes or mechanisms are postulated; (e.g., intramyelinic 

edema or the influx of inflammatory cells and macromolecules, 

neurotoxicity of antiepileptic drugs, seizure per se, or 

a combination of these). However, these hypotheses do not 

apply to 2 patients with neoplasm, because these 2 patients 

did not have seizure, antiepileptic drugs, or 

encephalitis/encephalopathy. 

CONCLUSION 

It is noted that transient lesion in the SCC with restricted diffusion 

can occur in patients with a wide variety of diseases 

or conditions, and even in patients without epilepsy or 

encephalitis/encephalopathy. Exact mechanisms of this phenomenon 

remain uncertain and an enigma. Complete 

reversibility without specific treatment suggests that unnecessary 

invasive examinations and therapeutic intervention 

should be avoided in a clinical setting. 

REFERENCES 

1. Kim SS, Chang KH, Kim ST, et al. Focal lesion in the splenium 

of the corpus callosum in epileptic patients: antiepileptic 

drug toxicity? AJNR Am J Neuroradiol 1999;20:125-129 

2. Takanashi J, Barkovich AJ, Yamaguchi K, Kohno Y. Influenzaassociated 

encephalitis/encephalopathy with a reversible 

lesion in the splenium of the corpus callosum: a case report 

and literature review. AJNR Am J Neuroradiol 2004;25:798- 

802 

KEY WORDS: Splenium, corpus callosum, diffusion-weighted 

imaging 

Scientific Exhibits



Diffusion-Weighted Imaging of Lesions in the Corpus 

Callosum 

Moritani, T. · Smoker, W. R. K. · Sato, Y. · Maley, J. · Lee, 

H. · White, M. L. 


Iowa City, IA 

PURPOSE 

1. Illustrate diffusion-weighted imaging findings of various 

pathologic conditions affecting the corpus callosum. 2. 

Understand the pathology and pathophysiology of these varied 

lesions. 


We reviewed the diffusion-weighted imaging findings in 

various pathologies involving the corpus callosum on more 

than 100 cases. These included infarction, hypoxic ischemic 

encephalopathy, CNS vasculitis and vasculopathy, multiple 

sclerosis, acute disseminated encephalomyelitis, progressive 

multifocal leukoencephalopathy, various tumors (lymphoma, 

glioma, etc.), hemolytic uremic syndrome, anticonvulsant 

drug- or seizure-induced lesions, diffuse axonal 

injury, toxic or metabolic leukoencephalopathy, and 

Marchiafava-Bignami disease. 

FINDINGS 

Areas of decreased or increased ADC values were observed 

in a variety of these pathologic conditions. Differential diagnosis 

is based on the ADC values as well as the location and 

shape of the corpus callosal lesion: focal or diffuse, symmetric 

or asymmetric, multiple or solitary, and whether there 

is involvement of the surrounding callosal white matter 

fibers. We also discuss the pathology and pathophysiology of 

these corpus callosal lesions. 

CONCLUSION 

The combination of ADC values and the location/shape of 

corpus callosal lesions permits narrowing of the differential 

diagnoses. 

KEY WORDS: Corpus callosum, MR imaging, diffusionweighted 

imaging 


Trigeminal Nerve: Imaging of Normal Anatomy and 

Pathologic Conditions 

Bermudez, P. · Carvajal, A. · Rovira-Gols, A. · Zauner, M. · 

Prenafeta, M. 

UDIAT-CD Corporació Sanitària Parc Taulí 


PURPOSE 

The aim of this exhibit is: 1. To explain and illustrate trigeminal 

nerve anatomy in its pathway. 2. To describe the imaging 

findings for different pathologies involving the trigeminal 

nerve in its pathway, based on its anatomical topography 

(brain stem, cisternal segment, parasellar, and skull basis 

location). 

392 


A pictorial review of the anatomical details of the trigeminal 

nerve, including the most significant cases of pathologies 

involving the trigeminal nerve at different points along its 

pathway from a retrospective review of the radiologic examinations 

(conventional radiology, CT, but basically MR 

imaging) carried out in our center. 

FINDINGS 

Understanding the different pathologies that can affect the 

trigeminal nerve requires thorough knowledge of the anatomy 

of this cranial nerve, from its real origin in the brainstem 

through its apparent origin and along its pathways through 

the cisterns and parasellar region. The different intrinsic and 

extrinsic pathologies that can cause trigeminal neuralgia 

have particular characteristics depending on their topography 

and etiology. 

CONCLUSION 

Knowledge of the anatomy and pathologic states of the 

trigeminal nerve is essential to understand imaging findings 

in the different pathologies that affect this cranial nerve and 

enable a correct differential diagnosis among the different 

pathologic entities. MR imaging is the method of choice for 

evaluating the trigeminal nerve. 

KEY WORDS: Brain anatomy, trigeminal nerve, MR imaging 

Scientific Exhibit 16A, See page 428 


Pebbles, Boulders, and Petrified Byways: Trekking the 

Terrain of Intracranial Calcifications 

Behr-Ventura, D. · Lev, S. · Waltz, D. · Ackerman, A. 

North Shore University Medical Center 

Manhasset, NY 

PURPOSE 

To review the radiographic spectrum of intracranial calcifications, 

emphasizing an approach based on anatomical location 

and morphology. 


We retrospectively reviewed the radiologic studies of 

patients with intracranial calcifications. Categorization was 

performed according to anatomical location - intraventricular, 

subependymal, parenchymal, and extraxial (with further 

subclassifications). Additional consideration was given as to 

whether the calcifications were associated with cystic or 

solid lesions. We demonstrate an array of “look-a-like” entities 

and unusual lesions, analyzing key morphologic features 

which aid in the differential diagnosis. The role of imaging 

modalities, including multislice CT and MR imaging is 

explored. 

FINDINGS 

Intracranial calcified lesions represent a broad spectrum of 

pathology. Intraventricular lesions include xanthogranuloma 

of the choroid plexus, meningioma, central neurocytoma, 

teratoma, and ependymoma. Subependymal/periventricular

lesions include hereditary entities such as tuberous sclerosis 

(to be differentiated from nodular heteropia, which does not 

calcify) and congenital infectious etiologies such as TORCH 

syndromes. Intraparenchymal entities can be divided into 

cortical-based, basal ganglia, and white matter lesions. 

Cortical calcifications may be congenital, as with Sturge- 

Weber syndrome, or postinfarction related. Calcification 

within the white matter is seen postradiation and/or 

chemotherapy. Other intraparenchymal etiologies include 

calcified infectious (parasitic, fungal) and neoplastic lesions 

(oligodendrogliomas, metastases), and calcifications related 

to underlying vascular malformations (AVMs, cavernous 

angiomas). We show an unusual example of a partially calcified 

and thrombosed giant vein of Galen malformation. 

Basal ganglia calcifications may be seen in toxic, metabolic, 

and congenital disorders (hypoparathyroidism, pseudohypoparathyroidism, 

carbon monoxide intoxication, Fahr’s disease, 

and Down’s syndrome). Sellar/parasellar calcifications 

commonly are associated with aneurysms, craniopharyngiomas, 

and meningiomas. Pineal lesions that calcify include 

cysts, pineal and germ cell origin neoplasms. 

Pinealocytomas “explode” gland calcification while germinomas 

and metastases “engulf” gland calcifications. 

Calcifications within the extraaxial spaces are varied. We 

present an unusual case of a chronic subdural hemorrhage 

with a residual framework of heavily calcified dura. Another 

case is presented of a patient with hyperparathyroidism, with 

excessive intracranial calcifications, including significant 

involvement of the falx and tentorium. Dural-based calcifications 

also can be seen in association with meningiomas 

and other dural neoplasms, such as meningiomatosis and 

rarely hemangiopericytoma. Subarachnoid space calcifications 

include the sequela of infectious or inflammatory 

processes, such as tuberculosis and neurocysticercosis, as 

well as dermoids and lipomas. 

CONCLUSION 

We demonstrate an approach for assessing intracranial calcifications, 

emphasizing the importance of lesion location and 

morphology. 

KEY WORDS: Calcifications, intracranial 


MR Imaging of Multicentric Brain Gliomas and 

Gliomatosis Cerebri 

Colosimo, C. 1 · Cerase, A. 2 · Di Lella, G. M. 3 · Caulo, M. 1 · 

Tartaro, A. 1 · Maira, G. 3 

1 University, Chieti, ITALY, 2 Policlinico, Siena, ITALY, 

3 Universita Cattolica del Sacro Cuore, Rome, ITALY 

Withdrawn 

PURPOSE 

The purposes of this exhibit are to: 1) define frequency of 

gliomatosis cerebri (GC) and multicentric gliomas (MG); 2) 

describe their most common MR imaging features, and 3) 

report on evolution of clinical and MR findings by presenting 

the results of the retrospective review of Gd-enhanced 

MR imaging of 53 out of 610 patients with pathologically 

proven brain gliomas (BG). 

393 


Fifty-three patients (8.6%; 31 male and 22 female) were considered 

to have true MG, since different tumor sites could not 

be considered disseminated through white matter tracts, due 

to satellite formation, nor explained by dissemination 

through CSF and perivascular spaces. Seven of fifty-three 

patients were under 18 years of age at diagnosis. Perfusionweighted 

imaging was obtained in 10 patients, fMRI in 7, 

proton MRS in 8, and FDG-PET in 6. 

FINDINGS 

Final pathologic diagnosis was GC in 15 patients, and MG in 

38. Thirty out of 38 MG were entirely astrocytic, 5 mixed, 

and 3 exclusively oligodendroglial. Duration of clinical and 

MR follow up ranges from 15 months to 14 years. 

CONCLUSION 

In conclusion, true MG and GC are more frequent than generally 

considered. In most cases, differential diagnosis is not 

possible by MR imaging. Despite poor prognosis in most 

patients, there is a considerable percentage of low-grade 

tumors. Early recognition of anaplastic evolution and/or evolution 

toward glioblastoma multiforme by MR imaging or 

perfusion-weighted imaging may allow better prognosis. 

Posters 

KEY WORDS: Multicentric brain gliomas, gliomatosis cerebri, 

MR imaging 


Assessment of Brain Tumors with Diffusion Tensor 

Imaging 

da Cruz, L. H. 1 · Magalhães, F. V. 1 · Ferreira, F. B. 1 · 

Domingues, R. C. 1 · Nogueira, J. 2 · Domingues, R. C. 1 

1 Clínica de Diagnóstico por Imagem-Barrashopping/Multi- 

Imagem Ressonancia, Rio de Janeiro, BRAZIL, 2 Instituto 

Nacional de Câncer, Rio de Janeiro, BRAZIL 

PURPOSE 

The assessment of brain tumors with diffusion tensor imaging. 


Conventional and diffusion tensor imaging (DTI) were performed 

in 19 patients with intracranial neoplasm (9 men; 10 

women; mean age 57 years) and 17 age- and sex-matched 

control subjects in a 1.5 T clinical scanner. Diffusion tensor 

imaging was obtained with echo-planar SE with diffusion 

gradient b values of 0 and 1000 sec/mm 2 applied in six different 

directions. Fractional anisotropy (FA) values were 

obtained from DTI data. Identical regions of interest (ROIs) 

were placed by a single observer on the tumor core, peritumoral 

abnormal T2-weighted signal intensity, contralateral 

normal-appearing white matter (NAWM) and white matter 

in controls subjects. 

FINDINGS 

Brain tumor cases were classified into: low-grade glioma (n 

= 8), high-grade glioma (n = 6), and metastasis (n = 5). 

Difference between FA values of the peritumoral T2-weighted 

signal intensity abnormality was observed in high-grade 

gliomas (0.125 to 0.203, mean 0.186) and vasogenic edema 

Scientific Exhibits


surrounding the metastatic lesions (0.189 to 0.273, mean 

0.251) when compared with the NAWM (0.380-0.572, mean 

0.447). When the tumor core was analyzed, no significant 

difference was observed between the FA values of metastatic 

lesions, high-grade or low-grade gliomas. Diffusion tensor 

imaging patterns of neoplasm involvement of the main fiber 

tracts were divided into: deviated, infiltrated, edematous, 

and destroyed. Five patients with low-grade glioma had 

deviation of the main fiber tracts; the patterns edematous and 

deviation were found in all cases of metastasis. Five patients 

with high-grade tumors presented with infiltration of tracts 

and in three patients destruction. 

CONCLUSION 

Changes of tumor core FA values were unable to distinguished 

low-grade gliomas from high-grade gliomas and 

metastasis. Although no statistical difference was demonstrated, 

FA values were lower in the peritumoral T2-weighted 

signal intensity abnormality when compared with the 

vasogenic edema surrounding metastasis. Diffusion tensor 

imaging appears to have the possibility to add important 

information to presurgical planning. While experience is 

limited, DTI appears to provide useful local information 

about the structures near the tumor, and this appears to be 

useful in planning. While DTI has some limitations, its 

active investigation and further study are clearly warranted. 

KEY WORDS: Brain neoplasm, diffusion tensor imaging, 

tractography 


MR Findings of Gliosarcoma 

Oka, M. · Hiwatashi, A. · Ohgiya, Y. · Ekholm, S. · 

Westesson, P. 

University of Rochester Medical Center 

Rochester, NY 

PURPOSE 

Gliosarcoma is a rare subtype of glioblastoma and composed 

of neoplastic glial cells with a sarcoma component. The purpose 

of our study is to present MR characteristics of gliosarcoma, 

including diffusion and perfusion imaging. 


MR findings of six patients with pathologically documented 

gliosarcoma were analyzed retrospectively. Diffusionweighted 

imaging had been performed in 4 patients and perfusion 

imaging in 3. We determined ADC values and signal 

intensity on diffusion-weighted images in the solid portion 

of the tumor and compared these results with 20 cases of 

classical glioblastoma. 

FINDINGS 

All tumors were well defined, relatively large lesions. 

Characterized 3 cases showed prominent surrounding 

edema, but the edema in the other 3 cases was only minimal 

even if it was a large tumor. Five cases were abutting a dural 

surface. One case was accompanied by a large cyst and the 

other cases showed intratumoral necrosis. Solid portions of 

the tumor showed hyperintensity on diffusion-weighted 

imaging in all 4 cases examined and the mean ADC values 

394 

was .0011 ± .0002 mm 2 /s, which was not significantly different 

from those of classical glioblastoma. Perfusion imaging 

showed increased rCBV in the solid portion of the tumor 

in the 3 cases examined. 

CONCLUSION 

Gliosarcomas have specific MR imaging characteristics. 

Thus, the tumor is well defined and abuts a dural surface. 

The surrounding edema is generally less prominent than seen 

in other tumors of same size but of different histologic characteristics. 

Diffusion and perfusion imaging is similar to 

other glioblastomas. 

REFERENCES 

1. Dwyer KW, Naul LG, Hise JH. Gliosarcoma: MR futures. J 

Comput Assist Tomogr 1996;20:719-723 

KEY WORDS: Gliosarcoma, MR imaging, diffusion 


Astrocytoma: Current and Future Imaging 

Esposito, F. J. 1 · Papa, M. 2 

1 Mercy Catholic Medical Center, Darby, PA, 2 Second 

University of Naples School of Medicine, Naples, ITALY 

PURPOSE 

To review the biology, classification, and imaging of astrocytomas. 


A poster presentation will review the biology, the revised 

World Health Organization classification, imaging characteristics, 

and epidemiologic factors of astrocytomas. 

Imaging modalities that will be described will include CT, 

positron emission tomography (PET) and MR imaging. 

Special emphasis will be placed on MR spectroscopy, diffusion 

imaging, perfusion imaging and diffusion tensor imaging. 

FINDINGS 

Illustrative images will be included. 

CONCLUSION 

After viewing the exhibit the viewer will: 1) Have an understanding 

of the basic biology of astrocytomas. 2) Have 

reviewed the revised WHO classification of astrocytomas. 3) 

Understand the current imaging modalities available for 

astrocytoma imaging and their limitations. 4) Appreciate the 

likely future of astrocytoma imaging. 

KEY WORDS: Astrocytoma


Comprehensive, Detailed, and Quickly Performed 

Functional MR Imaging Approach to Brain Tumors 

da Cruz, L. H. 1 · Ferreira, F. B. 2 · Domingues, R. C. 1 · 

Domingues, R. C. 1 · Nogueira, J. 3 

1Clínica de Diagnóstico por Imagem-Barrashopping/Multi- 

Imagem Ressonancia, Rio de Janeiro, BRAZIL, 2Multi- Imagem Ressonancia, Rio de Janeiro, BRAZIL, 3Instituto Nacional de Câncer, Rio de Janeiro, BRAZIL 

PURPOSE 

To determine whether new functional MR imaging 

sequences [diffusion tensor, tractography, MR perfusion, 

MR spectroscopy (MRS) and functional MR imaging using 

blood-oxygenation level dependent (BOLD) sequence] can 

help to add new information to presurgical approach of brain 

tumors. To determine the efficiency and efficacy of these 

tools in the evaluation of brain tumors. To characterize the 

MR features in the evaluation of brain tumors using these 

new modalities. 


MR exams of 29 patients with biopsy-proven brain tumors 

(17 men, 12 women; mean age 47.5 years) were performed 

in a 1.5 T clinical scanner using MR clinical standard protocol 

and functional MR imaging sequences (low-grade 

glioma, n = 8; anaplastic astrocytoma (AA), n = 6; glioblastoma 

multiforme, n = 9; metastasis, n = 5; gliomatosis cerebri, 

n = 1). Diffusion tensor imaging (DTI) and tractography 

were performed using a single shot spin-echo echo-planar 

sequence applied in six different directions and a DTI-fractional 

anisotropic map was obtained. MR perfusion was performed 

using TR of 1500 ms and TE of 90 ms, following 

bolus infusion of intravenous contrast material, acquiring 

relative cerebral blood volume (rCBV) of the tumor and of 

the contra-lateral normal white matter. MR spectroscopy was 

done with a multivoxel PRESS sequence (TE = 30 ms), 

obtained main metabolites ratios. Functional MR imaging 

used to identify eloquent areas (sensory-motor and language) 

of cortical activation based on the BOLD phenomenon. 

FINDINGS 

All high-grade gliomas (HGG) and metastasis had high perfusion. 

Low-grade gliomas (LGG) and gliomatosis cerebri 

(GC) had low perfusion. MR spectroscopy showed high 

Co/Cr ratio in AA (n = 5, mean 2.02) and in all GBM (mean 

2.35) and metastasis (mean 2.17) cases. The NAA/Cr ratio 

was low in these patients (mean 1.02). Co/Cr and NAA/Cr 

mean ratios in LGG were 1.17 and 1.24 respectively. High 

pick of mI was found in LGG (n = 6) and in the GC case. 

HGG caused disruption and dislocation of the main fiber 

tracts (MFT) verified by DTI-FA map. LGG caused dislocation, 

but not disruption. Main fiber tracts were infiltrated but 

not disrupted in GC. FA values were diminished in the vasogenic 

edema surrounding metastasis, causing misdiagnosis 

of MFT infiltration. BOLD sequences identified sensorymotor 

and language cortical areas. The time spent to perform 

each exam was shorter, as parallel acquisition system used 

makes functional sequences quicker. 

395 

CONCLUSION 

MR imaging has been used widely to evaluate brain tumors. 

With the advent of the new neuroimaging advanced tools, 

the assessment of these lesions has being done more accurately, 

precisely, and rapidly, contributing in their evaluation 

and management of these patients. 

KEY WORDS: Brain neoplasm, diffusion tensor imaging, 

tractography 


Role of Standard Clinical MR Spectroscopy in the 

Management of Brain Tumors with Gamma Knife 

Radiosurgery 

Lee, C. · Young, A. B. · Fields, T. M. · Green, E. B. · Given, 

C. A. 

University of Kentucky 

Lexington, KY 

PURPOSE 

In most current 1.5 T MR imagers, using either CHESS or 

STEAM to suppress the signal of water and fat, MR spectroscopy 

(MRS) is limited to what frequency or peaks it can 

resolve as a proton imager. Thus at least at the 1.5 T level 

with proton imaging there are limited metabolites that can be 

resolved. So on most routine MRS of normal brain inositol, 

choline, creatine, and NAA peaks are resolved. Using standard 

software on a 1.5 T MR, we set out to determine the role 

of MRS in management of CNS tumors undergoing gamma 

knife treatment. In gamma knife treatment radiation necrosis 

changes may occur early with as much as a 20% volume 

increase, which MRS may differentiate from tumor 

regrowth. It is the purpose of this exhibit to present our experience 

and discuss MR spectra focusing mainly on tumor 

spectra of primary and metastatic tumors in the management 

of this specific population. 


There were 38 patients undergoing gamma knife treatment 

of which 16 were primary, and 22 metastatic tumors all studied 

on a 1.5 T Siemen’s unit. Voxels were placed on the 

enhancing nodular portions avoiding the cystic/fluid portions. 

Standard MRS graphs and ratios were generated and 

reviewed. Some metastatic lesions were too small, and some 

to close to the skull. This select population excluded most of 

the GBM. Meningiomas and acoustic schwannomas were 

excluded since most occurred near bone. 

FINDINGS 

No characteristic spectra were found, but generally NAA 

was decreased and choline markedly increased. There 

seemed to be correlation with choline peaks and degree of 

malignancy. Creatine also was generally decreased. Lactate 

and lipid peaks were rarely seen. Radiation necrosis was evident 

with a noisy background MRS pattern. No change and 

decrease in size of tumor were considered favorable responses, 

and the MRS usually showed necrosis. 

Scientific Exhibits


CONCLUSION 

Characteristic MRS tumor patterns were not found, but this 

population is too small. Tumor specificity by this MRS software 

package was limited. The MRS was most helpful in determining 

whether the tumor was malignant or not, and thus helped 

characterize the tumor response to treatment. And in making 

this diagnosis, an elevated choline peak was most helpful. The 

overall diagnosis was still based primarily on the clinical history 

and MR features, and change in the size or tumor volume. 

Elevated choline peak, and lactate or lipid peaks reflect increasing 

degrees of malignancy. Metastatic lesions decreased in size 

even after one treatment, with MRS reflecting radiation necrosis. 

Sometimes biopsy was still necessary. MR spectroscopy 

clearly has an adjunct role in the management of these patients 

and can be incorporated into the routine clinical follow up. 

KEY WORDS: MR spectroscopy-brain tumors, gamma knife 

tumors, MR spectroscopy 


Intraventricular Masses 

Chaney, K. A. · Michals, E. 

University of Illinois Medical Center 

Chicago, IL 

PURPOSE 

To demonstrate the wide variety of intraventricular masses 

and to demonstrate how location, patient’s age, and imaging 

characteristics shape and narrow one’s differential diagnosis. 


Patients who had intracranial tumors and prior MR and/or CT 

imaging between 2000 and present (2004) were examined. 

FINDINGS 

Those whose imaging demonstrated intraventricular masses 

were selected and pathology was verified. Both males and 

females were included in this retrospective analysis, as were 

adults and pediatric patients. 

CONCLUSION 

Although the number of intraventricular masses encountered 

is substantially less than parenchymal masses, their relative 

rarity and large variety of etiologies often pose a clinical challenge. 

Both exophytic intraaxial tumors and extraaxial tumors 

can present intraventricularly resulting in a complex differential 

diagnosis. These tumors include extraaxial choroid plexus 

tumors, ependymal tumors, and meningeal tumors and 

intraaxial, exophytic tumors such as glioma and the primitive 

neuroendocrine tumor, medulloblastoma. Nonneoplastic etiologies 

(e.g., colloid cysts) and infectious etiologies (e.g., neurocysticercosis) 

also can present intraventricularly. 

Furthermore, systemic disease such as tuberous sclerosis and 

metastatic tumors can involve the ventricles. Thus, understanding 

the normal anatomy and histology of the ventricular 

system and surrounding brain parenchyma, combined with the 

knowledge of the location of the mass in the ventricular system, 

age of the patient, and imaging characteristics, facilitate 

in the creation of an accurate differential diagnosis. 

KEY WORDS: Neoplasm, intraventricular 

396 


Functional MR Imaging and Direct Cortical Stimulation 

for Language and Motor Function Mapping in Surgical 

Management of Brain Space-Occupying Lesions 

Causin, F. 1 · Iannucci, G. 1 · Pinna, V. 1 · Mondani, M. 2 · Budai, R. 2 

· Skrap, M. 2 · Cavedon, C. 1 · Stancanello, J. 1 · Francescon, P. 1 

1 2 St. Bortolo City Hospital, Vicenza, ITALY, S. Maria della 

Misericordia City Hospital, Udine, ITALY 

PURPOSE 

BOLD technique carries important advantages in terms of 

noninvasivity and of scanner system availability but still 

needs confirmation for it diffuse presurgical application. 


We compared direct cortical stimulation and BOLD functional 

MR imaging (fMRI) in the aim to estimate potentials 

and limits of BOLD-based brain mapping of language and 

motor functions and to understand advantages and drawbacks 

of both techniques in 75 surgical patients. 

FINDINGS 

In relation to lesion location and to neurologic symptoms 

motor and language tasks were performed using a 1.5 T 

scanner with GRE-EPI sequences and BOLD technique 

(TR/TE: 0.96 sec/66 msec, FA: 90°; matrix: 128 x 128). 

Statistical parametric mapping (SPM99) implemented in 

MatLab 6 was used for statistical analysis using Student’s ttest 

(correct p value = 0.05/0.01 and uncorrect p value = 

0.001/0.0001) corrected for false positive (Bonferroni correction). 

BOLD-fMRI results were analyzed with Talairach 

co-planar brain atlas for Brodmann areas (BA) assessment. 

After craniotomy and lesion exposition patients were awakened 

from anesthesia and direct cortical stimulation (DCS) 

was performed positioning bipolar electrical probes upon the 

lesion and on the surrounding brain tissue. Our report shows 

effectiveness of this modality in the detection of eloquent 

areas in patients undergoing surgical lesion removal. 

Importance of localization of eloquent functional areas 

should be useful to decrease the neurologic postsurgical 

residual deficit and to evaluate possible downsides of surgical 

procedures especially in benign or low-grade lesions, 

suggesting vigilance or alternative treatments. 

CONCLUSION 

We believe that in collaborative patients fMRI tasks performed 

should be able to give useful information about functions 

tested. The ability of this test to predict all key language 

functions and to give surgeons precise topographical 

information of essential cortex remains unclear. Functional 

MR imaging remains actually an important tool for incruent 

brain function evaluation and for presurgical analysis. 

KEY WORDS: Functional MR imaging, brain lesion, neurosurgery

397 



Cystic Intracranial Neoplasms: Neuroimaging and Diffusion-Weighted MR Imaging of Brain: Further than 

Neuropathology Correlation 

Stroke 

Wang, A. · Kincaid, J. · Silbergleit, R. · Wilson, J. · Kragha, 

K. · Barry, K. · Tech, K. · Noujaim, S. 


Royal Oak, MI 

PURPOSE 

To assess the neuroimaging findings of a wide variety of 

intracranial cystic neoplasms and their location and correlated 

with neuropathology; to provide a reasonable differentional 

diagnosis and to help for the patient’s care. 


This is a retrospective study for review of neuroimaging and 

neuropathology of 50 patients, age from 2 to 80 years old 

presented with intracranial cystic neoplasms. The neuroimaging 

studies included are as follows: brain MR imaging 

with and without contrast (all of the patients), diffusion 

imaging (most of the patients), MR spectroscopy (some of 

the patients). 

FINDINGS 

In the pediatric group, the posterior fossa cystic ganglioglioma 

can mimic juvenile cystic astrocytoma; the supratentorial 

cystic ganglioglioma can mimick juvenile cystic 

astrocytoma or PNET. In adult group, metastasis can mimick 

brain abscess or glioblastoma multiforme; hemangioblastoma 

in the cerebellum does present with nodular enhancement 

with signal void and can be differentiated with metastasis. 

In extraaxial neoplasms, cystic meningioma, Rathke’s 

cleft cyst, craniopharyngioma and epideromoid cyst can be 

diagnosed reasonably prior to the surgery because of thier 

location and MR findings. Restriction of diffusion can be 

seen in most cystic neoplasms. MR spectroscopy through the 

rim of the cystic neoplasm may provide useful information 

for the differential diagnosis of malignant versus benign neoplasm. 

CONCLUSION 

A wide variety of intracranial cystic neoplasms are presented 

with neuroimaging findings and correlated with neuropathology. 

The majority of cases can be reliably making 

preoperative diagnosis based upon their location, age group, 

and neuroimaing findings. However, cystic ganglioglioma 

can mimck juvenile cystic astrocytoma in pediatric group. 

Cystic metastasis can also mimick brain abscess. 

KEY WORDS: CNS, neoplasms, cystic 

Sawlani, V. · Powell, N. 

Morriston Hospital 

Swansea, UNITED KINGDOM 

PURPOSE 

Diffusion-weighted MR imaging provides image contrast 

that is different from that provided by conventional MR 

techniques. It is particularly sensitive for detection of cytotoxic 

edema. The objective of this Scientific Exhibit is to differentiate 

diseases that manifest with sudden neurologic 

deficit, based on diffusion contrast and apparent diffusion 

coefficient (ADC) values and review of hyperintense and 

hypointense lesions on diffusion-weighted MR imaging. 


Diffusion-weighted sequence is incorporated in the MR 

imaging protocols of patients clinically presented with acute 

and subacute neurologic deficit from January 2003 to June 

2004. The study was performed on a 1.5 T GE Twin speed 

MR imaging system. The standard EPI-based diffusionweighted 

imaging sequence with b value 1000 is used for 

evaluation of lesion and calculation of ADC values. The 

image contrast on diffusion-weighted imaging is dependent 

on the molecular motion of water, which is substantially 

altered by disease. The restriction of water movement inside 

the cell or in between the cells creates a unique diffusion 

contrast and alters the ADC values. The hyperintense and 

hypointense lesions on diffusion-weighted imaging 

sequence are evaluated and compared with standard MR 

imaging sequences. 

FINDINGS 

High signal intensity on diffusion-weighted MR imaging and 

low ADC values similar to the findings in acute cerebral 

infarction are seen also in many other conditions such as herpes 

simplex encephalitis, Japanese encephalitis, congenital 

rubella syndrome, pyogenic abscess, tuberculoma, solid 

hypercellular tumors, epidermoid, and liquid embolization 

material (glue and lipidol mixture). Similarly low signal 

intensity on diffusion-weighted MR imaging and high ADC 

values are observed in many conditions such as chronic 

infarct, necrotic tumor, neurocysticercosis, acute demyelinating 

plaque, adrenoleukodystrophy and arachnoid cyst. 

CONCLUSION 

A review of hyperintense and hypointense lesions on diffusion-weighted 

MR imaging and their differential diagnosis 

with emphasis on the concept of cytotoxic and vasogenic 

oedema is presented. Diffusion-weighted MR imaging may 

help also in differentiating arterial and venous infarct, active 

verses chronic demyelination in leukodystrophy and differentiating 

ring-disk enhancing lesions. 

KEY WORDS: Diffusion-weighted imaging, apparent diffusion 

coefficient, MR imaging 

Scientific Exhibits



Perfusion MR Imaging of Brain Tumors: One Step 

beyond Conventional MR Imaging 

da Cruz, L. H. · Domingues, R. 

Clínica de Diagnóstico por Imagem-Barrashopping/Multi- 

Imagem Ressonancia 


PURPOSE 

To explain the physical principles underlying perfusion MR 

imaging. To describe applications of perfusion MR imaging 

in the diagnosis and characterization of brain masses as well 

as their differential diagnosis. To highlight MR perfusion 

applications in the assessment of brain tumors and its utility 

to predict the grading of malignant lesions. To determine the 

contribution of MR perfusion in the differential diagnosis 

between posttreatment changes, as radiation necrosis, and 

recurrent tumor. 


Perfusion MR imaging is a contrast dynamic susceptibility 

contrast method. Perfusion MR imaging was performed 

using TR 1500 ms and TE 90 ms following a bolus infusion 

of intravenous contrast material in a 1.5 T clinical scanner. 

Then, a relative cerebral blood volume (rCBV) map can be 

obtained from the perfusion MR imaging data. 

FINDINGS 

The rCBV map of the tumor when compared with the normal 

contralateral white matter can predict whether the tumor has 

hyperperfusion or not. Hyperperfusion has been extensively 

associated with high grade lesions. 

CONCLUSION 

This new MR method has been used widely in clinical practice 

to evaluate patients with brain tumors. Its use can be 

focused in the correct diagnosis of these lesions, to assess the 

treatment response, and to know if there is recurrent tumor 

or not. 

KEY WORDS: Neoplasm, perfusion 


Prognostic Value of MR Imaging in Patients with Diffuse 

Axonal Injury 

Martín, A. 1 · Carvajal, A. 1 · Martinez, M. 2 · Rovira-Gols, A. 1 

· Zauner, M. 1 · Bermudez, P. 1 

1UDIAT-CD Corporació Sanitària Parc Taulí, Sabadell, 

SPAIN, 2Critical Care Center, Corporació Sanitària Parc 

Taulí, Sabadell, SPAIN 

PURPOSE 

Diffuse axonal injury (DAI) is one of the most important 

types of brain damage that can occur as a result of nonmissile 

head injury. Cerebral MR imaging has been shown sensitive 

for detection and characterization of these lesions. The 

aim of this exhibit is to study the prognostic value of MR 

imaging in patients with DAI. 

398 


We performed a prospective study between Jan 1999-Aug 

2003, with 20 patients with severe heat trauma (SHT) and 

discrepancy between the normal CT and the neurologic status 

defined as: glasgow coma scale (GCS) ≤ 8, no intracranial 

hypertension and abnormal awake when withdrawing 

sedatives. MR imaging was performed at medium of 10 days 

after injury to detect DAI type lesions. Patients with DAI 

were divided into three groups in correlation with three 

major anatomical areas: DAI type 1 with axonal injury in the 

white matter of the cerebral hemispheres, DAI type 2 with 

focal lesion in corpus callosum, and brain stem DAI type 3. 

We compare MR imaging of DAI types and the Glasgow 

Outcome Scores (GOS) at 6 months. 

FINDINGS 

In eighteen patients (90%) MR imaging demonstrated DAItype 

lesions. The GCS was lower in patients with DAI type 

2 and 3 (GSC 4.75 and 5 respectively) than in patients with 

DAI type 1 (GCS 7). The correlation between DAI groups 

and the GOS scale are shown. 

CONCLUSION 

MR imaging is a sensitive method in detection of DAI-type 

lesions. The abnormal neurologic status probably might be 

due to the DAI lesions. There is a good correlation between 

the GCS, DAI-type lesions and GOS at 6 months. More than 

50% of patients with DAI due to SHI had a good recovery at 

6 months. Despiste the little number with DAI-type 1 lesions 

(without corpus callosum or brain stem involvement) it 

seems that they have a better prognosis. 

KEY WORDS: Diffuse axonal injury, MR imaging, head trauma 


Imaging and Treatment of Craniocervical Arterial 

Dissection 

Place, C. · Srinivasan, A. · Lum, C. · Goyal, M. 

The Ottawa Hospital 


PURPOSE 

To highlight recent advances in imaging of craniocervical 

arterial dissection (CCAD) and discuss various therapeutic 

options. 


The clinical presentation, etiologies, imaging features, and 

therapeutic options in CCAD are reviewed with appropriate 

examples. 

FINDINGS 

Arterial dissection arises from an intimal tear allowing blood 

to enter the wall and form an intramural hematoma. This 

may result in luminal stenosis or aneurysmal dilatation 

(“pseudoaneurysm”). The extracranial carotid and vertebral 

arteries are more likely to undergo dissection than the 

intracranial segments. CCAD occurs predominantly in 

young or middle-aged adults with no sex predominance. The 

various etiologies include spontaneous, traumatic, iatrogenic, 

fibromuscular dysplasia, and inherited connective tis-

sue disorders. History of a minor precipitating event (e.g., 

chiropractic neck manipulation, coughing) frequently is 

elicited with spontaneous dissections. Craniocervical arterial 

dissection can be clinically silent or produce symptoms 

ranging from cervical pain, headache, incomplete Horner 

syndrome, transient ischemic attacks, infarcts, to even death. 

Noninvasive studies like MR angiogram (MRA), CT 

angiogram (CTA) and ultrasound are attractive alternatives 

to conventional angiography (gold standard) for the diagnosis. 

Apart from avoiding potential complications of invasive 

angiography, they can demonstrate directly the intramural 

hematoma. MR imaging typically shows a narrowed eccentric 

signal void surrounded by a hyperintense semilunar signal 

(mural hematoma) on T1- and T2-weighted images. 

Contrast-enhanced three-dimensional MRA with elliptic 

centric-view ordering has high spatial and contrast resolution 

(with venous-suppressed images) that can replace conventional 

angiography. It also depicts the origin of the aortic 

arch vessels better than time of flight MRA. MR imaging is 

also superior to angiography in dissection without luminal 

abnormalities and in nonspecific occlusions. CT angiography 

can be performed quickly, is independent of flow phenomena 

and useful where MR imaging is contraindicated. 

Intracranial complications like infarct and subarachnoid 

hemorrhage can be demonstrated also with CT and MR 

imaging. Ultrasound with Doppler and color-flow imaging 

can provide useful information in the initial assessment and 

follow up of carotid artery dissections. Overall, MR imaging 

with MRA is presently the method of choice for initial diagnosis 

and follow up of CCAD. Despite anticoagulation being 

advocated since the 1970s, there are no randomized trials, 

and its validity never has been proven. However, there is 

some indirect evidence of the appropriateness of anticoagulation. 

A target of 2.0 to 3.0 international normalized ratio 

generally is used for 3 to 6 months. Follow-up MR imaging 

is suggested at 3 months to look for luminal irregularity and 

changes in therapy instituted accordingly. The prognosis of 

CCAD depends on the presence and severity of ischemic 

brain damage and the extent of collateral circulation. With 

anticoagulation about 90 percent of stenoses resolve, two 

thirds of occlusions recanalize, and one third of pseudoaneurysms 

decrease in size. Presently, the indications for 

endovascular or surgical intervention are unclear. While 

some authors reserve such treatment for patients having 

pseudoaneurysms or persistent ischemic symptoms despite 

adequate anticoagulation, others prefer a conservative 

approach even in these scenarios. 

CONCLUSION 

CT and MR imaging are reliable noninvasive tools for rapid 

diagnosis and follow up of CCAD and may replace angiography 

as the gold standard. Further evidence is required to 

define the role of newer therapeutic options. 

KEY WORDS: Dissection, carotid, vertebral 

399 


Diagnosis of Carotid Arterial Dissection: Is Multislice 

Detector CT Angiography Better than Time-of-Flight 

MR Angiography? 

Elijovich, L. · Kazmi, K. · Law, M. 


New York, NY 

PURPOSE 

Cervical artery dissection (CAD) is an important cause of 

ischemic stroke, particularly in young patients. Noninvasive 

imaging essentially has replaced conventional angiography 

as the gold standard in diagnosis. MR imaging/time-of-flight 

(TOF) MRA and multislice CT angiography (MS CTA) have 

both demonstrated high specificity and sensitivity in making 

this diagnosis. However, no direct comparison of these two 

techniques has been reported in the literature. The purpose of 

this study was to present/compare the findings of CAD on 

MR imaging/TOF MRA and MS CTA and determine if MS 

CTA offered additional information that is not readily available 

MR imaging/TOF MRA. 


We reviewed six patients who had carotid artery dissection 

and had both MR imaging/TOF MRA and MS CTA performed 

as part of their evaluation. MR angiography was performed 

utilizing 2D and 3D TOF sequences. Both the source 

data as well as maximum intensity projection (MIP) reconstructions 

were reviewed. Axial fat saturated (FS) T1weighted 

images also were obtained. Multislice CTA was 

performed on a 16-slice CT scanner. A contrast bolus at a 

high injection rate (3.5-4.0 cc/second) was used. The axial 

source images as well as MIP reconstructions were 

reviewed. 

FINDINGS 

Given that the pathognomic angiographic findings of intimal 

flap and double lumen often are not present, visualization of 

a mural hematoma on FS T1-weighted imaging is widely 

accepted as the key finding in diagnosing carotid dissection 

with a sensitivity and specificity approaching 100% in some 

series. Separate series have demonstrated similar results for 

the MS CTA findings of mural thickening or eccentric luminal 

narrowing associated with a long segment of involvement. 

This was confirmed in our series as dissection was 

diagnosed in all six patients on MS CTA using these findings. 

It is well known that the reliance of TOF MRA on ferromagnetic 

properties of flowing blood makes it difficult to 

evaluate areas of complex flow. Therefore, arterial luminal 

narrowing and irregularities are assessed more readily with 

an arterial contrast study, such as MS CTA. Our series illustrates 

this point. In one patient, a pseudoaneurysm was 

detected on MS CTA that was not seen on MR imaging/TOF 

MRA. In another patient, MS CTA demonstrated a long segment 

of irregularity and beading suggestive of fibromuscular 

dysplasia whereas TOF MRA only showed a long segment 

of narrowing. In a third patient, MRimaging/MRA overestimated 

stenosis as occlusion, a limitation that has been well 

described in studies of carotid stenosis. All three of these 

additional findings provided by MS CTA had implications 

on the management of these patients. 

Scientific Exhibits


CONCLUSION 

Despite their independently reported high specificity and 

sensitivity in the diagnosis of CAD, these cases demonstrate 

that MS CTA and MR imaging/TOF MRA are not interchangeable 

in regards to the information that they provide. 

This series suggests that MS CTA may be superior in depicting 

the concurrent abnormalities and complications of this 

disease that may alter management. 

KEY WORDS: Carotid dissection, MR angiography, CT 

angiography 


Carotid and Vertebral Artery Dissection: A Pictorial 

Review 

Griffith, G. M. · Sudhir, K. · Wang, H. 

University of Rochester 

Rochester, NY 

PURPOSE 

Dissection of the carotid and vertebral arteries has been 

increasingly recognized as a cause of stroke in young and 

middle-aged adults. Prompt recognition of the disease is 

important for effective management by anticoagulation or 

endovascular interventions to minimize the risks of infarction, 

permanent neurologic deficits, or death. The aim of this 

exhibit is to review the neuroradiologic features of carotid 

and vertebral artery dissections based on a series of collected 

cases. 


The normal anatomy and neuroradiologic findings in carotid 

and vertebral artery dissections will be reviewed from case 

studies collected from the past 4 years at the Strong 

Memorial Hospital. The modalities used include CT angiography 

[axial sections with maximum intensity projection 

(MIP), multiplanar volume reformat (MPVR), MR angiography 

(coronal 2D-phase contrast and axial 3D time of flight 

with 2D/3D postprocessing] and digital subtraction angiography 

of the head and neck. A review of the current literature 

will be cross referenced with cases from our institution. 

FINDINGS 

The exhibit will include normal anatomy for reference and 

neuroradiologic findings of extracranial and intracranial 

carotid and vertebral arterial dissection of spontaneous, traumatic, 

and other etiologies. 

CONCLUSION 

CT angiography and MR angiography techniques are replacing 

conventional angiography for initial diagnosis of carotid 

or vertebral artery dissection. Familiarity with the radiographic 

features facilitates early diagnosis, which can have 

a significant impact on the reduction of the morbidity and 

mortality in the young adults affected with this disease. 

KEY WORDS: Arterial dissection, CT angiography, MR 

angiography 

400 


CT Angiography Using Multidetector Row CT in the 

Evaluation of Extracranial Carotid Stenosis: Utility of 

Newly Devised Volume Rendering Method and Delayed 

Enhancement Technique 

Imakita, S. · Tanaka, R. · Higashi, M. · Onishi, Y. · Yamada, 

M. · Naito, H. · Murao, K. · Iihara, K. · Miyamoto, S. 

National Cardiovascular Center 

Suita, JAPAN 

PURPOSE 

Our goal was to establish the utility of CT angiography using 

multidetector row CT with newly devised volume rendering 

method (Narrow Bandwidth Volume Rendering) and delayed 

enhancement technique for the evaluation of extracranial 

carotid stenosis. 


Patients with extracranial carotid stenosis were examined 

using multidetector row CT. Multidetector row CT scanner 

[Aquilion Multi (16-detector rows), Toshiba Medical 

Systems, Japan] was used. The scan parameters were as follows: 

high voltage 120 kV, tube current 300 mA, slice collimation 

16 x 0.5 mm, rotation time 0.5 sec, CT pitch factor 

0.6875:1 (helical pitch 11:16), scan length 85 mm. The contrast 

medium (Iohexol 350 mgI/ml) was injected via antecubital 

vein at the rate of 2.0 ml/sec with power injector. The 

total volume of contrast medium was 1.0 ml/body weight kg. 

The 0.5 mm thick axial images were reconstructed at every 

0.5 mm interval. The 3D reconstruction was performed using 

usual volume rendering method and narrow bandwidth volume 

rendering on a workstation (Zio M900 Quadra, Ziosoft 

Inc., Japan). Multiplanar reformation (MPR) images were 

obtained also in early enhancement phase and delayed 

enhancement phase. These images were compared with digital 

subtraction angiogram (DSA) for their characterization 

of extracranial carotid stenosis. 

FINDINGS 

Three-dimensional CT angiograms could clearly depict 

carotid stenosis from any direction except the cases surrounded 

by dense calcification and inserted metallic stent. CT 

angiograms reconstructed using narrow bandwidth volume 

rendering could clearly depict the lumen surrounded by dense 

calcification or metallic stent. In many cases surrounded by 

dense calcification MPR images could not clearly depict the 

stenotic lesion. Multiplanar reformation images could clearly 

depict the intimal thickening inside the metallic stent. 

Multiplanar reformation images could clearly depict carotid 

plaque. Multiplanar reformation images in delayed enhancement 

phase could more clearly depict the extent of carotid 

plaque than ones in early enhancement phase. 

CONCLUSION 

Narrow bandwidth volume rendering was a valuable method 

for the examination of extracranial carotid lesion with dense 

calcification and metallic stent. Multiplanar reformation 

image in delayed enhancement phase was an effective tool 

for evaluating the extent of carotid plaque. 

KEY WORDS: CT angiography, multidetector row CT, 

extracranial carotid stenosis

401 



Pathology of the Pericallosal Artery and Its Perivascular CT Venography: Distinct Advantages over MR 

Pial Plexus: A Pictorial Essay 

Venography 

Lum, C. · Srinivasan, A. · Miller, W. · Goyal, M. 

University of Ottawa 


PURPOSE 

The corpus callosum (CC) is the major white-matter interhemispheric 

pathway. It commonly is affected in demyelination 

and tumors. However, vascular-related pathologies in 

the CC are uncommon in part due to its distal territory and 

rich anastamotic plexus. This exhibit reviews the vascular 

supply of the CC, pericallosal artery (PeriCA) microvascular 

anatomy, and both rare and typical pathologic entities of the 

PeriCA and its branches. 


A review of our electronic teaching file database maintained 

since 2002 to present was performed. Interesting cases 

involving vascular pathology of the corpus callosum were 

reviewed and selected for this pictorial exhibit. A review of 

the microsurgical anatomy of the PeriCA was performed. 

FINDINGS 

Four patients with ruptured pericallosal artery aneurysms 

and typical callosal sulcus subarachnoid hemorrhage were 

identified. Three were treated by endovascular coil occlusion, 

one by surgery. In one patient, there were two other 

aneurysms in which the pattern of hemorrhage (diffuse with 

predominant callosal sulcus blood) was helpful in identifying 

the ruptured aneurysm. One patient had a ruptured pericallosal 

AVM treated by Onyx liquid embolic material and 

radiotherapy. Two patients had prior VP shunts and cystic 

lesions of the corpus callosum identified as scalloping deformity 

of the corpus callosum (CC). The presumed pathophysiology, 

softening of the CC from chronic hydrocephalus and 

tethering of the perforating branches arising from the PeriCA 

to the overlying cingulate gyrus, is reviewed, along with the 

salient PeriCA pial plexus anatomy. Two patients presented 

with acute infarction of the body and splenium of the corpus 

callosum. Cases of diffuse axonal injury involving the corpus 

callosum are presented. 

CONCLUSION 

The arterial supply to the corpus callosum is rarely emphasized. 

These cases highlight some rare and common vascular-related 

pathologies of the corpus callosum and provide an 

opportunity to review its microvascular anatomy. 

REFERENCES 

1. Menovsky T, van Rooij WJ, Sluzewski M et al. Coiling of ruptured 

pericallosal artery aneurysms. Neurosurgery 

2002;50:11-14 

2. Numaguchi Y, Kristt DA, Joy C et al. Scalloping deformity of 

the corpus callosum following ventricular shunting. AJNR 

Am J Neuroradiol 1993;14:355-362 

3. Ture U, Yasargil MG, Krisht AF. The arteries of the corpus callosum: 

A microsurgical anatomic study. Neurosurgery 

1996;39:1075 

KEY WORDS: Aneurysm, corpus callosum, anatomy 

Pekala, J. S. 1 · Shah, L. 2 · Enterline, D. 1 

1 2 Duke University Medical Center, Durham, NH, Duke 

University, Durham, NC 

PURPOSE 

CT venography (CTV) is becoming a first-line tool in the 

evaluation of the dural venous system. This exhibit examines 

the potential advantages of CTV over MR venography 

(MRV) including: improved special resolution, the ability to 

assess the osseous structures of the skull base, and excellent 

contrast between intraluminal filling defects and normal 

flow. 


We present cases that highlight the advantages of CTV over 

MRV. These cases include congenital sigmoid hypoplasia, 

dense arachnoid granulations initially thought to represent 

thrombus on MRV, and pseudotumor cerebri with dural vein 

stenosis (a contentious relationship). We also review the current 

literature supporting CTV for the evaluation of the dural 

sinuses and contrast that with recent literature supporting 

MRV for this indication. 

FINDINGS 

These cases demonstrate the potential advantages of CTV 

over MRV. However, the current literature continues to consider 

the strengths and weaknesses of each technique. 

Currently many authors feel that CTV provides a more sensitive 

exam, while MRV is more specific. 

CONCLUSION 

CT venography yields detailed images of the intracranial 

venous circulation. It is a rapid method for the diagnosis of 

dural sinus thrombosis and provides excellent special resolution 

and contrast as well as the ability to assess the osseous 

structures of the skull base. These factors can aid in achieving 

a diagnosis in certain clinical situations. 

KEY WORDS: Thrombosis, dural vein 


Multidetector Row (16 Slice) CT Angiography in Patients 

with Acute Cerebrovascular Disease 

Suh, S. I. 1 · Seol, H. Y. 1 · Kim, T. 2 · Lee, Y. H. 2 · Lee, N. J. 2 · 

Kim, J. H. 2 · Cha, I. H. 1 · Kwon, T. H. 1 

1Korea University Guro Hospital, Seoul, REPUBLIC OF 

KOREA, 2Korea University Hospital, Seoul, REPUBLIC OF 

KOREA 

PURPOSE 

To describe the optimized scanning protocols of 16-slice 

multidetector CT (MD CT) angiography for cervicocranial 

vascular disease and to provide examples of clinical applications 

in patients with various cervicocranial vascular pathology. 

Scientific Exhibits



We demonstrate the MD CT angiography findings (source 

images and various three-dimensional reconstruction techniques 

with or without subtraction) of cervicocranial vascular 

lesions as follows: intracranial and extracranial carotid 

stenosis and occlusion, dissection with pseudoaneurysm 

involving vertebrobasilar arterial system, various intracranial 

aneurysms, previously clipping aneurysm, vasospasm 

after subarachnoid hemorrhage, parenchymal arteriovenous 

malformation, recurred dural arteriovenous fistula and dural 

sinus thrombosis. 

FINDINGS 

We correlate the imaging findings of the MD CT angiography 

with digital subtraction angiography (DSA). There were 

good correlations between imaging findings of the MD CT 

angiography for cervicocranial vascular pathology and those 

of the DSA. 

CONCLUSION 

Multislice (16 channel) CT angiography may be a promising 

reliable new diagnostic tool for the rapid and comprehensive 

assessment of the arteriovenous cerebrovascular disease in 

patients with clinical signs of acute cervicocranial vascular 

disorders. 

KEY WORDS: Multidetector row CT, CT angiography, 

intracranial aneurysm, arteriovenous malformation 


Utility of “Volume Position Matching” in Subtraction 3D 

CT Angiography with Controlled-Orbit Helical Scanning 

Imakita, S. · Onishi, Y. · Yamada, M. · Tanaka, R. · Higashi, 

M. · Naito, H. · Iihara, K. · Miyamoto, S. 

National Cardiovascular Center 

Suita, JAPAN 

PURPOSE 

Our goal was to evaluate the utility of newly devised 

“Volume Position Matching” (the method for decreasing a 

misregistration artifact) in subtraction three-dimensional 

(3D) CT angiography with controlled-orbit helical scanning 

for the examination of pre and postoperative intracranial 

aneurysms. 


The patients with pre and postoperative intracranial 

aneurysms were examined using multidetector row CT with 

controlled-orbit helical scanning. Multidetector row CT 

scanner [Aquilion Multi (16-detector rows), Toshiba 

Medical Systems, Japan] was used. The scan parameters 

were as follows: high voltage 120 kV, tube current 300 mA, 

slice collimation 16 x 0.5 mm, rotation time 1.0 sec, CT 

pitch factor 0.6875:1 (helical pitch 11:16), scan length 55 

mm. Immediately after the data from the unenhanced helical 

scan were obtained, an injection of contrast medium was 

started. The contrast medium (iohexol 350 mgI/ml) was 

injected via antecubital vein at the rate of 2.0 ml/sec with 

power injector. The total volume of contrast medium was 1.2 

ml/body weight kg. Acquisition of data from the enhanced 

helical scan was begun 30 to 35 seconds after contrast injec- 

402 

tion was initiated. The 0.5 mm thick axial images were 

reconstructed at every 0.2 mm interval. The subtraction 

images with a misregistration artifact were corrected using 

volume position matching. Three-dimensional reconstruction 

was performed using volume rendering method on a 

workstation (Zio M900 Quadra, Ziosoft Inc., Japan). 

Subtraction 3D CT angiograms with and without using volume 

position matching were compared for their characterization 

of intracranial aneurysms and postclipping state of 


FINDINGS 

Subtraction 3D CT angiograms without or with a mild misregistration 

artifact could clearly depict intracranial 

aneurysms, the surrounding vessels, and the relation 

between the clips and necks of postclipping intracranial 

aneurysms. In several cases a moderate to severe misregistration 

artifact occurred due to the motion of patients, and 

subtraction 3D CT angiograms could not clearly depict 

intracranial aneurysms, surrounding vessels, and especially 

the relation between the clips and the necks of postclipping 

intracranial aneurysms. Subtraction 3D CT angiograms with 

using volume position matching were superior to subtraction 

3D CT angiograms without using volume position matching 

for their characterization of intracranial aneurysms and postclipping 

state of intracranial aneurysms. 

CONCLUSION 

Volume position matching was a valuable method in subtraction 

3D CT angiography with controlled-orbit helical 

scanning for the examination of pre and postoperative 


KEY WORDS: Three-dimensional CT angiography, multidetector 

row CT, intracranial aneurysm 

Functional 

40-43 


Methods and Clinical Applications of Brain Mapping 

with Functional MR Imaging 

Magalhaes, F. V. · Cruz, L. C. H. · Domingues, R. C. 

Clínica de Diagnóstico por Imagem-Barrashopping/Multi- 

Imagem Ressonancia 


PURPOSE 

Brain function mapping with functional MR imaging using 

the BOLD phenomena has progressed rapidly in the last 

years. This complex phenomena is based on various hemodynamic 

events that occur when there is neuronal activation, 

consisting of an indirect way of evaluating this activation. 

Besides experimental applications this method is gaining 

many clinical applications, being part of the everyday practice 

of many neuroradiologists. To demonstrate the methodology, 

the expected results, and the clinical applications for 

mapping of cortical areas associated with tactile, auditory, 

visual, motor, and language functions.


Ten healthy volunteers were examined on high field (1.5 T) 

MR equipments: Siemens Magnetom Symphony Maestro 

Class or Siemens Magnetom Avanto. Sensorial mapping 

consisted of right hand tactile, visual and auditory evaluation. 

Motor mapping included tongue and right hand and 

foot evaluation. Language mapping consisted of mentally 

generating sentences based on key words presented to the 

volunteers. The visual activation condition consisted of the 

projection of a reversing checkerboard and the rest condition 

of observing a fixation point. The activation and rest conditions 

of auditory evaluation consisted of the biauricular presentation 

or not presentation of instrumental music. The activation 

conditions of tactile evaluation consisted of stimulating 

the right hand with a small brush, and of motor evaluation 

of opening and closing the right hand, flexing and 

extending the right foot and moving the tongue in the mouth. 

The rest condition for these evaluations consisted of the volunteer 

staying without making any movements or receiving 

any stimulation. The images were analyzed with SPM 2 software 

and the BOLD evaluation application of Siemens 

Syngo platform. 

FINDINGS 

In all volunteers statistically significant (p < 0.01) signal difference 

was obtained when comparing the various activation 

and rest conditions. The cortical activation areas corresponded 

to the expected. The visual evaluation showed bilateral 

activation of the occipital cortex, around the calcarine 

sulcus (Brodmann area 17). Auditory evaluation showed 

bilateral activation of the temporal cortex on the Heschl 

gyrus (Brodmann areas 41 and 42). Tactile and motor evaluations 

showed activation on the post and prefrontal gyrus, 

corresponding to the stimulated body part (Brodmann areas 

1, 2, and 4). Language evaluation showed activation on the 

inferior frontal gyrus (Broca’s area - Brodmann 44 and 45) 

and on the insular cortex. 

CONCLUSION 

Functional MR imaging in a noninvasive method with good 

spatial and temporal resolutions, allowing critical brain function 

mapping, which can be used on surgical planning and 

surgical risk assessment, directing intraoperative electrophysiologic 

procedures, and opens an unlimited research 

field. 

KEY WORDS: Functional MR imaging, brain mapping 


Brainstem Lesions: MR Imaging, MR Spectroscopy, MR 

Perfusion and Positron Emission Tomography 

Horowitz, S. W. · Raizer, J. · Kasotakis, M. · Spies, S. 

Northwestern Memorial Hospital 

Chicago, IL 

PURPOSE 

To illustrate the roles of MR imaging, MR proton spectroscopy, 

MR perfusion and positro emission tomography 

(PET) in diagnosis and follow up of brainstem lesions. 

403 


Eight patients with brainstem lesions, including 6 unknowns 

and 2 severe lupus patients with old pontine infarction or 

ischemia were studied. MR imaging/MRS utilized multivoxel 

3D or 2D csi at TE 144 and TE 30 in the unknowns, and 

single voxel in the lupus patients. For unknown lesions, MR 

perfusion was included. Two of the five brainstem gliomas 

were followed every 3 months with PET, MR imaging, 

MRS, and perfusion. Each lupus patient had 3 MR imaging/MRS 

exams over 1 year, including pre and poststem cell 

transplant. The demyelinating disease patient had multiple 

brain and spinal cord MR images over 2.5 years. 

FINDINGS 

Of MRS ratios, only the Cho/NAA ratio at TE 144 was helpful 

to distinguish pontine tumefactive, ring-enhancing 

demyelinating disease (Cho/NAA = 1.5) from pontine tumor 

(Cho/NAA = 2.0 or greater). MR perfusion did not distinguish 

the large, rim-enhancing demyelinating lesion from 

tumor. Demyelinating disease was confirmed by multiple, 

separate nonenhancing lesions in normal size cord, clinical 

follow up and brainstem lesion shrinkage. 

In brainstem glioma at 3 months after RT, MRS was more 

sensitive in detection of brainstem tumor and delineation of 

extent than was FDG PET. At 3 months post-RT, PET was 

relatively insensitive for tumor detection and extent, with 

increased FDG uptake limited to small sites of gadolinium 

enhancement. At 6-month follow up, new location of tumor 

growth on MR imaging enhancement demonstrated abnormally 

elevated FDG uptake and further increased MR perfusion 

rCBV. The lupus patient with old pontine infarction pre 

and poststem cell transplant showed diminished NAA/Cr 

and NAA/Cho at TE 30 compared with controls, with no 

change on serial MRS. The lupus patient with pontine chronic 

ischemic T2 hyperintense signal changes without infarction 

showed interval improvement at 1 year posttransplant in 

the NAA/Cr and NAA/Cho ratios which had been diminished 

pretransplant. 

CONCLUSION 

In this limited patient group, MRS was sensitive in brainstem 

glioma detection, with elevated Cho/NAA ratio being 

higher in tumor (2.0) than in a tumefactive ring-enhancing 

demyelinating lesion (1.5). At 3-month follow up postradiation, 

MRS was sensitive to brainstem glioma detection, with 

interval improved degree of elevated Cho/NAA ratio. PET 

was relatively insensitive in detection and delineation of 

brainstem glioma at 3-month postradiation. Increased FDG 

uptake at a new site of tumor spread at 6-month follow up 

corresponded with MR imaging enhancement site and with 

further elevated MR perfusion rCBV ratio suggesting 

increased tumor grade. FDG-PET did not provide additional 

information beyond the combined MR imaging/MRS/MR 

perfusion in this preliminary data. In severe lupus patients, 

MRS showed diminished pontine NAA/Cr and NAA/Cho 

ratios in pontine infarction or ischemia which persisted 

unchanged in old pontine infarction reflecting permanent 

loss of neuronal NAA marker; however, the improved NAA 

ratios in ischemia without infarction raises the question of 

improved neuronal function following stem-cell transplant 

in pontine ischemia without frank infarction. 

KEY WORDS: MR spectroscopy, brainstem, glioma 

Scientific Exhibits


404 



Radiation Necrosis after Gamma Radiosurgery to Brain Diffusion Tensor Imaging and Fiber Tracking of the 

Metastasis: MR Imaging, MR Spectroscopy, MR Limbic System in the Normal Pediatric Brain and in 

Perfusion, Positron Emission Tomography 

Cerebral Malformations 

Horowitz, S. W. · O’Connor, E. · Raizer, J. · Spies, S. · 

Kaakaji, R. 

Northwestern University Memorial Hospital 

Chicago, IL 

PURPOSE 

To review radiation necrosis findings after gamma radiosurgery 

to brain metastasis, including the typical MR imaging 

appearance, multivoxel csi MR proton spectroscopy 

(MRS), MR perfusion, and positron emission tomography 

(PET). 


Eight patients with brain metastasis, treated by gamma 

radiosurgery in 7/8 (some included whole brain RT) and 

Linac-based radiosurgery in 1 patient, were diagnosed with 

radiation necrosis (RTN) by MR imaging/MRS. Multivoxel 

csi at both TE 144/30 was performed in 7/8, single voxel in 

1/8 patients. MR perfusion and PET were included in 2 

patients. Six lesions were resected in 5 patients, and followup 

MR imaging/MRS obtained in 3 patients up to 3 years in 

1/3 patients. 

FINDINGS 

In all 8 patients, MR imaging showed typical geographic, 

lobulated peripheral enhancement, many with speckled 

enhancement within the necrotic center. Increasing blood 

products on GRE due to radiation-induced vasculitis accompanied 

increased lesion size in 2 patients. Marked gyral 

enhancement was noted in 1 patient. MR spectroscopy 

showed dominant lipid peak at TE 30 at and peripheral to the 

enhancement, elevated Cho/Cr at TE 144 but decreased 

“absolute” choline amplitude relative to contralateral normal. 

MR rCBV hypoperfusion and decreased uptake on PET 

correlated with MRS/MR imaging of RTN. Of the 6 resected 

lesions, 5 showed radiation necrosis alone and 1 showed 

a small focus of metastatic tumor within a much larger volume 

of RTN. 

CONCLUSION 

Typical MR imaging appearance in radiation necrosis s/p 

gamma to brain metastasis is a geographic, lobulated, 

peripherally enhancing lesion, frequently containing speckled 

enhancement within central necrosis, and extensive 

peripheral edema. Gyral enhancement, or increasing blood 

products on GRE within an increasing size lesion may be 

seen. Enhancement size and area fluctuates over time. When 

MR imaging suggests possible RTN, correlate with MRS 

and MR perfusion. Pitfalls include possible tumor foci within 

predominant radiation necrosis. One could hypothesize a 

potential for misdiagnosis of a completely necrotic tumor 

which might be difficult to differentiate from radiation 

necrosis. 

KEY WORDS: Radiation necrosis, gamma radiosurgery, MR 

spectroscopy 

Rollins, N. 1,2 · Koral, K. 1,2 · Reyes, T. 1 · Halovanic, C. 1 · Chia, J. 3 

1 2 Children’s Medical Center, Dallas, TX, University of Texas 

Southwestern Medical Center, Dallas, TX, 3Philips Medical 

Systems, Best, THE NETHERLANDS 

PURPOSE 

Diffusion tensor imaging (DTI) and fiber tracking (FT) provide 

a novel way to evaluate white matter tracts. White matter 

of the limbic system, important in declarative memory 

and learning and known to be abnormal in many cerebral 

malformations, is readily apparent on DTI/FT from birth on. 

We evaluated the fornix and cingulum using DTI/FT in normal 

children from birth to adolescence and in multiple cerebral 



The investigation was HIPPA compliant and informed 

parental consent was obtained. In addition to routine 1.5 T 

MR imaging, DTI was acquired with a 6 channel SENSE 

head coil (Philips Medical Systems) and isotropic diffusionweighted 

gradients using single-shot spin-echo planar 

sequence [SENSE factor 2.5, EPI factor 35, b = 700, 7357/98 

(TR/TE), scan matrix 112 x 256, 246 x 246 mm FOV, 55 

slices, 2.2 mm slice thickness, no gap, 32 directions]. Fiber 

tracking was initiated by manually placing regions of interest 

(ROIs) over the cingulum and fornices on the color maps 

or anatomical coregistration images and using a “brute 

force” approach. To date 176 patients, (3 days-18 years) 

have been studied including 40 normal patients. Cerebral 

malformations include callosal agenesis without hydrocephalus 

[ACC] (n = 9), Chiari II malformation without 

ACC (n = 12), open lip schizencephaly (n = 7), lissencephaly 

and semilobar holoprosencephaly (HPE) in 1 neonate each. 

FINDINGS 

The fimbria, crura, columns, and fornix are visible by FT at 

birth as structures separate from the cingulum. The cingulum 

decreases relative to the thickness of the corpus callosum 

with increasing age. Asymmetry in the thickness of the cingulum 

is common in adolescents. The midportions of the 

cingulum and forniceal crura may appear fused in callosal 

agenesis and may be separate and normal appearing or show 

variable dysmorphism in Chiari II malformation without 

ACC. In three patients with Chiari II malformations, fibers 

of the cingulum were seen obliquely crossing the corpus callosum. 

Open lip parietal schizencephaly was associated with 

ipsilateral forniceal agenesis and hypoplasia of the ipsilateral 

cingulum; the contralateral fornix was caudally displaced 

due to septal agenesis. In lissencephaly, the columns and 

crura of the fornix and the rostral and superior portions of the 

cingulum were thickened and dysplastic while the temporal 

portions of the limbic tracts were not visible. In HPE, the 

fornices were thickened, dysplastic, and foreshortened in an 

AP direction but maintained the C-shaped configuration 

characteristic of fornices seen normal neonate. The rostral 

extension of the fornices divided into horizontally oriented 

asymmetric paramedian structures similar to the posterior 

projections of the anterior commissure and a single median

structure similar to the precommissural fornix but thicker 

and longer than the usual appearance of the precommissural 

fornix in a neonate. 

CONCLUSION 

The cingulum and fornices are abnormal in many cerebral 

malformations as seen by DTI/ FT. The appearance of the 

major limbic tracts cannot be predicted necessarily on the 

basis of findings at routine MR imaging. 

KEY WORDS: Diffusion tensor imaging, limbic system, pediatrics 

Head and Neck 

44-70 


Diffusion-Weighted Imaging in Head and Neck Tumors 

Maeda, M. 1 · Sakuma, H. 1 · Maier, S. E. 2 · Takeda, K. 1 

1 Mie University School of Medicine, Tsu, JAPAN, 2 Brigham 

and Women’s Hospital, Boston, MA 

PURPOSE 

Diffusion-weighted imaging studies comparing apparent diffusion 

coefficient (ADC) and histopathologic findings in 

brain tumors have strongly suggested that greater cellularity 

is associated with more restricted diffusion. This method 

may be useful also for the characterization of head and neck 

tumors. Although the initial studies are promising, the most 

serious drawback of echo-planar diffusion-weighted imaging 

is significant susceptibility artifacts in the head and neck. 

The purpose of this exhibit is to demonstrate advantages of 

line scan diffusion-weighted imaging over echo-planar diffusion-weighted 

imaging and to show the usefulness of ADC 

in the diagnosis of head and neck tumors. 


We have studied more than 90 cases with head and neck 

tumors, including squamous cell carcinoma, malignant lymphoma, 

and various types of benign (e.g. pleomorphic adenoma) 

or malignant tumors (e.g. leukemia, retinoblastoma). 

Line scan diffusion-weighted images were acquired using 

two different b values, with the maximum b value applied 

along the three orthogonal directions: one with a low diffusion 

weighting (b factor) of 5 s/mm 2 and the other with a 

high (maximum) b factor of 1000 s/mm 2 . Echo-planar difffusion-weighted 

imaging (9999/69.2, b = 0, 1000) was performed 

in a limited number of patients for comparison with 

line scan diffusion-weighted imaging. Apparent diffusion 

coefficient value measurements were obtained from the trace 

ADC map using line scan diffusion-weighted imaging. In 

ADC measurement of tumors, special care was taken to 

include the solid-appearing portions of the tumors and to 

exclude obviously necrotic or cystic regions demonstrated in 

the corresponding T2-weighted and contrast-enhanced MR 

images. 

405 

FINDINGS 

Line scan diffusion-weighted imaging successfully provided 

diagnostic images free from susceptibility artifacts and also 

permitted ADC values to be measured. On the other hand, 

severe susceptibility artifacts distorted echo-planar diffusion-weighted 

images, particularly in patients with dental 

work. Malignant tumors tended to be lower in ADC values 

than benign tumors. For example, mean ADC values were 

lower than 1.0 × 10 -3 mm 2 /s in squamous cell carcinoma 

(mean 0.96 × 10 -3 mm 2 /s) and malignant lymphoma (mean 

0.65 × 10 -3 mm 2 /s) while those in benign tumors such as 

pleomorphic adenoma were mostly higher than 1.0 × 10 -3 

mm 2 /s. Malignant lymphoma showed significantly lower 

ADC values than squamous cell carcinoma and other types 

of carcinoma, but ADC values in some tumors such as 

leukemia overlapped with those of lymphoma. It is a point to 

notice that ADC values in benign lymphadenopathy (lymphoid 

hyperplasia) overlapped with those of malignant lymphoma. 

The relationship between pathology and ADC values 

of tumors was briefly discussed also. 

CONCLUSION 

Line scan diffusion-weighted imaging is insensitive to susceptibility 

artifacts and permits the evaluation of ADC values 

in the head and neck tumors. Apparent diffusion coefficient 

values by line scan diffusion-weighted imaging can be 

a useful adjunct for distinguishing between benign and 

malignant tumors and narrowing the differential diagnosis. 

REFERENCES 

1. Guo AC, Cummings TJ, Dash RC, Provenzale JM. Lymphomas 

and high-grade astrocytomas: comparison of water diffusibility 

and histologic characteristics. Radiology 

2002;224:177-183 

2. Wang J, Takashima S, Takayama F, et al. Head and neck 

lesions: characterization with diffusion-weighted echo-planar 

MR imaging. Radiology 2001;220:621-630 

KEY WORDS: Diffusion-weighted imaging, neoplasm, head 

and neck 


Guide to the Olfactory System 

Fatterpekar, G. M. · Naidich, T. P. · Delman, B. N. · Hoff, P. 

R. · Som, P. M. 

The Mount Sinai School of Medicine of New York 

University 

New York, NY 

PURPOSE 

To present a teaching module on the normal anatomy and 

pathology of the olfactory system. 


The CT and MR anatomy of the olfactory system was studied 

in 5 normal volunteers, and selected cadaver specimens. 

High-resolution axial CT scans of the anterior skull base and 

nasal cavity were obtained on a 16-slice spiral CT (Somatom 

Sensation 16, Siemens Medical Solutions, Malvern, PA) 

using 120 kv, 200 mAs, and two different slice thicknesses 

(1 mm thick sections reconstructed at 0.7 mm intervals for 

the anterior skull base, and 3 mm thick sections for the nasal 

Scientific Exhibits


cavity). The axial dataset then was reconstructed in the 

orthogonal sagittal and coronal planes. Multiplanar MR 

scans of the anterior and middle cranial fossae were obtained 

using 2 mm thick T1-weighted (TR 500 ms, TE 12 ms, NEX 

6), and T2-weighted (TR 4000 ms, TE 130 ms, NEX 6) pulse 

sequences. MR microscopy of the olfactory pathway was 

performed on 3 formalin-fixed cadaver specimens using a 

9.4 T, 89 mm vertical bore MR system (Brüker Analytik, 

Rheinstetten, Germany), a 30 mm birdcage coil, and intermediate-weighted 

pulse sequences (TR = 800 - 2400 ms, TE 

= 14 - 30 ms, matrix 512 x 512, excitations = 25 - 50, and 

slice thickness 500 mm). The anatomy displayed was correlated 

with standard reference works and displayed for easy 

assimilation. The case files of the institution then were 

reviewed to select examples of olfactory pathology useful 

for illustrating anatomical points. 

FINDINGS 

The imaging studies display the full extent of the olfactory 

system. The CT images demonstrate the vomer, the expected 

locations of the vomeronasal organs, the roof of the nasal 

cavity with its olfactory epithelium, and the cribriform plate 

through which ~20 olfactory nerves pass to synapse in the 

olfactory bulb. The MR images depict the courses of the 

olfactory nerves through the cribriform plate, the relations of 

the olfactory bulb and tract to the floor of the anterior fossa 

and the overlying frontal lobe, and the division of the olfactory 

trigone into the medial, lateral, and intermediate olfactory 

striae, which pass to the septal region and medial temporal 

lobe. High resolution 9.4 T MR images of anatomical 

specimens display the cortical lamination of the olfactory 

bulb, the fibers of the olfactory tract, and the projections of 

the olfactory striae into the anterior perforated substance, the 

uncus, and the olfactory cortex of the temporal lobe. These 

anatomical relationships are summarized and “crystallized” 

in color diagrams and succinct descriptions of the anatomy 

and physiology of the olfactory system. The understanding 

gained then is utilized to explain the pathology observed in 

olfactory disorders such as holoprosencephaly, Kallmann 

syndrome, trauma, tumors like olfactory neuroblastoma 

(esthesioneuroblastoma), and uncal lesions. 

CONCLUSION 

This teaching module illustrates the normal anatomy of the 

olfactory system and provides the viewer with an approach 

to evaluating pathology of the “smell” pathway. 

KEY WORDS: Olfactory system, anatomy, pathology 

406 


Dynamic MR Imaging of Head and Neck Vascular 

Malformations and Hemangiomas 

Ohgiya, Y. 1 · Hashimoto, T. 2 · Gokan, T. 2 · Oka, M. 1 · 

Munechika, H. 2 · Ekholm, S. 1 · Westesson, P. 1 

1 University of Rochester School of Medicine & Dentistry, 

Rochester, NY, 2 Showa University School of Medicine, 

Tokyo, JAPAN 

PURPOSE 

Head and neck vascular lesions are common, particularly 

during childhood. However, physicians often confuse these 

lesions because the nomenclature for classifying these 

lesions often is used inappropriately. Therefore, knowledge 

of the imaging and treatment of these patients is essential for 

radiologists. This exhibit demonstrates characteristic imaging 

findings and practical classification for treatment of vascular 

malformations and hemangiomas using dynamic MR 

imaging. 


Between September 2001 and August 2004, nine patients 

who underwent conventional and dynamic MR imaging had 

head and neck vascular anomalies (5 hemangiomas, 4 high 

vascular malformations, 2 lymphangiomas, and 8 low-flow 

vascular malformations). The temporal resolution of the 

dynamic MR imaging was 5 sec. Percentage enhancement 

above baseline was calculated by using the formula: percentage 

enhancement above baseline = (signal intensity after 

enhancement - signal intensity before enhancement)/ signal 

intensity before enhancement ×100. Time intervals between 

the onset of the enhancement in the lesion and the time of the 

maximal percentage enhancement above baseline of the 

lesion within 120 seconds were measured. We defined these 

time intervals as contrast rise time of lesion. Statistically significant 

differences (p < 0.01) were determined using the 

Scheffe test for comparison among hemangiomas, high- and 

low-flow vascular malformations in terms of contrast rise 

time of lesion. Diagnosis of the peripheral vascular malformations 

was based on angiographic or venographic findings. 

Diagnosis of hemangiomas is based on the temporal growth 

history and appearance on physical inspection. 

FINDINGS 

The range of the contrast rise time of lesion of the hemangiomas 

was from 0 to 10 seconds. The range of the contrast 

rise time of lesion of the high-flow vascular malformations 

was 5 seconds. The range of the contrast rise time of lesion 

of the low-flow vascular malformations was from 50 to 105 

seconds. The lymphatic malformations showed lack of 

enhancement. The mean contrast rise time of lesion of the 

hemangiomas was significantly shorter than that of the lowflow 

vascular malformations (Scheffe; P < 0.01). The mean 

contrast rise time of lesion of the high-flow vascular malformations 

was significantly shorter than that of the low-flow 

vascular malformations (Scheffe; P < 0.01). No significant 

differences in the mean contrast rise time of lesion existed 

between the hemangiomas and the high-flow vascular malformations.

CONCLUSION 

MR imaging is excellent for defining the extension of 

hemangiomas and vascular malformations and their relationship 

to adjacent structures, such as neurovascular bundles. In 

addition to defining the extent of hemangiomas and vascular 

malformations on conventional MR images, dynamic MR 

images can provide information about hemodynamics of 

vascular lesions. MR imaging may become one-stop examination 

for evaluation of vascular malformations. 

KEY WORDS: Dynamic MR imaging, vascular malformations, 

hemangiomas 


F-18 Fluorodeoxyglucose Positron Emission 

Tomography and MR Imaging in Head and Neck Tumors 

Moritani, T. · Smoker, W. R. K. · Lee, H. · Menda, Y. · Maley, 

J. · White, M. M. · Graham, M. M. 


Iowa City, IA 

PURPOSE 

1. Review those head and neck structures that normally 

exhibit physiologic F-18 fluorodeoxyglucose (FDG) uptake. 

2. Present the positron emission tomography (PET) and MR 

appearances of various head and neck tumors. 


F-18 fluorodeoxyglucose PET is more sensitive and specific 

than CT for detection of primary and recurrent head and 

neck cancers. However, there is limited data in the literature 

comparing PET and MR imaging in this patient population. 

We performed PET, or PET/CT, and MR imaging on over 

100 patients with pathologically proven head and neck 

tumors. 

FINDINGS 

Physiologic FDG uptake is demonstrated in normal structures, 

including various muscles, vocal cord, salivary glands 

and lymphoid tissues, and correlated with MR imaging. 

Benign conditions with abnormal FDG uptakes include 

postradiation inflammation of salivary glands and mucosal 

tissue. Head and neck tumors we have studied include various 

types of carcinoma in the oral and nasal cavities, larynx, 

tonsils, and paranasal sinuses, metastatic lymph nodes, 

benign and malignant salivary gland and thyroid tumors, 

basal cell carcinoma, esthesioneuroblastoma, rhabdomyosarcoma, 

osteosarcoma, and melanoma. MR findings correlated 

very well with those of PET for both primary and metastatic 

lesions. MR imaging seems to be especially useful in 

demonstrating small anatomical structures, perineural 

spread, and intracranial extension of tumors. 

CONCLUSION 

We illustrate physiologic uptake of FDG, and abnormal 

uptake by benign and malignant lesions of the head and 

neck. We discuss the effectiveness and clinical applications 

of PET and MR imaging in the evaluation of patients with 

head and neck tumors. 

KEY WORDS: PET, MR imaging, head and neck 

407 


Imaging of Osteoodontokeratoprosthesis Surgery: What 

the Radiologist Should Know 

Sheah, K. 1 · Chan, L. 1 · Theng, J. 2 · Tay, A. 3 · Lim, W. 1 · Tan, D. 2 

1Singapore General Hospital, Singapore, SINGAPORE, 

2Singapore National Eye Centre, Singapore, SINGAPORE, 

3National Dental Centre, Singapore, SINGAPORE 

PURPOSE 

1) Increase awareness of the CT findings in different stages 

of osteoodontokeratoprosthesis (OOKP) surgery with intraoperative 

photographic correlation. 2) Highlight the important 

reporting points in OOKP surgery imaging. 


Five patients underwent osteoodontokeratoprosthesis surgery 

at our institution from 2003 to 2004. This procedure is 

rarely performed. It usually is considered only in patients 

with corneal blindness refractory to keratoplasty, and is performed 

as a 2-staged procedure. Strampelli initially 

described the procedure, using the patient’s own tooth root 

and alveolar bone as a vital support for an optical cylinder, 

and has been modified over the years by Falcinelli. Surgery 

is performed usually in two stages 2 to 4 months apart, which 

allows soft tissue to grow around the osteoodonto lamina 

scaffold and for the ocular globe graft (with buccal mucous 

membrane) to become vascularized. The first stage involves 

ocular surface reconstruction and creation of an osteoodonto 

lamina and optical cylinder from the patient’s tooth (usually 

canine). Buccal mucosa is grafted on the sclera, and the lamina 

placed in a submuscular pocket. CT scan during this period 

is important for assessment of the cylinder and to detect 

complications from the procedure. The second stage 

involves total iridodialysis, lens extraction, and anterior vitrectomy 

with implantation of the prosthesis. The eye is then 

reinflated with air and buccal mucosal flap replaced. The CT 

findings in both stages of the operation will be demonstrated. 

FINDINGS 

Measurement of the osteoodonto lamina on CT in axial plane 

and with three-dimensional (3D) reconstruction is important 

as a baseline for subsequent follow up to look for prosthesis 

resorption. We typically use 3D reconstruction in soft tissue 

VRT windows to image the lucent optical cylinder. The ideal 

measurement for the optical cylinder is about 3.5 x 8 mm in 

cross section, and we demonstrate the best window settings 

to view this based on correlation with intraoperative measurements. 

Measurement of the osteoodonto lamina is also 

important, and this usually is done in bone windows. The 

dimensions obtained immediately after surgery serve as a 

baseline for comparison to detect early prosthesis resorption. 

Finally, assessment of surgical sites for complications such 

as inflammation or collection should be included also in the 

postoperative CT scan reports. 

CONCLUSION 

CT with 3D reconstruction is an integral part of osteoodontokeratoprosthesis 

surgery for evaluation of the osteoodonto 

lamina and optical cylinder, implantation site in the submuscular 

pocket of the cheek, and buccal mucosal membrane 

Scientific Exhibits


408 

graft site on the orbit. We foresee greater use of this proce- Scientific Exhibit 50 

dure in the future, and it is important that the radiologist recognize 

the pertinent imaging findings. 

Imaging of Head and Neck Venous Malformations: A 

Pictorial Review 

KEY WORDS: Osteoodontokeratoprosthesis, method, CT 


Pterygopalatine Fossa: A Pictorial Review of the 

Anatomy and Related Disease 

Basmaji, C. N. · Noujaim, S. 


Royal Oak, MI 

PURPOSE 

The purpose of this pictorial exhibit is to familiarize and sensitize 

physicians to early recognition and diagnosis of various 

pathologic entities related to the pterygopalatine fossa 

and the mechanism of disease spread. 


The exhibit will incorporate a number of related cases from 

a large tertiary medical center utilizing various imaging 

modalities including CT, MR imaging, and angiography. 

Both common and uncommon pathologic processes involving 

the pterygopalatine fossa will be demonstrated as well as 

their mechanism of spread to adjacent intra and extracranial 

structures via various fissures and foramina. 

FINDINGS 

After reviewing the detailed anatomy of the pterygopallatine 

fossa using anatomical drawings and cross-section images, 

several examples of both benign and malignant pathologic 

entities will be used to demonstrate the mechanism of disease 

spread to adjacent intra and extracranial structures. 

Examples include minor salivary gland tumors, lymphoma, 

and both malignant and benign aerodigestive and sinonasal 

tract tumors (carcinomas, rhabdomyosarcoma, and juvenile 

nasopharyngeal angiofibroma), as well as meningioma of the 

skull base, and fungal and bacterial sinusitis. 

CONCLUSION 

Familiarization of the complex anatomy and pathologic entities 

related to the pterygopalatine fossa may lead to earlier 

and more accurate diagnosis of various disease processes by 

physicians. 

KEY WORDS: Pterygopalatine fossa, anatomy, disease 

Flis, C. M. · Sibtain, N. A. · Connor, S. E. J. 

King’s College Hospital 


PURPOSE 

To review the classification of venous malformations in the 

context of head and neck vascular malformations and 

tumors. To describe and illustrate the imaging findings using 

ultrasound, MR imaging and CT, conventional angiography, 

and percutaneous phlebography and explain how to differentiate 

from other vascular/lymphatic malformations. To outline 

the specific role of these imaging techniques and imageguided 

sclerotherapy. 


The plain film, ultrasound (including color and duplex 

Doppler), CT and MR imaging as well as conventional 

angiographic appearances of head and neck venous malformations 

will be described and illustrated together with examples 

of image-guided sclerotherapy. A range of head and 

neck subtypes including orbit, oral cavity, superficial and 

deep facial space, supraglottis, intraosseous and intramuscular 

will be illustrated. 

FINDINGS 

Venous malformations are nonproliferative low-flow vascular 

lesions that consist of dysplastic venous channels. 

Ultrasound and MR imaging are the imaging modalities of 

choice with a supplementary role for CT. Characteristic 

appearances of venous malformations include the presence 

of phleboliths and markedly hyperintense lobulated lesions 

on T2-weighted MR sequence. Image-guided sclerotherapy 

is the first-line treatment option in selected cases. 

CONCLUSION 

An understanding of the imaging characteristics of venous 

malformations of the head and neck is important to enable 

the correct diagnosis. The radiologist plays a pivotal role in 

selecting the most appropriate diagnostic and therapeutic 

modalities. 

KEY WORDS: Malformation, venous 


Dermoids of the Face and Skull 

Camacho, D. L. A. · Spampinato, M. V. · Grimme, J. D. · 

Castillo, M. 

University of North Carolina at Chapel Hill 


PURPOSE 

To review the embryology of dermoids of the face and skull, 

illustrate the spectrum of CT and MR appearances, and 

demonstrate potential pitfalls in diagnosis of these lesions.


Following a discussion of the embryology of these lesions, a 

pictorial gallery of dermoids of the face and skull is presented. 

The cases are organized by anatomical location, with 

numerous examples illustrating the spectrum of possible CT 

and MR imaging findings. Several cases are complemented 

by gross pathologic photographs. 

FINDINGS 

Dermoids arise from trapped pouches of ectodermal elements 

during neural tube closure. Dermoids commonly are 

located at embryologic sites of cutaneous or osseous fusion 

that form the structures of the eyes, ears, face, and skull. 

Whereas epidermoids typically are located laterally and contain 

only squamous epithelium, dermoids are typically midline 

and contain skin appendages — sweat glands, hair follicles, 

and sebaceous glands. Dermoids are unilocular, cystic 

lesions which expand slowly over years by the accumulation 

of cutaneous products. Lesions may rupture if they are large, 

or become infected if they are associated with a sinus tract. 

Rarely, they may undergo malignant transformation into 

squamous cell carcinoma. CT imaging reveals well delineated, 

cystic masses of fat attenuation. Diploic lesions cause 

expansile bone erosion with sclerotic remodeling. In contrast 

to epidermoids, the walls of dermoids tend to be thicker, and 

may enhance or calcify. On MR imaging, dermoid contents 

demonstrate T1 shortening simulating fat. Prominent lipid 

peaks can be seen on MR spectroscopy. If infected, the 

lesion may show restricted water motion on diffusionweighted 

imaging. 

CONCLUSION 

Understanding the embryology underlying the development 

and preferred anatomical locations of dermoids of the face 

and skull, the typical and atypical imaging manifestations, 

and potential complications — including rupture, superinfection, 

or malignant transformation — will aid in the accurate 

diagnosis and optimal management of these lesions. 

KEY WORDS: Dermoid 


Traumatic Soft Tissue Injuries of the Orbit 

Jayaraman, S. · Delman, B. N. · Naidich, T. P. · Som, P. M. 

The Mount Sinai Medical Center of New York University 

New York, NY 

PURPOSE 

To illustrate the spectrum of ocular and orbital disorders 

caused by blunt and penetrating trauma. 


We performed a retrospective analysis of CT and MR imaging 

of orbital trauma during the past 3 years at our institution. 

We correlated imaging characteristics to clinical and 

pathologic findings. 

FINDINGS 

Though the globe and orbit encompass a small portion of the 

craniofacial anatomy, trauma to this region has dire consequences 

due to impairment of vision. A variety of conditions 

caused by blunt and penetrating trauma are illustrated 

409 

including hyphema, dislocated lens, retinal detachment, 

choroidal detachment, vitreous hemorrhage, globe rupture, 

and optic nerve injury. In addition, direct and indirect effects 

of orbital fractures on the orbital soft tissues are described, 

including muscular impingement from a displaced fracture 

fragment and a subperiosteal hematoma causing ocular compression. 

CONCLUSION 

A systematic approach to the orbit, from the anterior chamber 

of the globe to the orbital apex, provides for accurate 

identification of traumatic pathology. Rapid diagnosis is 

essential to guide appropriate clinical management. 

KEY WORDS: Orbit, trauma, ocular 


CT Imaging for Dental Implants 

Kakimoto, N. 1,2 · Larheim, T. A. 1,3 · Murakami, S. 2 · 

Furukawa, S. 2 · Westesson, P. 1 

1University of Rochester Medical Center, Rochester, NY, 

2Osaka University Graduate School of Dentistry, Suita, 

Osaka, JAPAN, 3University of Oslo, Oslo, NORWAY 

PURPOSE 

The purpose of this exhibit is to present a systematic 

approach for evaluating CT imaging obtained for dental 

implants. This area is quite familiar to the oral surgeon and 

the dentist but the general radiologists often feel insecure 

about the anatomy and terminology. This educational exhibit 

will present a systematic step-by-step approach how to dictate 

a dental CT scan. 


This exhibit is based on many years of experience of dental 

CT imaging. Thin, contiguous images obtained parallel to 

the alveolar ridge of the maxilla and/or mandible and reformatted 

perpendicular to the alveolar ridge. Additional 

oblique images were obtained as well panoramic views. 

FINDINGS 

The height, width, and contour of the alveolar process is a 

first assessment in dental CT imaging. Exact measurements 

in mm are needed for determining where to place the 

implants. The exhibit will illustrate how to obtain these 

measures. Reformatted panoramic images are especially 

helpful to recognize the mandibular canal and measure the 

distance from the top of the alveolar ridge to the canal. This 

measurement will be illustrated. There also will be illustrations 

of unacceptable image quality and mistakes in measurements. 

Complications resulting from inaccurate radiographic 

diagnosis will be illustrated with CT images. 

CONCLUSION 

This educational poster will illustrate how to report dental 

CT imaging specifically related to placement of dental 

implants. 

KEY WORDS: Dental implant, CT imaging, reformat 

Scientific Exhibits



MR Imaging of Optic Nerve Degeneration 

Lee, H. K. · Smoker, W. R. K. · Sato, Y. · Moritani, T. · Lee, A. 

University of Iowa Hospital and Clinics 

Iowa City, IA 

PURPOSE 

Optic nerve degeneration (OND), which is not clearly understood, 

eventually produces irreversible damage to retinal 

ganglionic cell axons with loss of vision. Like wallerian 

degeneration, OND occurs in either the proximal or distal 

segment of the optic nerve, secondary to optic nerve axonal 

injury or retinal cell damage. The purpose of this exhibit is 

to discuss the pathogenesis of OND and demonstrate various 

conditions inducing OND from either an antegrade or retrograde 

direction. 


High-resolution MR imaging is required for precise optic 

nerve imaging. Antegrade OND occurs in the distal segment 

of the optic nerve secondary to retinal ganglionic cell damage. 

This antegrade degeneration is seen in long-standing 

severe ocular disease such as absolute glaucoma, following 

anterior ischemic optic neuropathy, trauma, or enucleation. 

Retrograde OND occurs in the proximal segment of the optic 

nerve secondary to axonal injury. It occurs following trauma 

to the intracranial optic nerve or optic chiasm, in association 

with optic nerve/chiasm compression from pituitary 

macroadenoma or internal carotid artery aneurysm, following 

occipital infarction with secondary OND, or in association 

with demyelinating lesions such as optic neuritis, toxic, 

and metabolic lesions that affect the OND. 

FINDINGS 

On fat suppressed T2-weighted MR images, the diagnosis of 

OND is made by identification of the small size and high signal 

intensity of the optic nerve following optic nerve damage. 

Additional contrast-enhanced T1-weighted images with 

fat suppression are required to differentiate OND from other 

optic nerve pathology including optic nerve edema or optic 

neuritis. Various conditions inducing OND are grouped into 

antegrade, retrograde and combined conditions. 

CONCLUSION 

High-resolution MR imaging is required to demonstrate 

OND. Knowledge of OND helps us to understand the clinical 

course of optic nerve damage and to differentiate it from 

other optic nerve pathology. 

KEY WORDS: Optic nerve, optic pathway, MR imaging 

410 


Pictorial Essay: Orbital Lesions 

Pinto, M. H. 1 · Brooks, M. 1 · Fertikh, D. 2 · Shah, V. 3 Grissom, L. 

· 

4 

1 2 Mercy Catholic Medical Center, Darby, PA, University of 

Pennsyvania, Philadelphia, PA, 3Thomas Jefferson 

University Hospital, Philadelphia, PA, 4A.I duPont Hospital 

for Children, Wilmington, DE 

PURPOSE 

The purpose of this exhibit is to display a multimodality pictorial 

review of imaging findings of various ocular and orbit 

pathology. 


Imaging studies of patients with orbital symptoms are analyzed 

retrospectively. 

FINDINGS 

Various pathologies are characterized as follows: 1) 

Inflammatory etiology such as pseudotumor, postseptal 

abscess, Wegener’s disease, thyroid ophthalmopathy; 2) 

Vascular lesions such as varix, carotid cavernous fistula; 3) 

Neoplasm including hemangioma, lymphangioma, optic 

nerve sheath meningioma, optic nerve glioma, lymphoma, 

and metastases; 4) Trauma: hematoma and fracture. 

Characteristic CT and MR appearances of these disease entities 

will be described. Pertinent patients’ demographic and 

clinical information which help narrow the differential diagnosis 

will be emphasized. 

CONCLUSION 

Radiologic evaluation is important in evaluating patients 

with orbital symptoms. We presented key findings of different 

orbital entities with the goal to improve diagnosis accuracy 

and hence guide appropriate patients management and 

treatment. 

KEY WORDS: Orbit, inflammatory, neoplasm 


Imaging Findings of the Giant Cell Tumors of the Skull 

Kashiwagi, N. 1 · Kikuyama, A. 2 · Ishikura, R. 3 · Morino, H. 2 

· Taki, T. 2 · Fujita, N. 4 · Hirabuki, N. 5 

1Osaka Medical Center for Cancer and Cardiovascular 

Diseases, Osaka City, JAPAN, 2Kansai-Rosai Hospital, 

Amagasaki City, JAPAN, 3Hyogo College of Medicine, 

Nishinomiya City, JAPAN, 4Osaka University Medical 

School, Suita City, JAPAN, 5Kinki-Chuou Hospital, Itami 

City, JAPAN 

PURPOSE 

To investigate CT and MR findings of giant cell tumor 

(GCT) of skull base which is unusual site on GCT.


Five patients from 12 to 35 years with histologically proven 

GCT of skull base were reviewed retrospectively. The 

images were analyzed for lesion location, extent of tumor, 

bone change on CT, signal intensities, and enhancement pattern 

on MR imaging. 

FINDINGS 

Three were originated from temporal bone and two were 

originated from sphenoid bone. In the GCT of the temporal 

bone, the tumors were located in the petrous and squamous 

portion with predominant extension to middle cranial fossa. 

In the GCT of the sphenoid bone, the tumors were centered 

in the body of the sphenoid bone and were symmetrical in 

shape. Bone changes in the tumor of temporal bone showed 

included cortical thinning, cortical resorption, expansive 

remodeling, endosteal scalloping, and woven bone fragment. 

Those of the sphenoid bone showed purely osteolytic 

change. Although MR signal intensities and enhancement 

pattern were variable in GCT of the both sites, all of the three 

tumors of the temporal bone had marked low intensity are on 

T2-weighted imaging in the peripheral area of the tumor; on 

the other hand, both tumors of the sphenoid bone did not 

have marked low intensity area on T2-weighted imaging. 

CONCLUSION 

Giant cell tumors of the skull have two preferential sites for 

occurring and have characteristic tendencies in its extent. 

Bone change and MR signal showed some differences 

between the two sites. 

KEY WORDS: Giant cell tumor, skull, MR imaging 


Imaging of Dental Infection 

Liu, X. · Westesson, P. 

University of Rochester School of Medicine & Dentistry 

Rochester, NY 

PURPOSE 

Describe CT and MR manifestations of dental infections. 


CT imaging of dental abscess, osteomyelitis, and dental peritonitis 

are reviewed and systematized. 

FINDINGS 

Dental abscesses starts with a nodular well defined hypodensity 

with high T2 signal. There is no enhancement. These 

lytical lesions can evolve into a more extensive lytic bone 

destruction consistent with osteomyelitis. 

CONCLUSION 

Osteoradionecrosis pathologic factors are seen as radiolucent 

lines and break of cortex. Subtissue abnormality in the 

adjacent soft tissue with the fatty is often clean. Contrastenhanced 

images are most helpful for determining the extent 

of the infection. This educational poster demonstrates CT 

and MR features of dental infections. 

KEY WORDS: CT, dental infection 

411 


Exploring the Buccal Space: A Review of Unfamiliar 

Territory 

Dunfee, B. L. · Sakai, O. · Begelman, K. · Barest, G. 

Boston University Medical Center 

Boston, MA 

PURPOSE 

The buccal space (BS) composes the substance of the cheek. 

Its contents and borders can be visualized well with CT and 

MR imaging; however, evaluation of this space often is overlooked 

both clinically and radiographically. The BS is bound 

by the buccinator muscle and alveolar ridge medially, by the 

masticator space and parotid space posteriorly, and by the 

superficial layer of the deep cervical fascia and muscles of 

facial expression anterolaterally. Superiorly the BS extends 

to the infratemporal fossa and inferiorly to the submandibular 

space. The goals of this exhibit are to review the anatomy 

of the BS, to discuss the imaging characteristics and differential 

diagnosis of BS lesions, and to familiarize radiologists 

with an often forgotten component of facial anatomy. 


We retrospectively reviewed CT and MR imaging studies of 

patients with buccal space pathologies from the imaging 

database. 

FINDINGS 

CT and MR imaging anatomy of the buccal space is 

reviewed. Both benign and malignant lesions can occur in 

the BS. Minor salivary gland tumor and vascular malformation 

are more common, though sarcoma, lymphoma, lipoma, 

schwannoma, and other rare entities have been reported. 

Malignant tumors can involve the BS by local extension 

from an adjacent space, commonly with gingival squamous 

cell carcinoma or by nodal or bone metastases. Inflammatory 

lesions are also common. Gas within the BS may serve as a 

clue to subtle trauma. 

CONCLUSION 

A thorough understanding of buccal space anatomy and 

imaging characteristics of various lesions within this region 

is crucial in developing an accurate diagnosis and impacting 

patient management. 

KEY WORDS: Buccal space, cheek 


It’s Not a Cervical Lymph Node — It’s a Vein! A Review 

of the CT and MR Findings of the Veins in the Head and 

Neck 

Escott, E. J. · Branstetter, B. F. 



PURPOSE 

The anatomy and caliber of veins can be quite variable. On 

unenhanced neck CT scans, aberrant veins may be difficult 

to differentiate from lymph nodes or other pathology. Even 

Scientific Exhibits


on contrast-enhanced studies, differential enhancement of 

veins or mixing of opacified with nonopacified blood can 

have a confusing appearance, particularly if the vein is focally 

dilated. Similarly, venous size and signal may be variable 

on MR scanning due to slow or turbulent flow, or variable 

enhancement. A thorough knowledge of venous anatomy and 

potential variants is necessary to properly differentiate an 

unopacified or focally dilated vein from lymphadenopathy 

or other pathology. 


The venous anatomy of the head and neck is reviewed, with 

an emphasis on the CT and MR manifestations of common 

anatomical variations. Examples of normal variants and 

pathologic mimics are shown. 

FINDINGS 

Venous anatomy and appearance can be variable, and normal 

veins may mimic pathology on CT and MR scans. 

CONCLUSION 

A thorough understanding of normal venous anatomy and 

common variants can help the radiologist to avoid misdiagnosing 

normal veins as lyphadenopathy or other pathology. 

KEY WORDS: Vein, vascular, neck 


Assessment of Coronal Reconstruction on Multislice CT 

for Lesion Extent in the Head and Neck 

Toyoda, K. · Shimizu, S. · Aoyagi, Y. 

Tokyo Dental College Ichikawa General Hospital 

Ichikawa-Shi, JAPAN 

PURPOSE 

Multislice CT enables imaging of the head and neck with 

high temporal resolution and reformatting coronal planes. 

The purpose of our study is to evaluate the usefulness of 

multislice CT (MSCT) coronal reconstructions for assessing 

lesion extent and to demonstrate the spread of the lesion in 

the head and neck, where anatomical structures are complicated. 


Subjects were 60 patients with head and neck disease who 

underwent MSCT: 15 with cellulitis or abscess, 12 with lymphadenitis, 

9 with metastatic lymphadenopathy, 10 with 

oropharyngeal cancer, 9 with laryngeal or hypopharyngeal 

cancer, and 5 with malignant lymphoma. Sixteen-row MSCT 

(IDT16; Philips, Netherlands) with collimation of 16 x 0.75 

mm was used. For each patient, sections were obtained from 

the skull base to the thoracic inlet; 4 mm thick transverse 

slices and 3 mm thick reconstructed coronal slices. 

Intravenous iodine contrast agents were used. On these coronal 

and transverse images, qualitative lesion conspicuity was 

evaluated, and the maximum diameter of the lesion was 

measured retrospectively. In 11 patients with complication of 

lymph node enlargement secondary to the abscess or malignant 

laryngeal or pharyngeal tumor, the conspicuity and size 

of the lymph node were evaluated also. 

412 

FINDINGS 

In all patients, high quality coronal images were obtained 

with few step artifacts, which offered the lesion conspicuity 

equal to that on transverse images. Coronal images provided 

easy evaluation of cranio-caudal extension and continuity of 

the lesion on fewer slices. In cases with deep neck abscess, 

coronal imaging is valuable in the evaluation of the pathway 

of spread to the retropharyngeal space or upper mediastinum. 

In cases with metastatic lymphadenopathy, inside 

noncontrast enhancing parts, nodal shape and extranodal 

spread were evaluated easily on coronal images in combination 

with transverse images. In patients with lymph node 

enlargement, the cranio-caudal extension tended to be larger 

than the anteroposterior extension. The maximum diameter 

of the lesion measured on coronal images was larger than 

that on transverse images in patients with 10 cellulitis or 

abscess (66.7%), 15 out of 17 inflammatory lymphadenopathies 

(88.2%), 11 out of 15 metastatic lymphadenopathies 

(73.3%), 3 oropharyngeal cancers (30%), 3 

laryngohypopharyngeal cancers (33.3%), and 7 malignant 

lymphomas (100%). Though transverse images were more 

visible in 3 patients with laryngohypopharyngeal cancer than 

coronal images, coronal images were useful in evaluating the 

subglottic or parapharyngeal extension. 

CONCLUSION 

Coronal images provided easy evaluation of cranio-caudal 

extension of the lesion. As head and neck diseases often 

extend in cranio-caudal direction, adding reconstructed coronal 

images is useful in MSCT examination. It provides helpful 

findings for better understanding of the lesion spread in 

cases with inflammation and malignancy, and for determination 

of planning of treatment such as radiation field in cases 

with malignancy. 

KEY WORDS: Multislice CT 


Nodular Fasciitis in the Head and Neck: CT and MR 

Imaging Findings 

Kim, S. 1 · Kim, H. 1 · Park, S. 2 · Baik, J. 1 · Byun, H. 1 · Weon, 

Y. 1 · Kim, Y. 2 

1 Samsung Medical Center, Sungkyunkwan University, 

Seoul, REPUBLIC OF KOREA, 2 Inha University Hospital, 

Incheon, REPUBLIC OF KOREA 

PURPOSE 

To describe the CT and MR imaging findings of nodular 

fasciitis occurring in the head and neck region. 


CT (n = 6) and MR (n = 4) images obtained from seven 

patients (3 men, 4 women; mean age, 19.6 years; age range, 

1 - 49 years) with surgically proved nodular fasciitis in the 

head and neck were reviewed retrospectively. All patients 

presented with a palpable mass in the head and neck, noticed 

1-3 months before; five in the face including the lateral 

malar area (n = 2), periparotid area (n = 1), upper lip (n = 1), 

and temple (n = 1), one in the occipital scalp, and the remaining 

one in the supraclavicular fossa. We investigated the CT 

and MR imaging characteristics with emphasis on the loca-

tion in relation to the deep cervical fascia, size, and margin 

of the lesion. The enhancement pattern and signal intensity 

on MR imaging were documented also. 

FINDINGS 

All lesions appeared as a discrete mass on imaging. Four 

lesions were solid and three lesions were predominantly cystic 

in appearance. Six lesions including one in the supraclavicular 

fossa were located superficial to the deep cervical fascia, 

mostly just beneath the superficial muscle of the face 

and neck. One lesion was located deep to the temporalis 

muscle. The lesion size ranged from 1.0 cm to 4.6 cm in 

greatest diameter with a mean of 2.2 cm. Five lesions were 

well marginated, while two lesions were poorly marginated. 

All four lesions of solid type showed moderate to marked 

diffuse enhancement, while three lesions of predominantly 

cystic type demonstrated peripheral, smooth (n = 2) or nodular 

(n = 1) enhancement. Four lesions also showed linear 

enhancement at the adjacent superficial or deep cervical fascia. 

Compared to the adjacent muscle, the signal of the solid 

portions of all four lesions studied with MR imaging were 

isointense on T1-weighted imaging and hyperintense on T2weighted 

imaging. The deep-seated lesion occurring at the 

temple grew aggressively, extending into the orbit and cranial 

cavity with massive bone erosion. 

CONCLUSION 

Although rare, nodular fasciitis may occur as a discrete solid 

or cystic mass in the head and neck, depending on the predominant 

stromal components. Superficial location and 

moderate to marked enhancement of the solid component 

seems to be rather characteristic of the disease, especially in 

patients with a recently developed, rapidly growing mass. 

KEY WORDS: Fasciitis, head 


Laryngocele: A Spectrum of Presentation in CT and MR 

Imaging 

Palacios, E. 1 · Robertson, H. J. 1 · Kroma, G. 1 · Eraso, A. 2 

1 Louisiana State University Health Sciences Center, New 

Orleans, LA, 2 Veteran’s Administration Hospital, New 

Orleans, LA 

PURPOSE 

The spectrum of imaging findings on CT and MR scans and 

clinical manifestations of laryngocele is presented. The association 

of laryngocele with underlying disease, including 

laryngeal carcinoma is noted. 


Eleven examples of laryngocele demonstrated on CT and/or 

MR scans were correlated with clinical manifestations. The 

radiologic differential diagnosis is illustrated. 

FINDINGS 

Laryngocele develops from the appendix or lateral sacculus 

of the laryngeal ventricle as a result of excessive laryngeal 

pressure or appendiceal orifice obstruction by carcinoma, 

trauma, or inflammation. There were 2 intralaryngeal, 3 

external, 4 mixed, and 2 bilateral laryngoceles. Three laryn- 

413 

goceles developed in association with carcinoma involving 

the laryngeal ventricle. Both air- and fluid-filled laryngoceles 

were noted. 

CONCLUSION 

Laryngocele presents a variety of findings on CT and MR 

scans. Careful imaging and direct laryngoscopy are necessary 

in every case to identify any underlying carcinoma or 

other contributing disease. 

KEY WORDS: Laryngocele, laryngeal carcinoma 


The Use of Different Imaging Modalities in the Diagnosis 

of Pediatric Neck Masses 

Chung, T. W. 1,2 · Hunter, J. V. 2 

1 Chonnam University Hospital, Gwangju, REPUBLIC OF 

KOREA, 2 Texas Children’s Hospital, Houston, TX 

PURPOSE 

Neck masses are a common problem in childhood and may 

pose a diagnostic dilemma. We propose to discuss the best 

use of the various imaging modalities in the differential diagnosis 

of neck masses in children, using illustrative examples. 


We retrospectively reviewed all cases of surgically confirmed 

masses presenting with a clinical lump in the neck, 

referred to a tertiary referral pediatric hospital over the past 

5 years. We analyzed the imaging findings, such as ultrasonography, 

nuclear medicine, CT and MR imaging, clinical 

records, and pathologic findings. We classify neck masses on 

the basis of their etiology into congenital and acquired 

lesions, vascular malformations, inflammatory lesions, 

benign and malignant tumors. 

FINDINGS 

Most neck masses result from congenital or inflammatory 

processes. Neoplastic tumor is rare in childhood. Congenital 

masses include branchial cleft cysts, thyroglossal duct cyst, 

ranula, dermoid/teratoma, and ectopic thymic cyst. While 

they often manifest as cystic masses in the case of a complication 

such as infection or hemorrhage, a thick irregular rim 

is seen on CT and MR imaging. Contrast enhancement may 

be useful. Visualization of a tract utilizing heavily T2weighted 

MR imaging is important for diagnosis and presurgical 

planning. Fetal MR imaging has proven invaluable in 

the predelivery assessment of the size and extent of neck 

masses and their impact on the airway which can determine 

the need for C-section. Vascular disorders include hemangioma, 

lymphangioma, arterial and venous vascular malformations, 

internal jugular phlebectasia, and fibromatosis coli. 

The noninvasive imaging modalities that are most useful in 

diagnosing hemangioma and vascular malformation are MR 

angiography and Doppler ultrasonography. Inflammatory 

disorders include lymphadenitis with or without abscess formation 

and sialadenitis. Ultrasonography can be beneficial 

in the detection of abscess formation, to exclude a solid mass 

and in the application of Doppler to rule out a vascular component 

as well as to guide catheter drainage. Tumors include 

benign tumors, such as neurofibroma, lipoma, aggressive 

Scientific Exhibits


414 

fibromatosis, and myofibroma, and malignant ones, such as Scientific Exhibit 66 

rhabdomyosarcoma, lymphoma, thyroid cancer, parotid 

tumor, metastatic lymphadenopathy, and neuroblastoma. 

Multimodality imaging studies for malignant tumors provide 

helpful informations for accurate diagnosis as well as tumor 

staging. 

Tumors and Tumor-Like Lesions of the Nasal Cavity, the 

Paranasal Sinuses, and the Adjacent Skull Base: 

Correlation of Apparent Diffusion Coefficients with 

Histologic Feature 

CONCLUSION 

A single study is sometimes insufficient to make a definitive 

diagnosis because of the overlap of imaging findings in various 

childhood “lumps and bumps” in the neck. It may therefore 

be necessary to image with a variety of modalities with 

or without contrast administration in order to differentiate 

masses as well as classify and treat them. This poster aims to 

help the radiologist triage and accurately diagnose neck 

masses in childhood in the most direct fashion while limiting 

radiation exposure. 

KEY WORDS: Neck, abnormalities 


Pictorial Review of Imaging Findings in Congenital 

Middle Ear Deafness on High-Resolution CT of the 

Temporal Bone 

Tang, P. 1 · Goh, J. 1 · Tan, T. 2 · Smoker, W. 3 · Gentry, L. 4 

1 Tan Tock Seng Hospital, Singapore, SINGAPORE, 2 Changi 

General Hospital, Singapore, SINGAPORE, 3 University of 

Iowa Hospitals and Clinics, Iowa City, IA, 4 University of 

Wisconsin Medical School, Madison, WI 

PURPOSE 

To illustrate the gamut of congenital causes of conductive 

hearing loss (CHL) on high-resolution CT (HRCT) in children 

and young adults with normal external ears and no past 

history of ear discharge or trauma. 


We performed a retrospective review of patients with CHL 

from 4 teaching hospitals in Singapore and the United States. 

All patients had undergone HRCT on single or multislice 

scanners. Both axial and coronal planes were acquired from 

the head of the malleus to the round window when a singleslice 

CT scanner was utilized. On the multislice CT scanner, 

axial images were obtained at 0.5 mm intervals with coronal 

reformats. 

FINDINGS 

We demonstrate various causes of congenital CHL which 

include congenital ossicular chain fixation, congenital oval 

window atresia, congenital round window atresia, absent 

lenticular process of the incus, and a monopodal stapes. 

CONCLUSION 

Most causes of congenital middle ear deafness can be 

demonstrated on HRCT without requirement for exploratory 

tympanotomy. 

KEY WORDS: Congenital 

White, M. L. · Zhang, Y. · Robinson, R. A. 


Iowa City, IA 

PURPOSE 

Measurement of apparent diffusion coefficient (ADC) values 

may be useful in grading tumors because of the significantly 

inverse correlation between ADC values and tumor cellularity. 

However, in our practice, ADC values do not always 

reflect only cellularity. The purpose of this exhibit was to 

show the spectrum of histologic features that influence ADC 

values of tumors and tumor-like lesions of the nasal cavity, 

the paranasal sinuses, and the adjacent skull base. 


A retrospective review of cases with histologically proven 

malignant tumors or benign tumor-like lesions in the nasal 

cavity, the paranasal sinuses, and the adjacent skull base was 

performed. The malignant tumors included rhabdomyosarcoma, 

squamous cell carcinoma, malignant melanoma, lymphoma, 

and neuroblastoma. The benign lesions were mesenchymal 

proliferative process, glomus tumor, polyp, mucocele, 

and neuroma and fibrosis. Diffusion-weighted MR 

imaging was performed with a 1.5 T MR unit in all cases. 

The b-values were1000 s/mm 2. Apparent diffusion coefficient 

maps were generated off-line on an ADW 4.0 GE 

workstation using FuncTool. Mean ADC values were compared 

with cellularity and other histologic feature derived 

from resected or biopsy material. 

FINDINGS 

The mean ADC values correlated well with cellularity. In 

some cases, however, ADC values also were affected greatly 

by other histologic features, such as necrosis, keratin, 

amount of stroma and cytoplasm, myxoid change, debris, 

and collagen. 

CONCLUSION 

Information provided through this exhibit will enable the 

reader to be familiar with the spectrum of histologic features 

that influence the ADC values of tumors and tumor-like 

lesions in the nasal cavity, the paranasal sinuses, and the 

adjacent skull base. 

KEY WORDS: Sinonasal tumor, histologic feature, apparent 

diffusion coefficient

415 



Imaging Features of Inverted Papillomas of the The Half-Smile: Intratemporal Causes of Facial Nerve 

Sinonasal Cavity 

Palsy 

Marin, H. · Jain, R. · Aho, T. · Patel, S. C. 


Detroit, MI 

PURPOSE 

Sinonasal inverted papillomas are rare tumors with local 

aggressive features and have tendency to recur after surgical 

resection. An association with squamous cell carcinoma has 

been described in up to 15% of the cases. An accurate preoperative 

staging is essential in surgical planning. The purpose 

of our study is to illustrate imaging features of inverted 

papillomas on CT and MR imaging. 


Contrast-enhanced CT and MR examinations were performed 

in patients with histologically proven inverted papillomas. 

Malignant transformation with squamous cell carcinoma 

was present in two cases. The imaging data was analyzed 

retrospectively using the following morphologic criteria: 

localization, tumor configuration, bony changes, and 

contrast-enhancement pattern on CT and signal intensity and 

tumor extension pattern on MR imaging. 

FINDINGS 

The tumor was seen arising from the lateral wall of the nasal 

cavity in most of the cases and also arising from the lateral 

wall of the maxillary sinus in one case. Bone abnormalities, 

better seen on CT were identified in all cases with primarily 

erosion of the lateral wall of the nasal cavity. Speckled calcifications 

within the mass were associated with extensive 

destruction of the sinonasal walls. Focal hyperostosis at the 

insertion site was identified when the tumor was arising from 

the lateral wall of the maxillary sinus. On MR imaging, the 

mass was isointense on T1- and slightly hyperintense to the 

muscles on T2-weighted images. T2-weighted images 

allowed precise delineation of the mass from retained secretions 

and inflammatory mucosal changes which was difficult 

with CT images. A convoluted cerebriform pattern on T2weighted, 

contrast-enhanced T1-weighted was identified in 

all cases. Aggressive features with invasion of the pterygopalatine 

fossa and the orbit as well as perineural tumor 

spread were demonstrated on MR images suggesting association 

with squamous cell carcinoma. 

CONCLUSION 

Contrast-enhancing sinonasal masses with a convoluted 

cerebriform appearance on MR imaging are suggestive for 

inverted papilloma. MR imaging allows a more precise characterization 

of the mass and better delineation of the tumor 

extent for preoperative planning. 

KEY WORDS: Inverted papilloma, sinonasal malignancy, 

schneiderian papilloma 

Goh, J. 1 · Chan, E. H. Y. 2 · Tang, P. 1 · Tan, T. 3 · Lee, H. 4 · 

Smoker, W. R. K. 4 · Gentry, L. R. 5 

1Tan Tock Seng Hospital, Singapore, SINGAPORE, 

2Singapore General Hospital, Singapore, SINGAPORE, 

3Changi General Hospital, Singapore, SINGAPORE, 

4University of Iowa Hospitals and Clinics, Iowa City, IA, 

5University of Wisconsin Medical School, Madison, WI 

PURPOSE 

To demonstrate the gamut of intratemporal causes of facial 

nerve palsy. 


The most common cause of facial nerve palsy is a benign 

inflammatory condition called Bell’s palsy. This usually has 

a viral etiology and is commonly self-limiting. Symptoms 

may sometimes be unremitting. However, there are many 

conditions which may injure the facial nerve producing 

unremitting facial nerve palsy. These may occur at any point 

along the course of the facial nerve. The severity of symptoms 

is affected also by the site of damage to the facial nerve 

itself. Oftentimes the modality of choice (i.e., CT or MR 

imaging) is dictated by the clinical presentation; on occasion, 

both modalities are required to accurately delineate a 

lesion. We performed a retrospective review of patients presenting 

with unremitting facial nerve palsy in four teaching 

hospitals in Singapore and the United States. All patients had 

undergone either MR imaging and/or CT. CT was performed 

with single or multislice scanners. Coronal and axial planes 

were acquired when a singleslice scanner was utilized. On 

the multislice scanner axial images were obtained at 0.5 mm 

or 0.75 mm intervals with coronal reformats. In this exhibit, 

we limit ourselves to the intratemporal causes of facial nerve 

palsy. We review the normal anatomy and imaging features 

of the facial nerve on CT and MR imaging. 

FINDINGS 

We present several conditions which may mimic an unremitting 

Bell’s palsy. These include schwannomas (facial and 

vestibular); ossifying and nonossifying hemangiomas of the 

facial nerve; inflammatory conditions such as otitis media; 

and tumors including chondrosarcoma of the temporal bone. 

We also illustrate the typical imaging features of Bell’s palsy, 

and demonstrate other causes of facial nerve palsy, including 

temporal bone trauma involving the geniculate ganglion. 

CONCLUSION 

Unremitting facial nerve palsy may be caused by other etiologies 

apart from Bell’s palsy. 

KEY WORDS: Facial, palsy, intratemporal 

Scientific Exhibits



Imagining Evaluation of Congenital Anomalies of the 

Temporal Bone with an Overview of the Embryologic 

Development 

Patel, P. 1 · Jain, R. 1 · Horia, M. 1 · Gandhi, D. 2 · Corrigan, J. 1 · 

Aho, T. 1 · Patel, S. C. 1 

1 2 Henry Ford Health System, Detroit, MI, University of 

Michigan, Ann Arbor, MI 

PURPOSE 

1. To review the embryologic development of the external, 

middle, and inner ear structures. 2. To systematically learn 

about the congenital anomalies of the external, middle, and 

inner ear based on the stages of the embryologic development. 


We will incorporate sketch diagrams with illustrative cases 

from our imaging database to review the embryologic development 

of the external, middle, and inner ear structures and 

to correlate various congenital anomalies with the stages of 

the embryonal development. High-resolution thin-section 

CT with multiplanar reformats and thin-section MR images 

will be used. 

FINDINGS 

The external and the middle ear structures develop from the 

first and second branchial arches, the first branchial cleft,and 

the first pharygeal pouch. The inner ear structures develop 

from the otic placode, an ectodermal thickening near the 

hindbrain on either side of a 3-week embryo. There is a 

defined order of development. Disruption in the normal 

development and the resultant congenital anomalies depend 

on the time of gestation that the insult occurred. The external 

and middle ear malformations tend to involve both the structures 

as their embryonal development is linked. The auricle 

is derived from the first and the second branchial arches. The 

first pharyngeal pouch forms the eustachain tube and middle 

ear cleft. The first branchial arch (Meckel’s cartilage) develops 

into the head of the malleus, body, and short process of 

the incus, and the tensor tympani muscle and tendon. The 

second branchial (Reichert’s cartilage) arch develops into the 

stapedius muscle and tendon, and the remainder of the ossicular 

chain, except for the footplate of the stapes. The inner 

ear structures start developing from the otic placode in third 

gestational week.. Major inner ear morphologic anomalies 

occur between the third and tenth fetal weeks. The otic placode 

develops into the otic vesicle that develops into the two 

pouches. By the sixth fetal week, the dorsal pouch develops 

into the utricle and semicircular canals and the ventral pouch 

develops into the saccule and cochlear duct. Cochlear duct 

starts elongating and has one turn by eighth week and by 

eleventh week has full two and a half turns. The semicircular 

canals begin to develop from the utricle and superior 

semicircular canal is complete by 19 weeks, while the lateral 

is the last to form at 22 weeks. 

CONCLUSION 

It is important to correlate imaging findings with embryologic 

development in order to better delinate and understand 

various complex congenital anomalies and their origin. 

Inner ear anomalies are rare but important cause of sen- 

416 

sorineural hearing loss. Combined external and middle ear 

anomalies are more common and cause conductive hearing 

loss. High-resolution CT with multiplanar reformats can better 

delineate the bony details of these very small structures, 

while MR imaging can demonstrate the membranous 

labyrinth. 

KEY WORDS: Temporal bone, congenital anomalies 


3 T MR Imaging of the Membranous Labyrinth 

Holtzman, A. W. · Eames, F. · Carfrae, M. · Parnes, S. 

Albany Medical Center 

Albany, NY 

PURPOSE 

While high-resolution MR imaging of the inner ear structures 

has been demonstrated in animal models, visualization 

in humans has posed a challenge. This exhibit will review 

the normal anatomy of the inner ear as it appears on 3 T MR 

imaging and the sensitivity of assessing subtle changes to the 

perilymph in normal pre and postcontrast subjects. The 

applications may serve as a basis for assessing patients with 

Meniere’s (endolymphatic hydrops) syndrome. 


Five normal subjects (3 male, 2 female) are imaged utilizing 

a phased-array coil with 3 T spin echo (SE) and gradient 

echo (GRE). Spin-echo T1 pre and postgadodiamide 

sequences, T2-weighted and 3D GRE MR imaging is used to 

visualize the semicircular canals, vestibulocochlear nerve 

and the membranous labyrinth divisions of the scala vestibuli 

and tympani. Postcontrast images are obtained hourly over 

a 5-hour period, and pixel intensity is measured at the basal 

turn of the cochlea. 

FINDINGS 

At the time of submission of this abstract, anatomical definition 

of the precontrast membranous labyrinth can be defined 

on both 2D and 3D acquisition. Based on conclusive animal 

models, we predict that a trend in postcontrast signal change 

within the perilymphatic space can be recorded and a maximum 

intensity period demonstrated. 

CONCLUSION 

The basis for differentiation of the perilymphatic and 

endolymphatic spaces depends on optimizing scan time in 

correlation with peak enhancement of the perilymph. 

Findings eventually may serve to better characterize different 

stages of endolymphatic hydrops and benefit intractable 

cases in diagnosis, staging, and surgical planning. 

REFERENCES 

1. Niyazov DM, Andrews JC, Strelioff D, Sinha S, Lufkin R. 

Diagnosis of endolymphatic hydrops in vivo with MRI. Otol 

& Neurotol 2001;22:813-817 

KEY WORDS: Membranous labyrinth, perilymph, Meniere’s


71-73 


Imaging of Carotid Disease Treated with Carotid Stents: 

Pre and Postprocedural Assessment 

Johnson, M. H. · Lin, F. · Scoutt, L. · Kelley, E. L. 


New Haven, CT 

PURPOSE 

To determine the most specific, sensitive, and cost-effective 

imaging strategies for the evaluation of carotid disease pre 

and posttreatment with carotid stents. 


Retrospective review of imaging practice patterns was performed, 

by a team of physicians, on patients treated with 

carotid stents. Patients were treated by a variety of practitioners, 

including interventional radiology, interventional 

neuroradiology, neurosurgery, vascular surgery, and cardiology. 

Pre and postprocedural imaging, and catheter angiography 

(DSA) were reviewed and assessed for diagnostic accuracy 

and utility for patient management. 

FINDINGS 

All patients with carotid atherosclerotic disease as an indication 

for carotid stent placement had preprocedural ultrasound and 

color Doppler evaluation. Patients with trauma and malignancy 

as an indication for carotid stent placement were assessed most 

commonly only by CT or CTA prior to catheter angiography 

and were imaged uncommonly by MR imaging or MRA. CT 

angiography represented the second and MR imaging/MRA the 

third most common imaging prior to DSA for all patients undergoing 

treatment with carotid stents. Diagnostic angiogram in the 

majority of cases was performed at the time of carotid stent 

placement. In a few complex cases the diagnostic DSA was performed 

several days to weeks prior to stent placement. Followup 

imaging included ultrasound in 100% of patients and either 

MR imaging/MRA or CT/CTA, usually within 4-6 weeks. Over 

the last 5 years, CTA has become more important than MR 

imaging/MRA in our management of patients with carotid disease 

requiring stent placement. CT angiography permits quantitative 

assessment of arterial diameter, while MRA is often limited 

due to metallic artifact. MR imaging is best for identification 

of early cerebral ischemic changes. 

CONCLUSION 

Ultrasound is the main imaging modality for carotid disease of 

all etiologies. For the atherosclerotic patient, brain imaging (MR 

imaging > CT) prior to therapy is useful, particularly in the clinically 

asymptomatic patient. Early ultrasound with follow-up 

CT/CTA in 4-6 weeks is accurate and cost effective in the postprocedural 

evaluation of the asymptomatic patient and those 

with traumatic injury or malignancy. Patients with complicated 

course or symptoms of TIA or stroke require more aggressive, 

early imaging often with ultrasound and MR imaging/MRA. 

KEY WORDS: Carotid stents, noninvasive imaging, pre and 

postprocedure assessment 

417 


Use of the BALT Self-Expanding Intracranial LEO Stent 

in Clinical Neurointerventional Practice 

White, P. M. · Sellar, R. J. 

Western General Hospital 

Edinburgh, UNITED KINGDOM 

The Leo stent from BALT is a self-expandable stent made up 

of woven braided nitinol wires. It is resheathable and repositionable 

until > 90% of the stent length has been deployed. 

The Leo stent exerts considerably greater radial force than 

the Neuroform stent and provides a very stable support platform 

to coiling. The hemodynamic effect of the Leo stent 

also is substantially greater due to increased mesh density 

arising from its woven design. A unique feature is the 2 helical 

platinum wires incorporated into the stent structure. 

These enable the entire length of the stent to be visualized 

and allow easy assessment of whether the stent is fully 

expanded along its entire length. Sizes up to and including 

3.5 mm diameter are delivered via a braided 2.6F Vasco 

microcatheter, 4.5-5.5 mm diameters via a 3.6F Vasco. A 

wide range of stent diameters (2.5-7.5 mm) and lengths (12- 

100 mm) are available. The Vasco tracks well but in larger 

sizes 2 guidewires can be required to negotiate around tortuous 

anatomy. Clinical cases demonstrating the clinical utility 

of the Leo stent are presented, which illustrate some of the 

advantages of this resheathable stent. Tips and tricks with 

regard to its use and also the potential problems as well as 

the benefit of increased haemodynamic effect are discussed. 

Standard Leo Stent Range 

Stent size in mm Vessel Length of External Vasco Guide 

[also available diameter stent when diameter size cath 

in length] accommodated deployed of delivery diameter 

system required 

2.5x12 [18] 2.3-3.0 11-13 2.6 F 21 .021” 

3.5x18 [12,25] 3.0-4.0 16-19 2.6 F 21 .021” 

4.5x25 [20,30,40,50] 4.0-5.0 23-26 3.6 F 28 .028” 

5.5x30 [25,35,50,75] 5.0-6.0 28-32 3.6 F 28 .028” 

6.5x40 [30,35,60,80] 6.0-7.0 38-42 4.2 F 35 .035” 

7.5x50 [40-100] 7.0-8.0 48-52 4.2 F 35 .035” 

KEY WORDS: Stent, self-expanding, intracranial 


Postmortem Neuroangiography: Improving the 

Evaluation of Neurovasculature in Forensic Pathology 

Plosker, A. D. · Kemp, W. L. · Perez, C. L. · Barnard, J. J. · 

McAnulty, A. L. · Chason, D. P. 

University of Texas Southwestern Medical Center 

Dallas, TX 

PURPOSE 

The neurovasculature structures are difficult to evaluate with 

post-mortem dissection alone. The purpose of this exhibit is 

to demonstrate how employing neuroangiographic techniques 

can improve the post-mortem evaluation of neurovasculature 

at the time of autopsy. 

Scientific Exhibits



Decedents undergoing autopsy at the Dallas County medical 

examiner’s office with gunshot wounds and blunt force 

injuries to the head and neck, as well as nontraumatic head 

and neck disease processes were selected for post-mortem 

neuroangiography. A diagnostic radiography unit with plain 

film radiographs, expired contrast agents, and catheter 

equipment were used to evaluate the neurovasculature. Postmortem 

neuroangiography was accomplished via exposure 

and direct puncture either of the aorta, carotid and/or, the 

vertebral arteries. 

FINDINGS 

Neuroradiologic techniques can aid forensic pathology by 

optimizing contrast injections, proper angiographic positions, 

as well as image generation and interpretation. In addition, 

post-mortem neuroangiography better identifies neurovascular 

pathology than dissection alone. 

CONCLUSION 

Post-mortem neuroangiographic techniques improve forensic 

evaluation of the neurovasculature. These procedures 

provide additional diagnostic information, and can be incorporated 

easily and inexpensively into routine autopsies. In 

addition, technical improvements, including c-arm equipment 

and digital radiography, will further advance the use of 

neuroangiography in forensic pathology. 

KEY WORDS: Neuroangiography, post-mortem 

Pediatrics 

75-81 


Conceptual Approach to Cerebral Features of Chiari II 

Malformation on MR Imaging 

Sabir, S. · Alorainy, I. 

King Khalid University Hospital 

Riyadh, SAUDI ARABIA 

PURPOSE 

To review the embryologic basis and consequent morphologic 

changes of Chiari II malformation and to emphasize 

the important midline sagittal findings that illustrate the concept 

behind posterior fossa changes. 


The approach to interpretation of cerebral features of Chiari 

II malformation on MR imaging relies on the concept that in 

this complex malformation of the hindbrain, spine, and 

mesoderm of the skull the normal-sized cerebellum develops 

in abnormally small posterior fossa. Therefore, as the brain 

grows there is upward, downward, and forward migration of 

posterior fossa structures and scalloping of its bony walls. 

This scientific exhibit illustrates the concept using MR 

images in multiple planes. 

418 

FINDINGS 

Upwards cerebellar towering through incisura with dens-like 

appearance. Downwards cerebellar herniation through 

enlarged foramen magnum with cerebellar peg formation. 

Forwards cerebellar creeping around brain stem structures. 

Medulla and upper cervical cord descend, with characteristic 

medullary spur and cervico-medullary kink. Fourth ventricle 

elongates, narrows and descends down. Tectal beaking is due 

to fusion of colliculi and upwards herniation of cerebellum 

through tentorial hiatus. Vertical orientation and low insertion 

of tentorium. The foramen magnum enlarges and scalloping 

of the posterior surface of petrous bone and clivus is 

seen because of growing cerebellum, which tries to fit itself 

into a small posterior fossa. Supratentorial associations 

include dysgenesis of corpus callosum, colpocephaly, absent 

septum pellucidum, parasagittal cyst, stenogyria, polymicrogyri, 

migration disorders and schizencephaly. 

CONCLUSION 

Keen observer will carry a visual impression of the posterior 

fossa configuration that would help him in easy recognition 

of Chirari II malformation, whenever encountered. 

KEY WORDS: Chiari II, malformation, herniation 


Fetal MR Imaging in the Evaluation of Central Nervous 

System and Spinal Anomalies 

Martín, C. 1 · Rovira-Gols, A. 1 · Escofet, C. 2 · Darnell, A. 1 · 

Durán, C. 1 · Carvajal, A. 1 · Zauner, M. 1 

1UDIAT-CD Corporació Sanitària Parc Taulí, Sabadell, 

SPAIN, 2Neuropediatria. Corporació Sanitària Parc Taulí, 


PURPOSE 

To describe the MR appearance of fetal central nervous system 

(CNS) and spinal anomalies and evaluate the influence 

of this technique in obstetric and neonatal management. 


From 1997 to 2004, 359 pregnant women underwent MR 

imaging at our center for evaluation of fetal anomalies 

detected or suspected at ultrasound (US). MR examinations 

were performed with 1.0 T-1.5 T MR units using T2W 

HASTE sequences and T1-weighted FLASH sequences. 

Images were analyzed by an expert pediatric radiologist and 

findings were compared and correlated with clinical, US, 

and postnatal imaging findings or autopsy. 

FINDINGS 

One hundred fourteen pregnant women with 119 fetuses presented 

115 anomalies of the CNS and spine: Anomalies of 

the corpus callosum (n = 13), diencephalic cyst (n = 2), holoprosencephaly 

(n = 5), septo-optic dysplasia (n = 1), cerebellar 

hypoplasia (n = 6), cerebellar cortical dysgenesis (n = 

2), rhombencephalosynapsis (n = 1), absence of the cerebellar 

vermis (n = 1), Chiari malformation (n = 5), Dandy- 

Walker complex (n = 9), cerebellar hemorrhage (n = 1), 

megacephaly (n = 3), dolichocephaly (n = 1), schizencephaly 

(n = 1), cerebral hemiatrophy (n = 1), hydranencephaly (n = 

1), hydrocephalus (n = 8), ventriculomegaly (n = 37), arach-

noid cyst (n = 2), choroidal plexus cyst (n = 3), anomalies of 

the spine (n = 12). Findings were normal in 20 fetuses with 

suspected brain abnormalities or neural tube defects at US. A 

single anomaly was found in 58 fetuses. Two or more CNS 

anomalies were found in 41 fetuses. Central nervous system 

and neural tube anomalies were associated with anomalies in 

other organs in 26 cases. Pregnancy was voluntarily interrupted 

in 31 patients and two fetuses died intrauterine. MR 

analysis provided additional diagnostic information in 41 

cases, changing the diagnosis in 36 and changing the therapeutic 

approach in 19. 

CONCLUSION 

Central nervous system and spinal anomalies are common. 

Prenatal diagnosis of these anomalies can determine the viability 

of the fetus, orient the prognosis and management of 

the newborn, and has implications in genetic counseling. MR 

imaging is a noninvasive technique for studying the fetal 

CNS and spine that provides additional information after US 

evaluation. 

KEY WORDS: Fetal MR imaging, CNS anomalies, spinal 

anomalies 


Brainstem and Cerebellar Findings in Joubert Syndrome 

Alorainy, I. A. 1 · Sabir, S. 2 

1 College of Medicine, King Saud University, Riyadh, 

SAUDI ARABIA, 2 King Khalid University Hospital, 


PURPOSE 

Joubert syndrome often is missed clinically and radiologically 

if not enough attention is paid to its subtle and variable 

presentation and brainstem and cerebellar malformations on 

imaging. The purpose of this exhibit is to illustrate the brainstem 

and cerebellar findings in Joubert syndrome on CT and 

MR imaging in view of its clinical presentation. 


The brainstem and cerebellar changes in children with clinically 

proven Joubert syndrome were illustrated as seen on 

CT and MR imaging. The components of the classical 

“molar tooth sign” and their explanation were also discussed. 

FINDINGS 

In the majority of patients with Joubert syndrome the superior 

cerebellar peduncles are thick and perpendicular to the 

brainstem and they lack decussation in the midbrain. The 

fastigium of the fourth ventricle is rostrally deviated causing 

abnormal rectangular shape of the fourth ventricle. Because 

of lack of decussation of the superior cerebellar peduncles in 

the midbrain, the ponto-mesencephalic junction (isthmus) 

becomes shallow and the interpeduncular cistern deep. The 

constellation of abnormal superior cerebellar peduncles, 

shallow ponto-mesencephalic junction, and deep interpeduncular 

cistern gives the classical “molar tooth sign” seen on 

the axial plane.The cerebellar vermis is either completely 

absent or hypogenetic, especially the inferior vermis, which 

is the reason for the “batwing” appearance of the fourth ven- 

419 

tricle. Enlargement of the prepontine cistern and presence of 

a retrocerebellar cyst are seen occasionally in Joubert syndrome. 

CONCLUSION 

Awareness about the clinical and neuroimaging findings in 

Joubert syndrome and the high index of suspicion are very 

essential for correct diagnosis of this rare congenital malformation. 

KEY WORDS: Joubert, brain malformation, posterior fossa 


Range of Preoperative Imaging Findings in Pediatric 

Patients with Intractible Epilepsy 

Faris, G. W. · Bardo, D. M. E. · Frim, D. M. · Marcuccilli, C. 

J.· Hecox, K. E. 

University of Chicago 

Chicago, IL 

PURPOSE 

The purpose of this exhibit is to describe, illustrate, and correlate 

clinical, EEG, imaging, surgical, and pathologic data 

used to reliably identify seizure activity and pertinent neuroanatomy 

in a group of pediatric patients that have undergone 

surgical resection of seizure foci. 


Twenty-five patients, 11(44%) female and 14 (56%) male, 

average age of 12.24 years, with intractable epilepsy, underwent 

clinical, functional, and imaging evaluation, surgical 

resection and pathologic assessment of their seizure focus at 

our institution between January 2002 and July 2004. All 

patients had preoperative 24-hour video-monitored scalp 

EEG recordings. Subdural EEG grids were placed in 24 

(96%). Preoperative CT was performed in 5 (20%); all 

patients with subdural EEG had CT following that procedure 

to evaluate grid position. Preoperative MR imaging was performed 

at our institution in 21 (84%); 4 (16%) had MR 

imaging at other institutions. WADA testing 8 (32%) and 

PET scans 11 (44%) were performed also. Resection of the 

epileptic focus and histologic evaluation occurred in all 

patients. 

FINDINGS 

All results are in comparison to histologic assessment which 

represents true positive findings. Scalp EEG showed only 

50% (10/20) accuracy. A variety of seizure types were classified 

using clinical and video-monitored EEG findings: 10 

(40%) combination of complex partial and general tonic 

clonic; 9 (36%) complex partial; 2 (8%) presented in status 

epilepticus; 3 (12%) mixed seizures; and 1 (4%) general 

tonic clonic. Preoperative MR imaging diagnoses correlated 

with the seizure focus in 14 of 20 (70%). MR imaging findings 

included 2 (8%) tumor; 5 (20%) previous infection or 

inflammation; 2 (8%) hemimegalencephaly; 4 (16%) remote 

trauma; 3 (12%) cortical dysplasia; 1 (4%) tuberous sclerosis; 

1 (4%) parietal hemorrhage; 1 (4%) cortical atrophy with 

porenchephalic cyst. Subdural EEG data accurately localized 

seizure foci in 19 (95%) in patients with MR imaging at our 

institution. CT in patients with subdural EEG grids was most 

Scientific Exhibits


useful to evaluate grid location and postprocedural changes. 

Preoperative CT identified the surgical lesion in 2 of 5( 40%) 

patients examined. WADA testing successfully localized 

speech centers in 100% of 8 patients examined. PET scan 

accurately localized seizure focus to the location of MR 

imaging and surgical abnormality in 100% of 11 patients. 

Resection of the epileptic focus and histologic evaluation 

was performed in all patients. Three (12%) patients had complete 

lobectomy; 21 (84%) patients had partial lobectomy; 1 

(4%) had hemispherectomy. Histologic diagnoses included 8 

(32%) cortical dysplasia, normal 5 (20%), mesial temporal 

sclerosis 3 (12%), cortical atrophy 3 (12%), 2 (8%) perivascular 

inflammation, 2 (8%) ganglioglioma and 2 (8%) gliosis. 

CONCLUSION 

Combinations of clinical, functional and radiographic diagnoses 

are essential to accurately detect seizure foci in 

patients with intractable epilepsy. Subdural EEG testing was 

most accurate in localization of seizure focus matching the 

surgical excision site in 95%. Scalp EEG was 50% accurate. 

MR imaging revealed anatomical abnormality in 14 (70%) 

of 20 patients examined at our institution. PET scan data correlated 

with MR imaging and pathologic abnormality with 

100% accuracy. Only 2 (40%) of patients with preoperative 

CT had accurate identification of the lesion. Cortical dysplasia 

was the most common pathologic diagnosis. Other diagnoses 

included normal 5 (20%), mesial temporal sclerosis 5 

(20%), cortical atrophy 5 (20%), perivascular inflammation 

2 (8%), ganglioglioma 2 (8%), fibrillary gliosis 1 (4%) and 

scar attributable to prior surgery in 1 (4%). 

KEY WORDS: Epilepsy, preoperative imaging, pediatrics 


Reversible MR Signal Changes in the Splenium of the 

Corpus Callosum in Various Central Nervous System 

Disorders: Estimated by Diffusion-Weighted MR 

Imaging and MR Spectroscopy 

Kawamura, Y. 1 · Watanabe, K. 2 · Itoh, H. 1 

1 University of Fukui, Fukui, JAPAN, 2 Okazaki Municipal 

Hospital, Okazaki, JAPAN 

In this study, reversible signal abnormality in the splenium 

of the corpus callosum was revealed on T2-weighted images, 

FLAIR images and EPI-diffusion images in various central 

nervous disorders (Table). Although the exact mechanism of 

the lesion in the splenium is unclear, its reversibility indicates 

that ischemic damage and cytotoxic edema is not likely 

the cause. The marked high intensity abnormality seen on 

diffusion-weighted images was reduced when motion probing 

gradient (MPG) was applied in X direction parallel to the 

myelin fibers course in the splenium. On T2-weighted 

images and FLAIR images, high intensity change is not as 

remarkable as on diffusion-weighted images. MR spectroscopy 

obtained in three patients (two infectious 

encephalopathy cases and one epileptic case) did not show 

significant decrease in NAA and no significant lactate peak. 

The authors presume that the myelin edema is the main 

cause of this transient signal abnormality in the splenium of 

the corpus callosum in various central nervous disorders. 

420 

Patients’ Details 

Case Age/ Symptoms Pathology 

Number sex 

1 8/female fever, convulsion, diarrhea, hallucination Salmonella Enteritidis 

2 2/female fever, convulsion, diarrhea, delirium Rotavirus 


4 5/male fever, convulsion, disorientation Influenza A 

5 5/male fever, disorientation Adenovirus 

6 4/male fever, convulsion, status epilepticus Influenza A 




10 15/male irritation Anticonvulsants for 


11 21/female no symptom Anticonvulsants for 


12 25/female lethargy Anticonvulsants for 

schizophrenia 

13 20/female headache Shering injury (traffic accident) 

14 18/female somnolence Osmolarity disorder due to 

Anolexia Nervosa 

KEY WORDS: MR imaging, diffusion, MR spectroscopy 


Diffusion-Weighted Imaging of Excitotoxic Brain Injury 

in Shaken Baby Syndrome 

Moritani, T. · Sato, Y. · Smoker, W. R. K. · Oral, R. 


Iowa City, IA 

PURPOSE 

1. Illustrate various diffusion-weighted imaging patterns of 

brain injury in shaken baby syndrome. 2. Understand the 

pathophysiology of brain injury in shaken baby syndrome 

based on excitotoxic machanisms. 


The pathogenesis of brain injury in shaken baby syndrome is 

multifaceted and remains unclear. Young infants have relatively 

large heads, weak neck muscles and thin skulls, predisposing 

to the shearing brain injuries. Unmyelinated white 

matter is more vulnerable to shearing stress than myelinated 

one. The pediatric brain is vulnerable to excitotoxic brain 

injury due to the different distribution of glutamate receptors 

at various ages. Cerebrospinal fluid glutamate levels are 

known to be extremely high in patients with shaken baby 

syndrome. Excitotoxic brain injury is related to increased 

release or leakage of glutamate associated with traumatic 

stimuli, and decreased re-uptake of glutamate with energy 

failure, which causes cytotoxic edema, resulting in decreased 

ADC values. In this exhibit, we demonstrate the various distributions 

of cytotoxic edema on diffusion-weighted imaging 

from our collection of more than 30 cases.of shaken baby 

syndrome. 

FINDINGS 

The patterns of brain parenchymal injury can be divided into 

focal, multifocal, or diffuse. The distribution of cytotoxic 

edema was not necessarily associated with a subdural 

hematoma or intraparenchymal hemorrhage. In fact, widespread 

parenchymal injury was often not related to the location 

and size of an acute subdural hematoma. The cause of 

brain injury in shaken baby syndrome could be related to 

traumatic stress such as shearing stress, hypoxia/ischemia, 

neuronal seizure activity, or a combination of these mechanisms. 

The severity of the diffusion-weighted imaging

abnormality correlated with patient outcome. Glutamate 

receptor antagonists may offer attractive possibilities for 

future therapy as neuroprotectants. 

CONCLUSION 

We demonstrate diffusion-weighted imaging findings of 

brain parenchymal injury in shaken baby syndrome and discuss 

the pathophysiology based on excitotoxic mechanisms. 

KEY WORDS: Shaken baby, diffusion-weighted imaging, 

MR imaging 


MR Imaging Findings and Clinical Presentation in 

Pediatric Nonaccidental Trauma 

Foerster, B. · Petrou, M. · Eldevik, P. · Maly, P. · Sundgren, 

P. C. 


Ann Arbor, MI 

PURPOSE 

To illustrate the MR imaging findings and describe the clinical 

presentation in suspected nonaccidental trauma in children 

under 2 years of age. 


We reviewed the hospital charts and the MR imaging reports 

of 15 patients examined due to clinical suspicion of nonaccidental 

trauma. We noted the clinical presentation and MR 

findings in each case. Both the radiologic diagnoses and the 

final diagnoses were compared and the impact of the MR 

examination on the clinical course was evaluated. 

FINDINGS 

Subdural hematoma, ranging from acute to chronic, was the 

most common finding seen in 13 of the 15 patients. Petechial 

hemorrhages, brain edema, and axonal injury were other 

findings in this population. These examples will be illustrated 

with a discussion of the mechanisms of injury. 

CONCLUSION 

Nonaccidental brain trauma is not uncommon and in many 

cases, the radiologist is an integral part of the diagnostic 

work up of these patients. Radiologists must maintain a high 

level of suspicion in this population and be familiar with the 

radiologic findings of nonaccidental trauma. MR imaging is 

performed in a limited number of cases; for example, when 

dating the abnormalities is crucial and to evaluate for additional 

findings such as axonal shear injury. 

KEY WORDS: Nonaccidental trauma, pediatrics, MR imaging 

421 

Socioeconomic 

82 


Spurious Hypocalcemia after Administration of 

Gadolinium Contrast Agents: An Algorithm for 

Minimizing False-Positive Laboratory Values 

Emerson, J. 1 · Kost, G. J. 2 

1 University of California Irvine, Orange, CA, 2 University of 

California Davis, Davis, CA 

PURPOSE 

Several gadolinium-based MR contrast agents have demonstrated 

varying degrees of interference with colorimetric laboratory 

methods used to measure total calcium. This can 

result in unnecessary and potentially dangerous calcium supplementation 

in patients with normal serum calcium levels. 

Alternatively, a clinically significant elevated calcium level 

may be overlooked due to the artificially lowered value. We 

undertook a review of the available literature to determine 

the magnitude of this problem and to identify an algorithm 

for minimizing laboratory interference after MR contrast 

agent administration. 


We reviewed the literature from 1995 to the present on laboratory 

interferences by MR contrast agents, including gadodiamide 

(Omniscan), gadopentetate dimeglumine 

(Magnevist), gadoversetamide (Optimark), and gadoteridol 

(ProHance). Six clinical reports and two experimental studies 

were identified and reviewed. 

FINDINGS 

The interference by gadodiamide in the laboratory determination 

of serum calcium was first noted in 1995. Since then 

seven other studies have examined this problem and have 

shown that another agent, gadoversetamide, also may produce 

laboratory interference with the measurement of total 

calcium. Two gadolinium-based contrast agents, gadoteridol 

and gadopentetate dimeglumine, do not interfere with total 

calcium measurement, although gadopentetate dimeglumine 

has been shown to interfere with the assessment of other 

metals or metabolites. The mechanism of interference is 

related to the propensity of gadolinium agents to dechelate in 

the body. Several linear, open-chain chelates, such as those 

used in gadodiamide and gadoversetamide, are less stable 

and more readily undergo dechelation. The most commonly 

used laboratory techniques rely on the use of reagents such 

as OCP than complex calcium and other metals. It is thought 

that these reagents bind with free gadolinium, making the 

reagent less able to bind calcium. Additionally, excess 

chelate added to the formulation of some MR imaging contrast 

agents may bind with serum calcium, making it unavailable 

to bind to the OCR reagent. Prince and colleagues 

demonstrated that laboratory interference after gadodiamide 

was more likely to occur after high doses of contrast were 

given, when phlebotomy was performed shortly after the 

contrast injection, and when contrast was given to patients 

with renal impairment in whom clearance of the agent was 

delayed. 

Scientific Exhibits


CONCLUSION 

The use of commonly used laboratory tests soon after contrast-enhanced 

MR imaging with gadodiamide or gadoversetamide 

may result in a spurious lowering of the total calcium 

level consistent with apparent hypocalcemia. This phenomenon 

may be seen more often in patients with renal insufficiency 

and in those receiving higher doses of contrast agent. 

The available evidence suggests that potential interference is 

not limited to calcium determinations, and that it is related to 

the chemical stability of the gadolinium chelate complex. 

Laboratory interference can be eliminated by using a 

gadolinium agent that does not interefere with lab tests, such 

as gadoteridol, or can be minimized by obtaining blood samples 

before administration of a contrast agent or by allowing 

sufficient time for clearance of the contrast agent before 

obtaining blood samples for analysis. 

KEY WORDS: Hypocalcemia, gadolinium, in vitro diagnostics 

Spine 

83-91 


Skeletal Manifestation of Neurologic Disease 

Alorainy, I. · Al-Nakshabandi, N. · A. Boukai, A. 

King Saud University 


PURPOSE 

To illustrate the role of imaging and imaging findings in neurologic 

diseases presenting with skeletal symptoms. 


Several neurologic conditions including congenital, neoplastic, 

neurocutaneous, infective, and traumatic neurologic diseases 

can present with skeletal manifestations. Careful attention 

to the skeletal involvement in such conditions is a very 

essential step in the radiologic evaluation of these diseases. 

FINDINGS 

Klippel-Feil anomaly and fused vertebrae are seen in 

patients with Chiari 1 malformation. An enlarged foramen 

magnum in patients with Chiari 2 is observed. A midline 

mandible cleft is seen in holoprocencephaly, and a bone bar 

is seen in patients with diastematomyelia. Hyperostosis of 

the skull is seen sometimes in patients with meningiomas. 

Sphenoid wing hypoplasia, bowing of the extremities, vertebral 

scalloping, ribbon ribs, as well as dural ectasia are the 

hall marks of neurofibromatosis 1. Bone islands are seen in 

patients with tuberous sclerosis. Muscle calcification can be 

seen in patients with neurocystosercosis. Charcot joint in the 

knee is a manifestation of neurosyphilis. Charcot joint in the 

shoulder can be seen in patients with traumatic syrinx of the 

spinal cord. Children who sustain head trauma to the skull 

are prone to develop a growing fracture. 

422 

CONCLUSION 

Imaging has a very important role in detailed evaluation of 

the skeletal manifestations of several neurologic diseases. 

KEY WORDS: Neurologic diseases, skeletal symptoms 

Scientific Exhibits 84 

Imaging of Neurologic Manifestations of Skeletal 

Diseases 

Alorainy, I. A. · Al-Nakshabandi, N. 

College of Medicine, King Saud University 


PURPOSE 

To illustrate the role of imaging and the imaging findings in 

skeletal diseases presenting with neurologic symptoms. 


Several skeletal conditions including congenital, neoplastic, 

and infective/inflammatory diseases can present with neurologic 

manifestations, mainly due to neuronal compression by 

the osseous lesion or the associated soft-tissue component. 

Careful attention to the neuronal involvement in such conditions 

is a very essential step in the radiologic evaluation of 

these diseases. 

FINDINGS 

1) Spinal canal stenosis at the craniocervical junction and in 

the lumbosacral region with associated compression of the 

spinal cord and nerve roots that is best evaluated by MR 

imaging is the hallmark of achondroplasia. Fibrous dysplasia 

and osteopetrosis cause bone expansion and compression of 

the adjacent neuronal structures, more commonly the cranial 

nerves. Fractures affecting the spine in osteogenesis imperfecta 

may cause cord or nerve root compression. 

Osteogenesis imperfecta may be associated with otosclerosis 

resulting in mixed conductive and sensorineural hearing 

loss. 2) Primary bone neoplasms such as aneurysmal bone 

cyst (ABC), osteoblastoma, and osteochondroma have 

predilection to the posterior element of the spine. Fluid/fluid 

levels are seen classically on MR imaging in cases of ABC. 

3) Vertebral metastasis is common in elderly patients and 

frequently associated with pathologic fracture or soft tissue 

component causing neuronal compression. 4) Diskitis and 

vertebral osteomyelitis may result from pyogenic, tuberculous, 

or brucella infection with associated inflammatory softtissue 

component or abscess causing neuronal compression. 

Rheumatoid arthritis, histiocytosis, and amyloidosis are 

inflammatory conditions with soft-tissue component that 

causes neural compression in the skull and spine. 

CONCLUSION 

Imaging has a very important role in detailed evaluation of 

the neurologic manifestations of several skeletal diseases. 

KEY WORDS: Spine neoplasms, skull base neoplasms


Expanding Role of the Neuroradiologist in Patients 

Undergoing Intensity-Modulated Radiotherapy for 

Spinal Tumors 

Lis, E. · Krol, G. S. · Yamada, J. · Bilsky, M. H. · Boland, P. 

· Cooperman, S. 


New York, NY 

PURPOSE 

To emphasize the expanding role of neuroradiologist in precise 

definition of tumor extent and location of normal intracanal 

structures in patients undergoing intensity-modulated 

radiotherapy (IMRT) for spinal tumors. 


Intensity-modulated radiotherapy is a technique which 

allows for delivery of high-dose irradiation with greater 

effect of sparing of normal tissues comparing to conventional 

radiotherapy. In the treatment of spinal tumors it is necessary 

not only to outline the lesion border but also to identify 

the exact position of the spinal cord and neuroradiology 

input is critical in achieving this goal. First, minute gold 

markers are implanted in subcortical bone of vertebrae (usually 

posterior elements) to serve as fiducial markings. 

Second, a myelography is performed shortly before the simulation, 

outlining exact position of the spinal cord within the 

canal and its reference to epidural component of the tumor. 

FINDINGS 

This scientific exhibit will review the technique for the percutaneous 

CT-guided placement of gold fiducial markers 

into the posterior elements of vertebrae, myelography and 

the process of integration of the information they provide 

into the dose distribution calculations. 

CONCLUSION 

It is likely that IMRT will replace conventional radiotherapy 

and will become a method of choice in the treatment of 

spinal tumors susceptible to irradiation. Neuroradiology is in 

a perfect position to provide information needed for more 

precise treatment planning, likely resulting in more favorable 

outcome. 

KEY WORDS: Radiation therapy, spine neoplasm 


Occipital Condyle: Normal Anatomy, Congenital 

Anomalies, and Pathology 

Lee, H. K. · Smoker, W. R. K. · Kanekar, S. G. S. 

University of Iowa Hospital and Clinics 

Iowa City, IA 

PURPOSE 

Radiologic evaluation of the occipital condyle (OC) traditionally 

has been difficult and often incomplete. Therefore, a 

number of OC lesions may go undetected or be poorly 

understood. The purpose of this exhibit is to review the normal 

radiographic anatomy of the OC and demonstrate vari- 

423 

ous OC pathologies. Thin section CT and high-resolution 

MR studies with 2D or 3D reformation permit optimal 

assessment of the OC. 


We demonstrate the OC on multislice CT and high-resolution 

MR studies obtained on normal subjects. Additionally, 

multiplanar reconstruction (MPR) and 3D shaded surface 

rendering produced on a Vitrea workstation are included. We 

present a variety of OC abnormalities evaluated by CT/MR 

imaging identified from a retrospective review of our case 

logs. 

FINDINGS 

1. Crucial normal anatomy of the OC with ligamentous 

insertions is demonstrated. The oblique sagittal reconstruction 

along the long axis of the OC is optimal for assessment. 

2. Abnormalities are categorized broadly as follows: 1) congenital 

abnormalities - hypoplasia (unilateral or bilateral), 

assimilation and pneumatization; 2) trauma - three types of 

OC fractures and atlanto-occipital joint distraction/dislocation; 

3) neoplasm - hematogenous or direct metastases, multiple 

myeloma, chordoma, paraganglioma, etc; 4) infection/inflammation 

- tuberculosis, rheumatoid arthritis, 

pseudogout etc; 5) miscellaneous conditions - Paget’s disease, 

myelofibrosis, etc. 

CONCLUSION 

High-resolution CT and MR images with 3D rendering and 

MPR permit superb depiction of the OC. Knowledge of OC 

normal anatomy and congenital anomalies significantly aids 

in detection and evaluation of OC pathology. 

KEY WORDS: Occipital condyle, cranio-vertebral junction, 

occipito-atlantal joint 


Compendium of Extra-Spinal Disease Detected on 

Conventional CT and MR Imaging of the Spine 

Evans, J. M. · Gilbert, S. C. · Ford, K. L. · Opatowsky, M. J. 

Baylor University Medical Center 

Dallas, TX 

PURPOSE 

To refamiliarize spinal imagers with potentially significant 

nonneuroradiologic findings visible within the chest, 

abdomen, and pelvis on state of the art CT and MR spinal 

examinations, many of which now utilize wide field of view 

acquisitions. In some cases, the identification of pathology 

situated beyond, but in near proximity to the spinal column, 

may offer key insights into the symptomatic presentation of 

these patients. 


Twenty-five illustrative cases of CT and MR imaging of the 

cervical, thoracic, and lumbar spine containing noteworthy 

nonspinal lesions are presented. These representative cases 

were selected prospectively from all studies of the spine at 

our large urban hospital setting during an 18-month period. 

All studies were performed on state of the art multidetector 

CT and 1.5 T MR scanners. 

Scientific Exhibits


FINDINGS 

Our pictorial review includes numerous interesting cases 

showing noteworthy extra-spinal lesions of the chest, 

abdomen, and pelvis. Among the cases are previously unsuspected 

bronchogenic carcinoma, lobar lung collapse, renal 

malignancy, lesions of the adrenal gland, abdominal aortic 

aneurysm, leaking abdominal aortic graft repair with erosion 

into vertebral bodies, and small bowel obstruction. 

Twenty-five Illustrative Cases of Nonneuroradiologic 

Pathology on CT and MR Imaging of the Spine 

CT MR Imaging Total 

Cervical Spine 4 2 6 

Thoracic Spine 6 4 10 

Lumbar Spine 6 3 9 

Total 16 9 25 

CONCLUSION 

Frequently, spinal CT and MR examinations demonstrate 

pathology lying beyond the typical confines of the osseous 

spinal axis. Heightened awareness and greater familiarity 

with extra-spinal lesions of the lung, mediastinum, abdominal 

solid and hollow organs, retroperitoneum, vascular system, 

and pelvis has become an additional responsibility of 

the spinal imager, particularly given the volume of peripheral 

structures on these examinations. As spinal imagers, we 

should develop search patterns that encompass the structures 

surrounding the spine as a means of detecting significant 

unsuspected abnormalities. In addition, we may offer an 

opinion as to the nature of further definitive diagnostic evaluation. 

KEY WORDS: Spine, extra-spinal, pathology 


Radiologic Spectrum of Intraspinal Cystic Lesions and 

Their Mimics: A Space-Based Approach 

Revzin, M. V. · Lev, S. 

North Shore University Medical Center 

Manhasset, NY 

PURPOSE 

To review the radiographic spectrum of cystic lesions of the 

spine, emphasizing an approach based on anatomical location. 

We discuss key distinguishing differential features as 

well as mimics. 


We retrospectively reviewed the radiologic studies and clinical 

features of patients with cystic intraspinal lesions. We 

examined the etiology and pathogenesis of these entities. 

Imaging modalities included helical multislice CT , enabling 

two-dimensional reformatted images, CT myelography and 

MR imaging (including advanced MR methods such as CSF 

flow studies and diffusion-weighted imaging when applicable). 

FINDINGS 

Intraspinal extramedullary cystic lesions represent a broad 

spectrum of pathology. Congenital entities include arachnoid 

cysts (which also may be acquired), posterior sacral 

424 

meningoceles, (lipo) myelomeningoceles and neurenteric 

cysts. CT and conventional radiography may be especially 

useful in detecting associated bony anomalies. Dural ectasia 

(associated with Marfan’s syndrome or neurofibromatous, 

for example) may produce diffuse dilatation of the thecal 

sac, which can be differentiated readily utilizing CT myelography 

or MR CSF flow studies. Posttraumatic or postoperative 

pseudomeningoceles also may be contiguous to the 

thecal sac, and may or may not contain neural elements. 

Neoplasms include dermoids and epidermoids, with lipomas 

as a mimic. Infectious processes, such as neurocysticercosis, 

can have a variegated appearance. Postinfectious arachnoiditis, 

with adhesive granulations, may be multiloculated. 

Perineural cysts (Tarlov’s cyst in the sacrum) are congenital 

dilatations of the arachnoid and dura comprising the spinal 

nerve root sleeve. Synovial cysts, associated with degenerative 

disease, are adjacent to the facet joints, posterior to perineural 

cysts. A cystic structure containing two nerve roots, 

in an enlarged composite root sleeve, represents a conjoined 

nerve root. Intraspinal meningoceles, on the other hand, displace 

nerve roots around the cyst. In the setting of trauma, 

nerve root avulsion may produce a pseudomeningocele, usually 

at the neural foramen. Schwannomas (and neurofibromas) 

also may occur at the neural foramen, and may mimic 

cystic lesions. MR imaging with the additional use of a contrast 

agent is essential in distinguishing these entities. 

Intraspinal intramedullary lesions include ventriculus terminalus, 

to be differentiated from hydrosyringomyelia, as well 

as cystic neoplasms. 

CONCLUSION 

Intraspinal cystic entities are diverse in etiology and appearance. 

Consideration of lesion location and key imaging features 

can often help to narrow the differential diagnosis. 

KEY WORDS: Cyst, spine 


Spinal Epidural Space: Anatomy, Normal Variations, 

and Pathologic Lesions on MR Imaging 

Batra, K. 1 · Chhabra, A. 1 · Satti, S. 1 · Patel, S. 1 · Koenigsberg, 

R. 1 · Gonzales, C. 1 · Faerber, E. 2 

1Drexel University College of Medicine (formerly MCP 

Hahnemann), Philadelphia, PA, 2St. Christopher Hospital, 

Drexel University College of Medicine (formerly MCP 

Hahnemann), Philadelphia, PA. 

Spinal epidural space pathologies have been described in 

scattered articles in literature. We attempt to give a comprehensive 

overview on MR imaging, so that the reader will be 

able to: 1. Review the cross-sectional anatomy of spinal 

epidural space, its contents, and venous plexus anatomy. 2. 

Gain knowledge of optimal MR technique including use of 

intravenous contrast in imaging. 3. Learn key imaging findings 

of various benign and malignant epidural lesions with 

illustrative case examples. 4. Familiarize with potential pitfalls 

and pseudomasses. 

KEY WORDS: Epidural, lesions, pseudotumor


Imaging of Spinal Bone Marrow: Normal and Pathologic 

Spectrum 

Scott, B. E. 1 · Ifthikharuddin, S. F. 1,2 · Spitzer, E. M. 1 Monajati, A. 

· 

1 

1 2 Rochester General Hospital, Rochester, NY, University of 

Rochester Medical Center, Rochester, NY 

PURPOSE 

To review the normal imaging characteristics and to provide 

a review of common pathologic processes affecting the bone 

marrow of the spine. 


The normal spectrum of imaging characteristics and spectrum 

of disease of the spinal bone marrow will be reviewed 

using CT and MR imaging. 

FINDINGS 

Several pathologic processes including degenerative, traumatic, 

neoplastic and metabolic conditions affect the spinal 

bone marrow. Specific examples presented include normal 

evolution of bone marrow signal with age, degenerative 

changes, renal osteodystrophy, sickle cell disease, metastases, 

myeloma, hemangioma, anemia and osteomyelitis. 

CONCLUSION 

A practical discussion of the imaging characteristics of normal 

and diseased bone marrow of the spine can facilitate the 

radiologist in the generation of a reasonable differential 

diagnosis. 

KEY WORDS: Spine, bone marrow, signal 


Vertebral Compression Fracture: Multimodal Imaging 

Approach for Percutaneous Vertebroplasty 

Uemura, A. · Numaguchi, Y .· Kobayashi, N. · Yamashita, 

S.·Matsusako, M. 

St. Luke's International Hospital 

Tokyo, JAPAN 

PURPOSE 

There has been no consensus on the role of diagnostic imaging 

in patients with vertebral compression fracture for percutaneous 

vertebroplasty (PVP). The purpose of this exhibit 

is to demonstrate the spectra of multimodal imaging findings 

of vertebral compression fracture which may provide crucial 

information in planning PVP. 


We have reviewed imaging studies for vertebral compression 

fractures including MR imaging, multidetector CT, and 3dimensional 

rotational imaging (3D RI) and correlated 

pathologic findings and postprocedural outcome. 

425 

FINDINGS 

MR imaging allowed detailed information about compression 

fractures as follows: presence of malignancy, bone marrow 

edema/inflammation, intravertebral granulation tissue, 

intravertebral necrotic lesions and fracture-related disk 

changes including secondary Schmorl’s node formation. 

Multiplanar reconstructed images of preprocedural multidetector 

CT and 3D RI had advantages over MR imaging in 

preprocedural assessment of disrupted sites of vertebral margin 

causing cement leakage into adjacent disk. 

CONCLUSION 

Multimodal imaging approach allows detailed analysis of 

vertebral compression fractures and provides crucial information 

in PVP. 

KEY WORDS: Vertebral compression fracture, percutaneous 

vertebroplasty, imaging approach 

Scientific Exhibits


Notes:

Electronic Scientific 

Exhibits (eSE) 1-41 



12:00 PM – 9:00 PM 


6:30 AM - 9:00 PM 


6:30 AM – 11:45 AM 

NOTE: A missing Electronic Scientific Exhibit number 

indicates an abstract has been withdrawn. 

Electronic Scientific Exhibits (eSEs) 32-41 are on 

individual computers. All other eSEs can be viewed 

from any computer. 

Adult Brain 

1-10 

Electronic Scientific Exhibit 1 

Carotid and Vertebral Artery Dissections: Spectrum of 

Diagnostic and Therapeutic Neuroradiology 

Vallone, I. M. · Gentili, M. · Cerase, A. · Bracco, S. · 

Gennari, P. · Venturi, C. 

Policlinico 

Siena, ITALY 

PURPOSE 

Carotid and vertebral dissections are an increasingly known 

cause of stroke, and represent the most common nonatherosclerotic 

cause of stroke in young adults. They may result in 

significant residual stenosis, occlusion, and/or pseudoaneurysms, 

despite prompt systemic anticoagulation therapy. 

The purpose of this scientific exhibit is to present a spectrum 

of clinical presentation, US, CT, MR imaging, and angiographic 

findings at diagnosis, clinical and neuroradiologic 

follow up, and indications for endovascular treatment of 

carotid and vertebral arteries dissections. 


This is achieved by reviewing clinical and neuroradiologic 

charts of 60 dissections in 53 patients (14 females, and 39 

males; age range at diagnosis, 12-77 years, follow up, 6 

months-5 years). 

FINDINGS 

This review includes: 1) unilateral and/or bilateral extracranial 

and/or intracranial traumatic and nontraumatic dissections 

with or without pseudoaneurysm at diagnosis or during 

follow up, extending or not from aortic arch dissection, 2) 

subarachnoid acute hemorrhage at diagnosis, 3) catheter 

427 

angiography-related dissections, and 4) 12 dissections with 

single or double pseudoaneurysm in 11 patients which have 

been treated by endovascular therapy including stenting (n = 

11) and/or vessel occlusion (n = 2). 

CONCLUSION 

Neuroradiologic follow up of carotid and vertebral dissections 

is essential for pertinent treatment options including 

endovascular management. 

KEY WORDS: Carotid and vertebral dissections, diagnostic 

neuroimaging, endovascular treatment 


Automatic Interpretation of Stroke Images by Means of 

an Anatomical Atlas 

Nowinski, W. L. 1,2 · Volkau, I. 1 · BhanuPrakash, K. 1 · Anand, 

A. 1 · Ivanov, N. 1 · Beauchamp, N. 2 

1Bioinformatics Institute, Singapore, SINGAPORE, 

2University of Washington, Seattle, WA 

PURPOSE 

Our goal is to automate the interpretation of multimodal 

stroke scans. This is achieved by using a deformable anatomy 

atlas with a fast method for warping the atlas and overlaying 

it onto the scan. 


An electronic version of the Talairach brain atlas with gross 

anatomy was developed earlier (1). The electronic atlas is 

segmented fully and labelled including the cortical areas 

(gyri and Brodmann’s areas) and subcortical structures. A 

fast transformation warping the atlas to the scan is developed. 

This transformation is done in five steps: 1) extraction 

of the midsagittal plane, 2) approximation of the intercommissural 

plane, 3) determination of cortical extents, 4) piecewise 

linear scaling of the atlas to fit the midsagittal plane, 

intercommissural plane and extents of the cortex, and 5) 

nonlinear fitting of the atlas cortical outline to that of the 

scan. 

FINDINGS 

The proposed method is able to register the atlas to the scan 

automatically within a few seconds. The method works for 

numerous types of scans including CBF, CBV, MTT, TTP, 

PKHT, DWI, MRA, and T2. After registration, the individualized 

atlas provides underlying cortical and subcortical 

anatomy for these multimodal data. 

CONCLUSION 

The developed solution is a useful aid potentially speeding 

up the interpretation of stroke images and increasing the 

interpreter’s confidence. 

REFERENCES 

1. Nowinski WL, Thirunavuukarasuu A. The Cerefy Clinical Brain 

Atlas on CD-ROM. Thieme, New York, 2004. 

KEY WORDS: Adult brain (stroke), electronic brain atlas 

Electronic Scientific Exhibits

Electronic Scientific Exhibits 


Anatomy of the Hippocampus Using 3 T MR Imaging 

Salamon, N. · Schultze-Haakh, H. · Laub, G. · Salamon, G. · 

Sinha, U. 



PURPOSE 

To demonstrate high-resolution anatomy of the hippocampus 

and surrounding structures. 


Using 3.0 T MR imaging (Siemens Sonata), we scanned 15 

normal volunteers (20 to 40 years old). Diffusion tensor 

imaging also was obtained to visualize the extrahippocampal 

connections of the white matter fibers. 

FINDINGS 

Three T images showed detailed anatomy of the Ammon’s 

horn, subiculum, and fimbria. Adjacent structures such as 

amygdala, optic tracts, lateral geniculate bodies also are well 

visualized. Fiber tracking images show white matter fibers 

communicating through the temporal stem to the frontal and 

occipital regions. 

CONCLUSION 

Three T MR imaging is a promising tool to learn detailed 

anatomy of the hippocampus and surrounding structures. 

KEY WORDS: Hippocampus, high-resolution imaging, 3 T 

MR imaging 

Electronic Scientific Exhibit 4 (Moved to SE 16A) 

Review of Imaging Findings in Neurosarcoidosis 

Lury, K. M. · Smith, J. · Matheus, M. G. · Castillo, M. 



PURPOSE 

To present an interactive electronic educational exhibit 

describing the clinical and imaging findings of sarcoidosis in 

the central nervous system (CNS). 


We have performed a thorough review of current neuroradiology 

literature relating to neurosarcoidosis. Our interactive 

PowerPoint presentation integrates this knowledge with our 

own plethora of fascinating case material. The clinical material 

is a blend of the unusual, and “classic” imaging findings 

in this fairly common, yet unusual, disease. 

FINDINGS 

Five to 25% of patients with systemic sarcoidosis also have 

CNS involvement. Patients with systemic disease initially 

may present with neurologic symptoms, and rarely, the sarcoidosis 

may be isolated to the CNS. Sarcoidosis has protean 

manifestations in the central nervous system and can involve 

the intracranial, spinal, and extra cranial head and neck. 

428 

CONCLUSION 

This exhibit will provide the viewer with concise and indepth 

knowledge of the clinical and imaging findings of sarcoidosis 

affecting the brain, spinal cord, and extra cranial 

head and neck. 

KEY WORDS: Neurosarcoidosis, brain, spine 


Serial Imaging Procedures of Early and Delayed 

Radiation Damage to the Central Nervous System 

Mironov, A. 

Kantonsspital Aarau 

Aarau, SWITZERLAND 

PURPOSE 

Renewed growth of a mass at the site of previously treated 

brain tumor raises the issues of indication for and choices of 

treatment. A variety of functional neuroradiologic imaging 

attempts to distinguish between recurrent tumor and radiation 

necrosis based on their different metabolic characteristics. 

Thus far, SPECT studies, PET scans, and MRS, 

although helpful, have not demonstrated definitive means of 

providing a noninvasive differentiation between these 

lesions. We analyzed, retrospectively, the serial MR imaging 

and PET scans over a wide period after the treatment in an 

effort to determine whether the value of imaging characteristics 

would improve our ability to differentiate between 

recurrent tumor and radiation necrosis. 


The follow-up imaging of a cohort of 37 patients have been 

analyzed, retrospectively, for a period of 10 years. All 

patients had been treated previously with surgery for brain 

mass and consequently underwent radiation therapy. 

Additionally, 29 patients underwent chemotherapy. The 

included cases harbored glioblastoma multiforme (15), 

anaplastic astrocytoma (7), low grade glioma (5), oligodendroglioma 

and mixed oligoastrocytoma (6), meningeoma 

and hemangiopericytoma (2), vestibular schwannoma (2). 

All patients underwent serial follow-up MR imaging and 

PET scans (6 cases) in different time sequences over a maximum 

period of 10 years. 

FINDINGS 

The influence of radiation can act on the natural history of 

brain tumors by causing different morphologic and metabolic 

changes over an unlimited interval after the treatment. 

They become manifest as: 1. Early radiation reaction occurs 

during or shortly after the treatment. The induced findings 

include focal or diffuse brain edema; 2. Early delayed reactions 

arising a few weeks to 12 - 24 months after treatment 

and involve the intrafocal and/or the perifocal blood brain 

barrier damage and demyelination. The imaging reveals an 

intensive and remarkable bright peripheral or intrafocal 

enhancement. There is a remarkable discrepancy between 

the extension of tissue enhancement (some times broadly 

based) and the existing nonrelevant mass effect; 3. Late 

delayed reactions occur more than 24 months after treatment, 

following an occult phase in which several radiationinduced 

cellular and molecular pathologic mechanisms had

een discussed. The onset of morphologic change may be 

manifested more as long as 10 years later. The contrastweighted 

imaging demonstrates a broadly based bright 

intensity enhancement with festoon-like or facet-like configuration. 

It develops often centrifugal from the margin of the 

postsurgery brain damage and involves both the gyral surface 

and the subependymal periventricular space. The mass 

effect is usually inappropriately small. In case of rapid development 

the mass can be increased considerably by brain 

edema. The natural history of radiation necrosis implies an 

intermittent course and can persist for several months or 

years. The response to the steroid treatment may be abounding. 

CONCLUSION 

The radiation necrosis is a dynamic process with unknown 

latent period. It accompanies the recurrent/persistent intrinsic 

brain lesions far more often than recognized. Evidence 

then is provided suggesting that serial diagnostic imaging in 

addition to steroid treatment may provide information that is 

useful in discriminating pure necrosis and specimens of 

mixed tumor and necrosis. 

KEY WORDS: Radiation necrosis, radiotherapy, brain neoplasm 


Perfusion MR Imaging of Brain Mass Lesions 

Hakyemez, B. · Erdogan, C. · Parlak, M. 

Uludag University Medical School 

Bursa, TURKEY 

PURPOSE 

Vascularity and the degree of angiogenesis are important 

components of tissue perfusion which can serve in the evaluation 

of different types of brain mass lesions. 


Perfusion MR imaging is a dynamic technique that measures 

cerebral perfusion via assessment of several hemodynamic 

parameters such as cerebral blood volume, cerebral blood 

flow, and mean transit time. Among these parameters cerebral 

blood volume seems to be most useful for the mass 

lesions. Beyond the applications like the evaluation of tissue 

at risk after acute stroke or preoperative grading of brain 

tumors, perfusion imaging may provide useful data for the 

differential diagnosis of various types of brain lesions. 

FINDINGS 

In this presentation, the basic methods of dynamic contrastenhanced 

susceptibility-weighted echo-planar perfusion MR 

imaging are described first. Second, selected conventional 

and perfusion MR findings of various brain lesions obtained 

from 144 patients are discussed emphasizing the effect of 

perfusion data in the differential diagnosis and management 

of mass lesions. 

CONCLUSION 

The lesions consisted of a series of typical and atypical 

meningiomas, high and low grade gliomas, cerebral metastases 

(lung, breast, renal cancer, and colon), lymphomas, 

429 

hemangioblastomas, epandimomas, schwanomas, PNETs 

and medulloblastomas, choroid plexus papilloma, abscesses, 

tumefactive demyelinating lesion, and tuberculosis lesion. 

KEY WORDS: Brain, MR imaging—perfusion, brain neoplasm 


Preprocedural Vascular Assessment for Posterior Fossa 

Decompression Surgery Utilizing CT Angiography: A 

Review of 280 Cases 

Grimaldi, G. · Woldenberg, R. · Shah, T. C. 


Manhasset, NY 

PURPOSE 

Surgical decompression of the posterior fossa is the dominant 

method of treatment for patients with symptomatic 

Chiari malformations. However, given the complex anatomy 

of both the arterial and venous systems in this region, the 

need for preoperative evaluation of these structures is often 

extremely helpful in preoperative planning. 

Withdrawn 


The availability of noninvasive vascular imaging via MR 

and CT has contributed significantly to the reduction of complications 

related to vascular injury. CT offers the added benefit 

of providing information about vascular and skull anatomy 

simultaneously. 

FINDINGS 

We reviewed CT angiography examinations of 280 patients 

who were undergoing preoperative evaluation for possible 

posterior fossa decompression for symptomatic Chiari malformation. 

CONCLUSION 

We focus our review on the craniocervical junction venous 

anatomy, specifically the suboccipital venous complex and 

the related condylar and emissary veins. 

KEY WORDS: Chiari, suboccipital venous plexus 




Visualization of Hemodynamics in the Field of 

Neuroradiology Using Time-Resolved Three- 

Dimensional Phase-Contrast MR Imaging 

Isoda, H. 1 · Hirano, M. 2 · Yamashita, S. 1 · Inagawa, S. 1 · 

Kosugi, T. 3 · Alley, M. T. 4 · Markl, M. 4 · Pelc, N. J. 4 · 




3Renaissance of Technology Corporation, Hamamatsu, 

JAPAN, 4Stanford University School of Medicine, Stanford, 

CA 

PURPOSE 

Hemodynamics affects the development and growth of vascular 

lesions, such as aneurysms and atherosclerotic plaques. 

Being able to predict the occurrence of vascular lesions and 

their growth, based on hemodynamics, would enable us to 

judge the outcome of the patient. This information would be 

beneficial for the prevention of disease and the planning of 

treatment for each patient. To date, two-dimensional cine 

phase contrast MR imaging technique and other imaging 

techniques have not been enough for this purpose. Recently 

a new cine phase contrast MR imaging technique named 

time-resolved three-dimensional phase-contrast MR imaging 

(4D Flow) (1) has been developed. It provides us with invivo 

four-dimensional hemodynamic information including 

space and time. The purpose of this exhibit is to introduce 

the 4D Flow technique and to show its usefulness in the field 

of neuroradiology. 


The 4D Flow technique for intracranial arteries and carotid 

arteries in human beings and vascular models was carried 

out with a 1.5 T MR unit (Signa Infinity Twinspeed with 

EXCITE XI (version 11), GE, Milwaukee, Gmax = 40 

mT/m). The 4D Flow technique was based on radiofrequency-spoiled 

gradient-echo sequence and velocity encoding 

was performed along all three spatial directions. Fourdimensional 

data including time dimension were obtained. 

Measurements were gated retrospectively to the electrocardiogram 

and cine series of three-dimensional data sets were 

generated. Utilized imaging parameters were as follows; 

TR/TE/NEX = 4-5.8/1.9-2.1/1, FA = 15, FOV = 160 x 160, 

voxel size = 1 x 1 x 1-3, VENC = 20-60 cm/s, 20 phases during 

one cardiac cycle, imaging time = 20-40 min. Timeresolved 

images of three-dimensional stream lines, threedimensional 

particle traces and two-dimensional velocity 

vector fields on arbitrary planes were calculated from fourdimensional 

data sets by flow visualization software 

(EnSight). 

FINDINGS 

Time-resolved two-dimensional velocity vector fields on the 

planes traversing the carotid siphon, basilar tip, and carotid 

bulb were visualized clearly. When we generated threedimensional 

stream lines originating from the bilateral C2 

segment of internal carotid arteries and the basilar artery, we 

were able to see the three-dimensional stream lines from the 

Willis circle to the bilateral M2 segment of middle cerebral 

arteries. Time-resolved images of three-dimensional particle 

traces also clearly demonstrated intracranial arterial flow 

430 

dynamics. Two-dimensional velocity vector fields on arbitrary 

planes in the aneurysm model clearly showed vortex 

flow within the aneurysm. Time-resolved images of threedimensional 

stream lines and three-dimensional particle 

traces clearly demonstrated that the aneurysm had threedimensional 

complex vortex flows within it during systolic 

phase and that flow streams ran out into the two adjacent M2 

branches of the middle cerebral artery. These images could 

be observed as cine images. 

CONCLUSION 

The 4D Flow technique is a promising noninvasive technique 

and provides us with time-resolved three-dimensional 

in-vivo hemodynamic information about intracranial and 

carotid arteries. In the future we hope to be able to predict 

the occurrence of vascular lesions and their growth. 

Knowing this would aid us in preventing disease and making 

good therapeutic plans. 

REFERENCES 



KEY WORDS: Aneurysm, hemodynamics, phase-contrast 

MR imaging 


Cerebral Vasospasm following Aneurysmal 

Subarachnoid Hemorrhage: Is It Related to Surgical or 

Endovascular Treatment? 

Said, A. H. M. · Lownie, S. P. · Pelz, D. M. · Saleem, S.N. 

University of Western Ontario 

London, ON, CANADA 

PURPOSE 

In patients with aneurysmal subarachnoid hemorrhage, the 

incidence of vasospasm following surgical vs endovascular 

intervention is unresolved. In one recent study involving 515 

patients, there was no significant difference between the two 

regarding incidence of vasospasm. Other studies described a 

lower incidence of vasospasm with coiling. Our objective is 

to evaluate the incidence of vasospasm following surgical 

and endovascular intervention in our patients and whether 

the selected line of treatment has an impact on the final outcome. 


We are reviewing retrospectively the charts of patients who 

had aneurysmal subarachnoid hemorrhage (SAH), and who 

had undergone either endovascular coiling or surgical clipping 

at our institution over the last 3 years. We have identified 

approximately 150 such patients. The anatomical features 

of the aneurysms are defined. The amount of bleeding 

and the development of vasospasm as monitored clinically 

and by transcranial Doppler (TCD) and/or cerebral angiography 

are recorded. Patients are classified into three groups: 

1) severe vasospasm warranting intraarterial papaverine 

therapy or angioplasty, 2) symptomatic vasospasm responding 

to medical treatment and 3) no, or asymptomatic 

vasospasm. We are excluding patients: 1) who had undergone 

both procedures within 2 weeks, 2) who had an addi-

tional surgical or radiologic intervention in the same session, 

3) those who had delayed intervention more than 1 week 

after SAH and 4) who had vasospasm at the time of intervention. 

Patients are classified also according to their clinical 

assessment before and after the procedure using Hunt and 

Hess criteria into five groups: a) complete recovery, b) 

incomplete improvement, c) equivocal, d) worsening and e) 

death. 

FINDINGS 

Initial analysis of 90 patients shows that 54 had coiling and 

36 had clipping. Among the coiling group, 20 patients (37%) 

had vasospasm, of which three patients (5.6%) had undergone 

balloon angioplasty. Twenty-four of 36 patients (67%) 

who had clipping developed vasospasm, of which seven 

patients (19%) needed angioplasty. Initial evaluation of clinical 

outcome suggests more favorable outcomes in the coiling 

group. Full statistical analysis of all data will be presented. 

CONCLUSION 

Our results show a definite trend towards lower incidence of 

symptomatic vasospasm with endovascular coiling of 

aneurysms than with surgical clipping, with a lower incidence 

of interventional vasospasm therapy in the coiling 

group. Initial analysis suggests a trend toward better clinical 

outcome in the coiling group. 

KEY WORDS: Vasospasm, subarachnoid hemorrhage, 

aneurysm 


Anatomy and Pathology of the Cerebral Veins 

Salamon, N. · Juncosa, A. · Hooshi, P. · Valentino, D. · 

Salamon, G. 



PURPOSE 

To show the value of various neuroimaging techniques to 

demonstrate cerebral veins, including MR imaging, MRV, 

and 3D volume-rendered images of CTA as a learning tool of 

venous anatomy. To demonstrate a variety of venous anomalies, 

normal variants and pitfalls, seen in routine brain 

imaging. 


Illustrative cases of various venous anomaly and pathology 

were collected between 2002 to 2004. 

FINDINGS 

This exhibit includes normal intracerebral venous anatomy, 

normal variant, and anomaly such as absent sagittal sinus, or 

hypoplastic sinus, developmental venous anomaly, venous 

thrombosis, AVM, and AVF. 

CONCLUSION 

MRV, MR imaging, and CT angiography became a routine 

noninvasive method to evaluate intra or extracranial venous 

diseases. Precise knowledge of venous anatomy is important 

431 

for interpretation of these studies. Venous anatomy is not 

well documented in classic imaging textbooks. This exhibit 

will be a core learning tool of venous anatomy. 

KEY WORDS: Anatomy, cerebral veins 

Functional 

11 


Central and Precentral Sulci: Display of Reproducible 

Patterns on MR Imaging 

Kim, B. · Jung, S. · Ahn, K. · Kim, Y. · Byun, J. 



PURPOSE 

The central sulcus (CS) is important for localization of motor 

and sensory cortices. The purpose of this study was to 

demonstrate reporducible patterns of CS and precentral sulcus 

(PS) on MR imaging, and to enable physicians to recognize 

these patterns on imaging studies. 


Brain MR images including T1-, T2-weighted and FLAIR 

images were obtained in 80 healthy persons (male: 40, 

female: 40, age : 20-49, median: 40) were analyzed to determine 

the patterns displayed on axial and sagittal images; 

their relationship to the sylvian fissure, superior and inferior 

frontal sulci, presence and type of motor hook region, pattern 

of gray-white difference in CS and PS. 

FINDINGS 

Axial MR images displayed the PS intersecting the superior 

frontal sulcus (93.8%) anterior to the CS. Motor hook region 

was identified (96.9%) as single (68.1%) or multiple 

(28.8%) hump. CS intersects interhemispheric fissure 

(81.3%). Gray-white distinction more distinct at the CS than 

that at the precentral sulcus on T1-weighted (86.3%) and 

FLAIR images (90%). On lateral sagittal images, CS does 

not intersect directly the sylvian fissure, but to the subcentral 

gyrus (87.5%). Precentral sulcus, which intersects inferior 

frontal sulcus (98.1%), intersects sylvian fissure (93.1%). 

CONCLUSION 

The precentral and central sulci show reproducible patterns 

on routine clinical MR images obtained at 1.5 T. The exhibit 

will display these patterns on MR and fMR images with 

review of literature. 

KEY WORDS: Brain anatomy, central sulcus 




“Doctor! What Is This Bump on My Head?”: Evaluation 

of Lesions Presenting in the Scalp and Calvarium 

Pivawer, G. 1 · Bajwa, Z. 1 · Reede, D. L. 1 · Smoker, W. R. 2 · 

Gentry, L. 3 · Holiday, R. 4 

1 2 Long Island College Hospital, Brooklyn, NY, University of 

Iowa, Iowa City, IA, 3University of Wisconsin Medical 

Center, Madison, WI, 4New York Eye and Ear Infirmary, 

New York, NY 

PURPOSE 

This exhibit reviews the anatomy of the scalp and calvarium. 

Imaging characteristics of the lesions originating in each of 

the radiographically discernable layers of the scalp and calvarium 

are presented. Pertinent anatomical considerations 

for surgical planning are discussed. 


A recent article by Hayman, et al (1) reviewed the five 

anatomic layers of the scalp (skin, connective tissue, epicranial 

aponerousis, loose areolar tissue, and pericranium); only 

three of these five layers are discernable using cross- sectional 

imaging. This article sparked our interest to analyze 

patients who present with focal or diffuse bumps on the 

head. We reviewed 76 cases that presented at four major 

teaching institutions in the United States. Theses cases were 

analyzed with regard to size, location, imaging characteristics, 

and potential patterns of disease spread. 

FINDINGS 

This exhibit reviews the anatomy of this region and provides 

illustrative pathologies originating in each of the three radiographically 

discernable layers of the scalp and calvarium. 

Some of the lesions presented include: 1) Superficial layer 

(skin): carcinoma of the skin, sebaceous cyst, superficial 

infections, and calcifications; 2) Middle layer (connective 

tissue layer): lipomas, lymphoma, and vascular lesions; 3) 

Deep layer (the galea/subgalea/periosteal complex): 

cephalohematoma, leptomeningeal cyst; 4) Calvarium: 

Paget’s disease, fibroosseous disease, metastases, 

osteomyelitis, etc. 

CONCLUSION 

Lesions in scalp and calvarium are relatively rare; an understanding 

of the layer of origin enables the radiologist to provide 

a reasonable differential diagnosis and predict potential 

patterns of disease spread. 

REFERENCES 

1. Hayman LA, et al. Clinical and imaging anatomy of the scalp. 

J Comput Assist Tomog 27(3):454-459 

KEY WORDS: Scalp lesions 

Head and Neck 

12-25 

432 


“Hey!! I Have a Lump in My Jaw!”: Nonodontogenic 

Lesions of the Jaw 

Kanekar, S. G. 1 · Smoker, W. R. K. 1 · Gentry, L. R. 2 · Reede, 

D. L. 3 · Phillips, C. D. 4 

1 2 University of Iowa, Iowa City, IA, University of Wisconsin, 

Madison, WI, 3Long Island College Hospital, New York, NY, 

4University of Virginia, Charlottesville, VA 

PURPOSE 

The great majority of jaw lesions have an odontogenic etiology. 

In this exhibit we present a large collection of nonodontogenic 

maxillary and mandibular lesions. 


We retrospectively reviewed the imaging studies (CT and 

MR) on 784 patients with clinically suspected or incidentally 

diagnosed jaw masses from 4 major academic medical 

centers. Odontogenic masses, being most common, were 

excluded from further review and only nonodontogenic 

(total 214 patients) masses were selected for further consideration. 

Some of these lesions exhibited a classic “Aunt 

Minnie” benign appearance and had no pathologic confirmation. 

The remaining lesions had undergone biopsy or surgical 

resection and were proven pathologically. 

FINDINGS 

For the purposes of this exhibit, we have chosen to present 

these nonodontogenic lesions in the follow manner, based 

upon pathologic classification. 1) Benign bony lesions: 

Tumors: Exostosis, osteoma, chondroma, synovial chondromatosis, 

and germ cell lesions. Histiocytosis. Fibroosseous 

lesions: Fibrous dysplasia and Paget’s disease. Giant cell 

tumors: Brown tumors associated with hyperparathyroidism. 

“Reparative granulomas.” 2) Malignant bony lesions: 

Metastatic: From breast, kidney, lung, prostate and hepatoma 

primary tumors. Primary: Multiple myeloma, malignant 

lymphoma, osteosarcoma, ewings sarcoma and mucoepidermoid 

carcinoma. 3) Vascular lesions: Hemangioma, vascular 

malformation. 4) Neurogenic masses. 5) Cystic masses: 

Epithelial-nasopalatine duct cyst, Nonepithelial-simple bone 

cyst, aneurysmal bone cyst, Stafne bone cyst. 6) 

Miscellaneous: Infection: Osteomyelitis, 

Osteoradionecrosis, Gorham’s syndrome (vanishing bone). 

The classic imaging features of these lesions are presented 

along with diagnostic “pearls” to their diagnoses. 

CONCLUSION 

Although odontogenic lesions are the most common lesions 

in the jawbones, various nonodontogenic lesions also must 

be considered in the differential diagnosis. This exhibit illustrates 

a variety of these common and uncommon nonodontogenic 

lesions and offers salient diagnostic “pearls.” 

KEY WORDS: Nonodontogenic, jaw, lesions



Fractures of the Zygomaticosphenotemporal Buttress: Lacrimal Gland Pathology: Tumor and Tumor-Like 

Classification, Diagnostic Evaluation, and Common Lesions 

Complications 

Khosla, A. 1,2 

433 

Ziegert, A. J. 1 · Gentry, L. R. 1 · Mount, D. L. 1 · Smoker, W. 

R. K. 2 · Reede, D. L. 3 · Pulfer, K. A. 1 

1 2 University of Wisconsin, Madison, WI, University of Iowa, 

Iowa City, IA, 3Long Island College Hospital, New York, NY 

PURPOSE 

This scientific exhibit reviews the imaging findings, classification, 

and complications related to fractures of the zygomaticosphenotemporal 

buttress (ZSTB). The ZSTB is 

defined as the vertically oriented triangular strut of bone at 

the junction of the zygomatic, sphenoid, and temporal bones. 

It is strategically located adjacent to the lateral aspect of the 

orbit, orbital apex, superior orbital fissure, central skull base, 

middle cranial fossa, and orbital roof. 


We retrospectively reviewed all CT scans (n = 980) obtained 

over a 3-year period of time that were obtained to evaluate 

patients with craniofacial trauma. All patients with fractures 

that involved the ZSTB were entered into the study. The 

cases were analyzed in detail to identify the primary mechanism 

of injury, type of ZSTB fracture, and associated facial 

fractures. The medical records and imaging studies (CT, MR 

imaging, MR angiography, angiography) were reviewed to 

assess for orbital and intracranial complications. 

FINDINGS 

A total of 144 fractures of the ZSTB were identified. 

Numerous complications resulting from the ZSTB fractures 

were identified and could be grouped into several categories 

(ocular injury, enophthalmos, extraocular muscle restriction, 

optic nerve injury, cranial nerve injury, orbital hematoma, 

CSF leaks, vascular injuries, arteriovenous fistula, epidural 

hematoma, contusions). Four main patterns of ZSTB fractures 

were identified and these patterns often suggested a 

high likelihood of specific complications. 

CONCLUSION 

Zygomaticosphenotemporal buttress fractures occur in four 

main patterns and are associated with a high frequency of 

orbital, ocular, vascular, and intracranial complications. This 

exhibit analyzes the patterns of injury and associated complications 

for each subtype of ZSTB fracture. 

KEY WORDS: Facial, trauma, fractures 

1 Mallinckrodt Institute of Radiology, St. Louis, MO, 

2 Veterans Administration Medical Center, St. Louis, MO 

PURPOSE 

Lesions affecting lacrimal glands can be classified as: a) 

epithelial (40-50%) and nonepithelial lesions and b) benign 

and malignant. The learning objectives are: 1) to learn about 

benign and malignant tumors of lacrimal glands and 2) to 

learn about imaging findings and an approach to imaging 

diagnosis. c) Correlation with pathologic findings. 


Retrospective analysis of pathologically proven cases of 

lacrimal gland pathology was evaluated for imaging findings. 

Findings: Epithelial lesions are mostly neoplastic. The 

most common benign mass is a benign mixed tumor and the 

most common malignant neoplasm is adenoid cystic carcinoma. 

Benign mixed tumors are well defined, homogenous 

and usually moderately enhancing. Bone erosion, calcification, 

nodularity and infiltrative pattern favor malignancy. 

Nonepithelial lesions mostly include inflammatory and lympho-proliferative 

pathology. These include dacryoadenitis, 

pseudo tumor, sarcoidosis, and the lympho-epithelial 

lesions. These have enlarged soft tissue contour and may be 

unilateral or bilateral. Miscellaneous lesions include: congenital 

abnormalities, amyloid tumor, TB, cysticercosis, and 

fibrous and neurogenic tumors of lacrimal glands. 

FINDINGS 

The imaging findings seen on CT, MR imaging, 

Ultrasonography and Isotope imaging including PET and 

will be described with appropriate examples and differentiating 

features will be discussed. An imaging approach to 

diagnosis of lacrimal gland lesions will be discussed. 

CONCLUSION 

Knowledge of clinical findings and imaging findings can 

help to shortlist the differential diagnosis of lacrimal gland 

pathology. 

KEY WORDS: Lacrimal gland, orbit, neoplasms 


Diagnostic Evaluation of Head and Neck Cancer 

McCue, J. 1 · Gentry, L. R. 1 · Smoker, W. R. K. 2 · Reede, D. 

L. 3 · Hartig, G. K. 1 · Harari, P. M. 1 · Corliss, R. F. 1 

1 University of Wisconsin, Madison, WI, 2 University of Iowa, 

Iowa City, IA, 3 Long Island College Hospital, New York, NY 

PURPOSE 

Provide residents, fellows, and other physicians a comprehensive 

teaching and reference module in the diagnostic 

evaluation and treatment of squamous cell cancer of the head 

and neck. 




Shared display electronic interactive format. 

FINDINGS 

This comprehensive module provides background in squamous 

cell cancer of the head and neck. Anatomy is reviewed 

using illustrations, diagrams, imaging, and surgical/endoscopic 

techniques. The TNM staging system is reviewed for 

all commonly encountered sites including: 1) paranasal sinus 

and nasal cavity, 2) lip and oral cavity, 3) oropharynx, 4) 

hypopharynx, 5) larynx, 6) nasopharynx. Illustrative cases 

including diagrams, imaging, and surgical/endoscopic findings 

are provided for each T and N stage disease. Common 

treatment options are reviewed and provided for reference 

during the evaluation of cases. Typical pathologic/histologic 

findings are reviewed as well as routes of spread in head and 

neck cancer. 

CONCLUSION 

This module provides a comprehensive teaching module 

addressing the issues commonly encountered in the diagnostic 

evaluation and treatment of squamous cell cancer of the 

head and neck. The module provides a framework which is 

intended to be useful for those physicians inexperienced in 

the field of head and neck cancer as well as provide a reference 

module which will be useful for experienced providers 

in the staging and treatment of these cancers. 

KEY WORDS: Cancer, neoplasm, staging 


Eponymous Syndromes of the Head and Neck 

Jennings, J. E. · Gentry, L. R. 

University of Wisconsin 

Madison, WI 

PURPOSE 

The head and neck radiologist must be familiar with a multitude 

of eponymous syndromes as they pertain to the head, 

neck, and skull base. Classification of complex imaging 

findings with their appropriate eponyms serves to enhance 

communication with our ENT and neurosurgical colleagues, 

who commonly employ eponyms in everyday clinical parlance. 

Unfortunately, eponyms as medical terms tend to suffer 

from a lack of clinical precision, compounded by the relative 

infrequency with which they arise in practice. The purpose 

of this exhibit is to review a number of such syndromes, 

ranging from the familiar to the obscure, so as to solidify 

one’s understanding of the diseases and their eponymous 

monikers. 


Using varied search protocols including the indices of the 

major neuroradiology textbooks and the world wide web, we 

identified 30 syndromes named for those who discovered or 

described them. We supplemented our own case files with 

those of colleagues at several major teaching institutions in 

order to assemble multimodality images which clearly depict 

each syndrome. The images are classified according to 

region (neck, orbit, skull base, etc.) and presented with text 

434 

discussing the major imaging features of each syndrome, as 

well as a brief historical summary of the origin of the 

eponym. 

FINDINGS 

The material is presented in an interactive style using hypertext 

format in which the user can select to view the cases 

either as unknowns with “multiple-choice” style questions, 

or in an indexed format which allows a rapid targeted review. 

CONCLUSION 

This hypertext-driven computer-aided exhibit is intended as 

a focused and informative review of some of the more clinically 

important syndromes of the head and neck. After viewing 

the exhibit, the user should feel more comfortable with 

the precise definitions of many of the eponymous lesions 

and, as a result, be better able to use eponyms in everyday 

clinical parlance. 

KEY WORDS: Syndromes, eponyms, head and neck 


Neck Masses in Adults: Imaging, Clinical, and 

Anatomical Considerations 

Watson, T. D. 1 · Reede, D. L. 1 · Holiday, R. A. 2 · Smoker, W. 

R. K. 3 

1 2 Long Island College Hospital, Brooklyn, NY, New York 

Eye and Ear Infirmary, New York, NY, 3University of Iowa, 

Iowa City, IA 

PURPOSE 

1) Review imaging characteristics of common neck masses 

in adults. 2) Learn key anatomical landmarks that impact on 

the diagnosis and management of patients with neck masses. 

3) Learn an algorithm for the differential diagnosis of neck 

masses based on clinical and radiographic findings. 


Imaging is an integral part of the current evaluation of an 

adult presenting with a neck mass. Traditionally, the radiographic 

evaluation of neck masses focuses on the location 

as defined by cross-sectional imaging (carotid space, posterior 

cervical space, etc.); then the differential diagnosis is 

derived from broad categories of disease (congenital, inflammatory, 

etc.). We suggest a different approach, using tailored 

algorithms that address specific locations of clinical complaints. 

We reviewed 215 cases of adults presenting with 

neck masses at three institutions over a 1-year period. Based 

on this review there were six common locations of clinical 

complaint as follows: 1) under the chin, 2) under the jaw, 3) 

lateral to the jaw, 4) anterior neck, 5) lateral neck, and 6) diffuse 

swelling. 

FINDINGS 

This exhibit reviews the common neck masses encountered 

in adults. The importance of the clinical findings as well as 

the key anatomical landmarks impacting on diagnosis and 

management are stressed. Illustrative pathologies encountered 

in the six common locations of clinical complaints are 

presented. Some of the lesions presented include: 1) under 

the chin: dermoid; 2) under the jaw: submandibular gland 

pathology, ranula; 3) lateral to the jaw: parotid tail lesions; 4)

anterior neck: thyroglossal duct cyst, ectopic thyroid tissue; 

5) lateral neck: branchial cleft cyst, carotid body tumors, 

laryngocele, neural tumor; and 6) diffuse swelling: lymphoma. 

CONCLUSION 

A knowledge of the location of the clinical complaint and 

key anatomical landmarks on cross-sectional imaging is 

essential when evaluating a patient with a neck mass. This 

enables one to develop a differential diagnosis based on both 

clinical and anatomical findings. 

KEY WORDS: Neck, masses, adult 


Lymphoma of the Head and Neck: A Pictorial Essay 

Go, J. L. · Tran, T. · Hoang, P. · Zee, C. S. 

University of Southern California Los Angeles 


PURPOSE 

Lymphoma accounts for 3-5% of all malignancies and may 

be subdivided into the Hodgkin’s (HL) and non-Hodgkin’s 

type (NHL). Because of its increased frequency in the head 

and neck, imaging plays a key role in diagnosis. 


This exhibit will review the salient features which distinguish 

HL from NHL within the head and neck including 

demographics, natural history, histopathology, and utilize a 

multimodality approach with CT, MR imaging, and PET/CT 

imaging to describe the differences between these two different 

disease processes. 

FINDINGS 

Nodal masses typically present as well circumscribed, 

homogeneous bulky masses. However, extracapsular extension, 

necrosis, and calcification may occur which may make 

diagnosis difficult. In addition, nonnodal disease within the 

head and neck, especially in NHL, may mimic squamous cell 

carcinoma or minor salivary gland malignancies and always 

should be considered in the differential of submucosal masses. 

This exhibit will demonstrate by example the myriad 

appearances of lymphoma of the head and neck on CT, MR 

imaging, and PET/CT. 

CONCLUSION 

Lymphoma, in many ways, mimics other pathologic processes 

within the head and neck and always should be in the differential 

for nodal and nonnodal masses of the head and 

neck. 

KEY WORDS: Lymphoma, neoplasm, head and neck 

435 


Interactive Tutorial of Lymph Node Imaging 

Lee, J. S. · Delman, B. N. · Jayaraman, S. · Som, P. M. 

The Mount Sinai School of Medicine of New York University 

New York, NY 

PURPOSE 

Understanding and using the accepted terminology in reporting 

lymph nodes are essential, as the ENT community has 

widely accepted the imaging-based lymph node classifications 

described in 1999. This computer-based lymph node 

tutorial covers: a) the classification of lymph nodes (levels I 

through VI) and its clinical utility; b) typical CT and MR 

protocols; c) imaging appearance of reactive lymph nodes; 

d) illustrations of necrotic, extracapsular, and conglomerate 

lymph nodes with their differential possibilities; and e) pattern 

of nodal spread of a known primary tumor. 


This computer program will allow users to learn the classification 

scheme in an interactive manner. This is followed by 

a challenge mode in which a user has an opportunity to classify 

and stage the lymph nodes from a series of axial images. 

The necrotic, extracapsular, and conglomerate lymph nodes 

then are illustrated along with their differential considerations. 

An opportunity to correlate with PET imaging is given. 

Finally, the nodal spread pattern of a known primary tumor 

is shown. In the appendix, the imaging protocols used at our 

institution are provided. 

FINDINGS 

In short period of time, the users will review, learn, and 

appreciate the simplicity of the imaging-based lymph node 

classifications used by many clinicians. 

CONCLUSION 

This interactive tutorial allows a comprehensive review of 

cervical lymph nodes in a challenging and enjoyable format. 

KEY WORDS: Lymph node, tutorial 


Craniosynostosis and Craniofacial Malformations in 

Children: Optimizing the Low-Dose Spiral CT 

Technique 

Vazquez, E. · Auger, C. · Piqueras, J. · Poch, J. M. · Lucaya, J. 

Hospital Vall d’Hebron 


PURPOSE 

CT imaging is required in children with craniosynostosis, 

particularly in cases with inconclusive findings on simple 

views, presurgical assessment, or in multiple or syndromic 

craniosynostosis. Radiologist should be familiar with the 

main imaging signs, the optimal CT imaging technique, and 

the differential diagnosis of this condition, in order to 

achieve earlier diagnosis and best surgical results. 




Craniosynostosis is the premature closure of one or several 

cranial sutures. Plain films are used to first evaluate suspected 

craniosynostosis. Helical CT with 3D and MIP reconstructed 

images is valuable in inconclusive cases or presurgical 

planning due to better detail of the calvarium and 

intracranial associated anomalies, such as hydrocephalus, 

associated cerebral malformations or venous drainage anomalies. 

Differential diagnosis has to be made with secondary 

craniosynostosis and positional flattening. Authors present a 

retrospective review of 185 children with suspected craniosynostosis 

seen over the last 13 years. 

FINDINGS 

They were studied with low-dose spiral CT with posterior 

reconstructed 3D and MIP images. MR imaging was performed 

when intracranial malformations were suspected, 

particularly in syndromic or multisutural synostosis. 

MDCT versus SDCT in Craniosynostosis 

Advantages: Improved temporal/spatial resolution, greater anatomical 

coverage, higher quality reconstrutions 

Decreased need for sedation from 30% (SD) to less than 

5% (MD) 

Disadvantages: Protocols need to be adapted from adult to minimize the 

radiation burden of children: Lowest MAs and Kv 

settings, appropriate section thickness, and faster table 

speed. 

Acceptable image quality can be achieved by reducing 

the scanning technique parameters, reducing radiation, 

although may increase the noise. 

CONCLUSION 

Particular emphasis is made in the CT imaging protocol used 

in our pediatric institution as well as in the differential diagnosis 

with nonsurgical causes of calvarial deformity, mainly 

with the positional plagiocephaly. 

KEY WORDS: Craniofacial malformation, craniosynostosis, 

children 


Teratomas of the Central Nervous System 

Grimme, J. D. · Camacho, D. L. A. · Spampinato, M. · 

Castillo, M. 



PURPOSE 

To review the imaging findings of teratomas arising in the 

brain, the head and neck, and spine. 


We reviewed our teaching file and chose representative 

examples of teratomas in the head and neck and central nervous 

system. Approximately 20 cases were found from those 

archived during a 10-year period. MR, CT, radiographic and 

sonographic images were chosen. A review of the literature 

was performed also to supplement the gathered material. 

FINDINGS 

Teratomas are germ cell tumors that develop from embryonic 

cells which become “misinvolved” during the formation of the 

primitive streak in the third week of gestation. Some of these 

436 

cells eventually become “misenfolded” as intracranial rests of 

tissue. Teratomas contain all three germinal layers and can be 

composed of multiple tissues of varying maturity, foreign to 

the site from which they arise. Intracranial teratomas are relatively 

rare, representing approximately 0.5-2.0% of intracranial 

tumors. Although they have been described in multiple 

areas of the brain, they are typically midline and common in 

the pineal and suprasellar regions. Most of our intracranial teratomas 

arose in the pineal or suprasellar regions but occurred 

in other locations as well. Teratomas of the head and neck 

comprise 9% of head and neck tumors in children. They commonly 

occur in the anterior midline, usually in the oropharynx 

or nasopharynx, but may involve also the orbit, temporal 

fossa, and face. Some teratomas, especially those arising in the 

nasopharynx, may traverse the skull base and have extensive 

intracranial extension. In our cases head and neck teratomas 

were more common anteriorly and low in the neck. In children 

the sacrococcygeal region is the most common location for 

teratomas; however, in adults they most often occur in the thoracolumbar 

region. Spinal teratomas can be intra or 

extramedullary. The most consistent features for all teratomas 

were an inhomogeneous appearance and presence of fat. 

Calcifications and enhancing soft tissues were less common. 

CONCLUSION 

Teratomas involving the head and neck and central nervous 

system are rare tumors, nearly always having a heterogeneous 

appearance and containing fat. 

KEY WORDS: Teratoma, central nervous system, head and 

neck 


Postoperative Temporal Bone 

Go, J. L. · Wallace, E. · Becker, T. 

University of Southern California Los Angeles 


PURPOSE 

The purpose of this exhibit is to describe common approaches 

for surgery of the temporal bone, their purpose, and imaging 

features that will aid the attendee in understanding the 

appearance of the postoperative temporal bone. 


The head and neck teaching file data base at a major academic 

institution on the West Coast was reviewed and correlated 

with the clinical history and operative reports. The 

postoperative CT scans were reviewed and utilized for this 

exhibit. 

FINDINGS 

The purpose of this exhibit is to acquaint the attendee with 

the approaches utilized by head and neck surgeons in operating 

on the temporal bone. Various approaches related to 

inflammatory disease including myringotomy, tympanoplasty, 

and mastoidectomy will be discussed. In addition, 

approaches for neoplasms of the cerebellopontine angle cistern 

and internal auditory canal, will be described also such 

as the retrosigmoid, translabyrithine, and middle cranial 

fossa approaches to the temporal bone. Imaging correlates of 

all of these procedures will be provided.

CONCLUSION 

For the uninitiated, evaluating the postoperative temporal 

bone may be a daunting task. By reviewing the various surgical 

approaches to the temporal bone and providing imaging 

correlates with CT, it is the goal of this exhibit to educate 

the attendee in interpreting these imaging studies. 

KEY WORDS: Postoperative, temporal bone 


3D CT of the Temporal Bone 

Fatterpekar, G. M. · Delman, B. N. · Naidich, T. P. · Lee, B. 

C. · Som, P. M. 

The Mount Sinai School of Medicine of New York University 

New York, NY 

PURPOSE 

This exhibit illustrates the role of 3D CT for a) displaying 

the normal anatomy of the temporal bone, b) evaluating temporal 

bone pathology, and c) planning/follow up of surgical 

procedures. 


High-resolution CT scans of the temporal bone were obtained 

in the axial (inferior orbito-meatal) plane in five normal subjects 

using a 16-slice spiral CT (Somatom Sensation 16, 

Siemens Medical Solutions, Malvern, PA), 120 kv, 200 mAs 

and overlapping 1 mm thick sections reconstructed at 0.7 mm 

intervals. The axial data sets were postprocessed on the Vital 

Images, Vitrea 2 (Plymouth, MN) to provide both 3D projections 

and special oblique planes designed to display specific 

structural relationships. Cases from the teaching files of the 

authors institution were selected to highlight the role played 

by 3D CT for evaluating temporal bone pathology. 

FINDINGS 

Three-dimensional CT successfully depicts the normal 

anatomy of the temporal bone. Reconstructing the anatomy 

along selected oblique planes helps to display the entire 

course of the facial nerve through the temporal bone and of 

the vestibular and cochlear nerves from the internal acoustic 

meatus into the vestibule and cochlea. Generating 3D anatomy 

in circumferential “rotational” “360º” views is especially 

beneficial for displaying the ossicular chain and the complex 

relationships among the structures of the inner ear. The 

ability of the computer to specify the same point on both the 

rotational and the oblique displays simultaneously helps to 

assess the integrity of the ossicles in cases such as fracture 

and cholesteatoma and to evaluate placement of ossicular 

prostheses and cochlear implants. 

CONCLUSION 

Three-dimensional CT with its multiprojectional display 

simplifies understanding of the complex anatomy of the temporal 

bone, and helps in both the preoperative planning and 

postoperative assessment of temporal bone pathology. 

KEY WORDS: Temporal bone, 3D CT, anatomy 

437 


Radiologic Evaluation of the Cochlear Implant 

Candidates 

Ozgen, B. · Sarac, S. · Sennaroglu, L. · Saatci, I. · Cila, A. 

Hacettepe University 


PURPOSE 

This exhibit will review the important aspects of imaging in 

the evaluation of the patients before cochlear implantation. 

The different types and components of the cochlear implant 

devices; standard surgical techniques also will be described 

briefly to familiarize the reader with this treatment method. 


The conventional radiographs, CT and MR imaging of the 

temporal bones, before and after implantation, will be used 

to illustrate the pertinent radiologic findings. 

FINDINGS 

Preoperative assessment should detect entities that may 

increase the degree of surgical difficulty such as mastoid sclerosis, 

labyrinthine ossificans, or high riding jugular bulb. The 

ear to be implanted should be free of infection. Cochlea should 

be assessed carefully for malformation, as the results of the 

implantation may be variable according to the severity of the 

dysplasia and there may be an increased risk of CSF gusher. 

The potential sites of complication such as aberrant facial 

nerve or carotid artery, large emissary veins also should be 

noticed and conveyed to the referring physician. 

Contraindications for the implantation including cochlear aplasia 

and absence of the cochleashould be assessed rigorously. 

CONCLUSION 

For the evaluation of the patients before cochlear implantation, 

the radiologists should be familiar with the surgical 

techniques and be able to identify clinically and surgically 

relevant findings that may contraindicate the implantation or 

alter the surgical methods. 

KEY WORDS: Cochlear implant, temporal bone 


26 


Thermo-Reversible Liquid Embolic Agents for the 

Treatment of Aneurysms 

Takao, H. 1 · Murayama, Y. 1 · Saguchi, T. 1 · Ishibashi, T. 1 · 

Ebara, M. 1 · Irie, K. 1 · Abe, T. 1 · Mori, Y. 2 · Vinuela, F. 3 

1Jikei University School of Medicine, Tokyo, JAPAN, 

2 3 Waseda University, Tokyo, JAPAN, University of 

California Los Angeles, Los Angeles, CA 

PURPOSE 

We developed a new embolic agent, thermo-reversible gelation 

polymer (TGP). This polymer’s own unique characteristic 

is that it can be solidified by body temperature. We eval- 



uated angiographic and histopathologic findings of this new 

liquid embolic agent for the treatment of experimental 

aneurysms. 


Thermo-reversible gelation polymer has a sol-gel transition 

temperature (TT). It becomes a liquid state at a temperature 

lower than the sol-gel transition temperature and becomes 

gel state above TT. It also can slowly deliver biologically 

active substance such as growth factor or engineered cells. 

The polymer was mixed with radiopaque material without 

solvent. Sixteen lateral aneurysms were surgically constructed 

on 16 common carotid arteries of swine. Three aneurysms 

were embolized with TGP without any protection device. 

Thermo-reversible gelation polymer was used also in combination 

with microcoils (n = 3), microstent (n = 3) and balloon 

protection across the neck of aneurysms (n = 3). One 

aneurysm was not treated as a control. Angiographic follow 

up was performed at day 14. Subsequently animals were sacrificed. 

Histopathologic evaluation was performed. 

FINDINGS 

All aneurysms were embolized successfully by TGP with 

protection devices (stent coils, and balloon). Distal TGP 

migration was observed when the aneurysms were 

embolized without protection. However these migrated TGP 

were dissolved successfully by the injection of cold saline. 

Histopathologic findings demonstrated neoendothelialization 

across the neck of the aneurysms. 

CONCLUSION 

This TGP can embolized an experimental aneurysm with 

protection devices. Further modifications such as mechanical 

stability or drug delivery system will be necessary prior 

to clinical application. 

KEY WORDS: Thermo-reversible gelation polymer, 

aneurysm, liquid embolic agent 

Pediatrics 

27-29 


“Bright and White” Deep Gray Matter Nuclei 

Kanekar, S. G. · Smoker, W. R. K. · Moritani, T. · Lee, H. · 

Kademian, J. C. 

University of Iowa 

Iowa City, IA 

PURPOSE 

To review the clinical and imaging features of deep gray 

matter nuclei pathology on MR imaging and CT. 


One hundred and fourteen patients with signs and symptoms 

of deep gray matter nuclei (DGMN) abnormalities (including 

irritability, lethargy, seizures, and dystonia) were evaluated 

by MR imaging and/or CT. Seventy of these patients 

had positive imaging findings. Patient age ranged between 1 

day and 70 years. MR imaging of the brain consisted of rou- 

438 

tine sequences (axial T1, FSE T2, FLAIR, and T1 sag followed 

by postcontrast T1 axial, sagittal, and coronal). 

Depending on the clinical presentation, these pathologies 

were classified into either acute or chronic. 

FINDINGS 

Acute hypoxia was the leading etiology among acute abnormalities, 

followed by hypoglycemia. Other entities producing 

acute changes in DGMN density/signal intensity were 

viral infections (ADEM, CJD, Japanese and West Nile 

encephalitis), poisoning and toxicity (carbon monoxide, kernicterus), 

extrapontine myelinolysis, and shaken baby syndrome, 

as well as hypotensive, arterial and venous infarction. 

Among the chronic causes, inherited metabolic diseases 

dominated the list. For better understanding, these diseases 

were further classified as per their enzymatic abnormality. 

For example: Glycolysis defect—Leigh’s disease; Fatty acid 

defects—glutaricaciduria and methylmalonic aciduria; 

Amino acid defect—MSUD; Urea cycle defect—citrullinemia; 

Lysosomal storage disease—Fabry’s, Krabbe’s, gangliosidosis, 

and metachromatic leukodystrophy; miscellaneous: 

Wilson, mitochondrial and Fahr. Other pathologies 

causing signal changes in DGMN were: Endocrine causes 

(hypo and hyperparathyroidism); neurofibromatosis type I, 

tumors (bilateral thalamic glioma, gliomatosis cerebri), 

infection, leukodystrophies and miscellaneous conditions 

(Halloverden Spatz disease, manganese deposition in TPN). 

CONCLUSION 

1) MR imaging and CT were very useful for diagnosis and 

treatment planning of acute abnormalities. 2) Although most 

chronic abnormalities were identified easily on MR imaging, 

especially inherited metabolic diseases, the final confirmation 

of the defect always required biochemical studies, since many 

of the inherited abnormalities had overlapping imaging findings. 

3) Knowledge of these various acute and chronic conditions 

producing DGMN signal abnormalities significantly aids 

in limiting the differential diagnosis when “Bright White” 

DGMN are encountered on cross-sectional imaging. 

KEY WORDS: Deep gray matter nuclei, MR imaging, bright 

signal intensity 


Tumors and Tumor-Like Lesions of the Pineal Region: 

MR Imaging Findings 

Korogi, Y. 

University of Occupational and Environmental Health 

Kitakyushu, JAPAN 

PURPOSE 

The pineal region is a heterogeneous area that includes the 

pineal gland and several parapineal structures. Pineal gland 

neoplasms are relatively uncommon, while pineal cysts are 

seen frequently as incidental findings. The purpose of this 

educational exhibit is to review the normal anatomy and provide 

an overview of the neoplastic and nonneoplastic lesions 

affecting the pineal region.


From a retrospective review of more than 100 patients with 

pineal region mass examined by MR imaging, clinical and 

MR imaging features of the lesions are presented. The list of 

pineal region masses includes germ cell tumors, pineal 

parenchymal cell tumors, glioma (astrocytic tumors, ependymoma, 

and other glial tumors), meningioma, metastases, and 

nonneoplastic masses such as pineal cysts, lipoma, epidermoid, 

cavernous hemangioma, and vascular malformations. 

FINDINGS 

MR imaging has allowed a marked improvement in the preoperative 

delineation of benign and malignant pineal masses 

and in distinguishing true pineal masses from parapineal 

masses impinging into the region of the gland. The solid 

components of the germ cell tumors are iso- to slightly 

hypointense relative to the gray matter on T1-weighted 

imaging, and mixed iso- and hyperintense on T2-weighted 

imaging. In particular, germinomas are usually mildly 

hypointense on T1-weighted imaging and mildly hyperintense 

on T2-weighted imaging, and may be isointense on 

both pulse sequences. Teratomas of the pineal are strikingly 

heterogeneous. After IV injection of Gd-DTPA, almost all 

lesions are enhanced markedly and heterogeneously. Unlike 

the previous contrast CT and MR reports, which showed 

homogeneous enhancement, the heterogeneous enhancement 

is a common finding. With high-resolution MR imaging, 

cystic components are seen more frequently than reported 

before; about 50% of germinomas and 90% of other 

GCTs in our series have cystic components. MR findings of 

other tumors including rare pathologies will be demonstrated 

also. 

CONCLUSION 

MR imaging is useful for delineation of benign and malignant 

pineal masses and for distinguishing true pineal masses 

from parapineal masses. MR findings are sometimes nonspecific, 

and age, sex, tumor markers as well as presence of 

cystic components are important for differential diagnosis. 

KEY WORDS: Pineal region tumor, germ cell tumor, MR 

imaging 


Imaging of “Treated” Hydrocephalus: Documentation of 

Function and Malfunction 

Patton, A. · Chao, C. P. 

Mayo Clinic 

Jacksonville, FL 

PURPOSE 

To describe the systematic evaluation of the adult and pediatric 

population who have been treated for hydrocephalus. In 

addition, the imaging appearance of the various causes of 

shunt malfunction will be illustrated. 


Hydrocephalus is defined as progressive increase in cerebrospinal 

fluid within the ventricular system. This is usually 

secondary to obstruction in the ventricular cerebrospinal 

fluid pathway, such as a mass lesion and less commonly due 

to cerebrospinal fluid overproduction. Red blood cells, white 

439 

blood cells, and neoplastic cells may compromise the flow of 

cerebrospinal fluid in the arachnoid granulation and dural 

sinuses. The treatment for hydrocephalus is shunt insertion. 

Shunt options include: redirection of cerebrospinal fluid 

from the ventricular system into the basal cistern (third ventriculostomy), 

peritoneum, atrium, pleura and into the gallbladder. 

Headache and vomiting are common nonspecific 

symptoms; however, may represent clinical manifestation of 

shunt malfunction. A retrospective review of these patients 

was performed. 

FINDINGS 

The following pathologic processes were identified in a population 

with “treated” hydrocephalus: distal shunt tubing 

within the peritoneal cavity, pleural cavity, right atrium, gallbladder, 

shunt disconnection, mal positioning of the distal 

shunt tubing, atonic gallbladder, encysted intraperitoneal 

cerebrospinal fluid, “malfunction” of the third ventriculostomy 

and spontaneous intracranial hypotension. Shunt series 

and contrast (iodine or technetium DTPA) studies may be 

used to evaluate for possible fracture, disconnection, and 

patency of the shunt system. Sonography is helpful in the 

assessment of some of these unusual shunts, from the ventricular 

system into the pleural, peritoneal cavity, and gallbladder. 

Furthermore, the diagnosis and treatment of 

intraperitoneal encysted cerebrospinal fluid collection at the 

distal shunt tubing can be achieved with ultrasound. The 

nonspecific finding of dural enhancement may be seen in 

patients with history of trauma, infection, inflammation, 

metastatic neoplasm, or intracranial hypotension. Cine MR 

imaging exquisitely demonstrates the patency of the third 

ventriculostomy. 

CONCLUSION 

The diagnosis of shunt malfunction is best defined by the 

change in the cerebrospinal fluid volume from the individual’s 

postsurgical asymptomatic “baseline” CT or MR 

images. Conventional radiographs of the shunt, contrast 

study of shunt system, CT, MR imaging including cine MR 

imaging play an important role in the evaluation of a nonfunctioning 

shunt system. Prompt diagnosis of shunt disruption 

and or associated complications may decrease morbidity 

in these patients. 

KEY WORDS: Hydrocephalus, shunt malfunction, imaging 



Spine 

30-31 


The Keck School of Medicine of the University of 

Southern California Comprehensive Spinal Imaging 

Course: Part 2 

Rothman, S. L. G. 


Los Angeles 


PURPOSE 

We live in a period of super specialization in radiology. 

Radiology residents and fellows migrate from section to section 

learning MR imaging, CT, and invasive neuroradiology. 

It is unusual for there to be an opportunity for the trainee to 

be in the position of seeing the entire picture of spine disease. 

Furthermore, the radiologist tends to be isolated from 

the clinicians who care for the patients. During the last 2 

years a comprehensive 10-hour, interactive, course in diagnostic 

imaging of the spine has been created including plain 

X-ray, videoflouroscopy, CT and MR imaging. The course is 

designed to present the spine in its totality, as an entity with 

three functions, supporting the head and body, allowing 

mobility, and protecting the spinal cord. The purpose of this 

computer exhibit is to demonstrate the second segment of the 

interactive CD-based course in spine radiology dealing with 

the anatomy, pathophysiology, and imaging characteristics 

of the intervertebral disk. It is designed to teach basic concepts 

rather than the details of rare entities. Symptom correlation 

with observed abnormalities is stressed. 


A PowerPoint lecture will be read from CDs. Imbedded in 

these presentations is a combination of text and voice used to 

challenge the trainee’s powers of observation and deduction. 

There are many quiz questions about the images pervading 

the lectures. 

FINDINGS 

A portion of the University of Southern California radiology 

course will teach the radiographic anatomy and physiology 

in a most unusual way. 

CONCLUSION 

The course in its entirety provides a novel approach for 

learning spinal imaging. 

KEY WORDS: Spine, anatomy, intervertebral disk 

440 


Comprehensive Diagnostic Evaluation of Cervical Spine 

Trauma: Part 2 - Upper Cervical Spine 

Gentry, L. R. 1 · Smoker, W. R. K. 2 · Reede, D. L. 3 · Pulfer, K. A. 1 

1 2 University of Wisconsin, Madison, WI, University of Iowa, 

Iowa City, IA, 3Long Island College Hospital, New York, NY 

PURPOSE 

This scientific exhibit reviews the current state of arts for 

diagnostic imaging of upper cervical spine trauma using 

multiple imaging modalities. The mechanisms of injury, typical 

patterns of fracture, and common complications of cervical 

spine fractures are reviewed. The clinical features of 

trauma patients that dictate which should be imaged and 

what imaging modality should be used are discussed. An 

algorithmic approach is presented for guiding the diagnostic 

evaluation. Diagnostic pitfalls are presented. The results of 

this study are presented in an interactive multimedia format 

(audio, video). 


The diagnostic images (conventional radiographs, CT, 3D 

CT, MR imaging, MR angiography, CT angiography, flexion-extension 

radiographs) of 174 consecutive patients with 

cervical spine fractures were reviewed retrospectively and 

the diagnostic studies analyzed. 

FINDINGS 

A wide array of fracture mechanisms was identified (hyperflexion, 

hyper-extension, lateral hyperflexion, axial loading, 

shearing, complex). The major complications that were 

observed included (cord contusion-transection, vertebral 

artery injury, intraspinal hematoma, spinal cord hematoma, 

dural tear, nerve root and brachial plexus avulsion, traumatic 

disk herniation, ligamentous injury, cervical dislocation). 

CONCLUSION 

Multiple imaging modalities often are required to identify 

and classify cervical fractures and to assess for the myriad of 

complications resulting from these fractures. The use of MR 

imaging and occasionally flexion-extension radiographs are 

essential for assessment of ligamentous injury that may be 

missed on imaging. Multislice thin-section CT is of paramount 

importance for identification of fractures and 3D CT 

may be helpful for surgical planning. MR angiography and 

CT angiography are essential for assessing for potential vascular 

injury. 

KEY WORDS: Trauma, fractures, spine

Stand Alone 

32-41 


Reference Atlas of Cerebral Vasculature 

Nowinski, W. L. 1 · Volkau, I. 1 · Thirunavuukarasuu, A. 1 · 

Baimouratov, R. 1 · Hu, Q. 1 · Luo, S. 1 · Huang, S. 1 · Runge, 

V. M. 2 

1 2 Bioinformatics Institute, Singapore, SINGAPORE, Scott 

and White Memorial Hospital, Temple, TX 

PURPOSE 

A digital reference atlas of cerebral vasculature is not available. 

At ASNR 2004 we presented our initial vascular model, 

which did not contain veins and the number of arteries was 

limited (1). The objective of this work is to develop such a 

reference cerebrovascular atlas for clinical, research, and 

educational purposes. 


High quality data were acquired on 1.5 T Magnetom Sonata 

Maestro Class scanner using an advanced gradient system 

and 8-channel receive head coil. MRA and MRV 3D TOF 

scans were acquired for the whole brain with 1.0 mm slice 

thickness. The scans were enhanced and the centerlines and 

radii of the vessels extracted automatically. The centerlines 

were smoothed and the branching points determined. A modeling 

technique was developed to construct smooth and efficient 

surfaces from the centerlines and radii. Small vessels 

were extracted interactively by means of a dedicated vascular 

editor. The vascular segments were labeled with names 

based on Terminologia Anatomica (2). 

FINDINGS 

The proposed methods enabled the construction of the cerebrovascular 

model and develop a user-friendly, interactive 

application. The user can rotate, zoom, and pan the model. 

By pointing to any vessel in three dimensions from any point 

of view, its name is given. In addition, any vascular submodel 

can be constructed by selecting its component vessels 

interactively from the vascular index. 

CONCLUSION 

A new MR imaging-based reference cerebrovascular atlas is 

constructed. It is useful to study and rapidly explore cerebral 

vasculature in three dimensions. Moreover, the tool is helpful 

in generating teaching materials. 

REFERENCES 

1. Nowinski WL, Thirunavuukarasuu A, Baimouratov R, Volkov I, 

Hu Q, Aziz A, Huang S. Three-dimensional atlas of brain 

anatomy and vasculature. Proceedings, American Society of 

Neuroradiology ASNR 2004, 42nd Annual Meeting, Seattle, 

WA, 5-11 June 2004;455-456 

2. Terminologia Anatomica, International Anatomical 

Terminology. Thieme, Stuttgart - New York, 1998 

KEY WORDS: Adult brain (vascular intracranial), electronic 

brain atlas, vascular atlas 

441 


Web-Based Real-Time 3D Visualization of 

Neuropathologic Entities on MR Angiography 

Su, A. · Knopp, E. 

New York University 

New York, NY 

PURPOSE 

Three-dimensional (3D) visualization of neurovascular MR 

angiography (MRA) is performed commonly by rendering 

maximum intensity projection (MIP) images from the source 

volumetric data set. However, MIP images may be unable to 

show vessels pathways in complex anatomy because the 

images usually are rendered at a predefined angle interval 

about a predetermined rotation axis. This limitation is compounded 

further by the loss of depth information inherent on 

MIP imaging. There are commercial workstations available 

that perform real-time 3D visualization (RT3DV) and render 

MIP images about any desired rotation axis and angle. In 

addition, other methods of evaluating volumetric data, 

including direct volume rendering (DVR) in which all voxels 

of a volume are considered for visualization, are available. 

However, referring surgeons or neurointerventionalists 

often do not have access to a RT3DV workstation to allow 

collaborative visualization. With the widespread availability 

of web access and rapid increase in computing power of consumer 

PCs, the authors will demonstrate the value of webbased 

RT3DV in postprocessing source MRA volumetric 

data sets. 


Cases highlighting various neuroradiologic pathologies evaluated 

using MRA were collected. The source MRA data set 

was transferred to a workstation that served as a DICOM 

receiver (Java DICOM Toolkit, Pixelmed) and a web server 

(HTTP Server, Apache Software Foundation). After logging 

into the web server from a remote client computer available 

commercially (Pentium IV, Dell), a user selected the desired 

MRA examination, which subsequently launched a custom 

programmed, web-based Java DICOM viewer onto the 

client’s computer. The authors integrated MIP and DVR 

functions into the DICOM viewer with the main goal of 

achieving interactive frame rates so that RT3DV could be 

performed about any axis and angle. This was achieved with 

a set of optimizations and efficient techniques, including 

data preprocessing, efficient data representation, and rapid 

rendering using the fast shear/warp-based method. 

FINDINGS 

Cases demonstrated include: 1) aneurysm; 2) tumors; 3) 

AVM; 4) developmental variants; 5) embolic stroke; 6) atherosclerosis; 

7) moyamoya. 

CONCLUSION 

By creating a web-based solution, RT3DV of MRA volumetric 

data sets can be performed independent of the availability 

of a workstation. This particularly benefits referring 

surgeons and neurointerventionalists. For the neuroradiologist, 

RT3DV of the source MRA data set about any axis and 

angle increases spatial perception of complex anatomy. 

KEY WORDS: Web-based, 3D visualization, real-time 



442 



MR Spectroscopic Imaging Software for the The Anatomy Wiz: An Interactive Internet Tutorial and 

Neuroradiologist 

Reference Tool 

Bonekamp, D. · Barker, P. B. 

The Johns Hopkins University 

Baltimore, MD 

PURPOSE 

The routine clinical application of brain MR spectroscopy 

(MRS), and particularly MR spectroscopic imaging (MRSI), 

is hampered by the lack of suitable offline processing tools 

for the processing, display and analysis of spectroscopy data. 

The purpose of this study was, therefore, to develop a userfriendly, 

interactive software package for the processing of 

both single- and multivoxel MRSI data, that will run on 

commonly available computer systems. 


The program dsx was developed in C++ on an Intel Pentium 

system running Redhat LINUX version 8.0, and subsequently 

cross-compiled for Microsoft Windows XP. The graphical 

user interface was implemented using the GTK/GDK toolkit. 

All standard spectroscopy processing, analysis, and display 

functions are implemented, with support for singlevoxel 

as well as 1D, 2D, multislice 2D, and 3D MRSI data 

from multiple MR vendors. The package also includes an 

image viewer function for display of metabolic and localizer 

MR images. 

FINDINGS 

The dsx program is a user-friendly interactive software package 

for processing MRS and MRSI data. In addition to its 

function as an offline MRSI processing tool, dsx also may be 

directly installed on MR scanners for real-time, online processing 

(for scanners based on the LINUX or Windows-XP 

operating systems). The software also may be used on 

portable laptop computers. Because of its interactive nature, 

dsx is helpful as a spectroscopy teaching tool, demonstrating 

the steps required to process MRS and MRSI data. For 

instance, dsx allows the direct visualization of the effects of 

digital filtration, voxel shifting, baseline and phase corrections 

on spectroscopy data. 

CONCLUSION 

The electronic scientific exhibit will allow a hands-on 

demonstration of this software on sample data sets from the 

human brain, including single voxel MRS and multislice 

MRSI data from brain tumors. 

KEY WORDS: Spectroscopy, spectroscopic imaging, software 

Muro, G. J. · Shen, C. 

Bridgeport Hospital 

Bridgeport, CT 

PURPOSE 

This is an Internet version of an existing anatomy program 

designed by the author. The existing program consists of several 

unique functions. The goal of this project is to make the 

program widely available over the Internet. 


The existing program consists of several modules. Unique 

features of the program include the ability to scroll through 

a set of images and select anatomical structures by clicking 

on them. The structure is highlighted and remains so on consecutive 

images. A structure can be selected also from a sorted 

list. There is also accompanying text, references, and supplemental 

images. Each module is created from an authoring 

program. The authoring program allows anyone to create a 

teaching module of his or her choice which is a powerful 

learning experience in itself. 

FINDINGS 

The Internet version is designed using the .NET Framework. 

The application is scalable allowing future modules to be 

added. The appearance and functionality of the Internet version 

is preserved. 

CONCLUSION 

The result is a powerful interactive anatomy teaching tool 

and reference tool available to anyone with Internet access. 

KEY WORDS: Anatomy, tutorial, reference 


Virtual MR Endoscopy of the Middle and Inner Ear 

Lane, J. I. · Camp, J. · Witte, R. J. · Driscoll, C. L. · Robb, R. 

Mayo Clinic 

Rochester, MN 

PURPOSE 

An understanding of the three-dimensional relationships of 

anatomical structures that make up the middle and inner ear 

is critical for radiologists and otolaryngologists to adequately 

interpret imaging studies in the clinical environment. 

Three-dimensional reconstructions of the temporal bone 

from clinical imaging studies suffer from limited resolution 

and relatively large fields of view. 


We have taken advantage of existing technology employed 

in the field of MR microscopy to construct a computer-based 

learning module of middle and inner ear anatomy. Images of 

a cadaver temporal bone specimen scanned at 9 T after

opacification of the middle ear space with dilute gadolinium 

solution were used to construct the three-dimensional environment. 

FINDINGS 

Imaging dataset achieved a voxel size of 78 microns. Image 

data were color-segmented to allow better definition of the 

structures of the middle and inner ear particularly in endoscopic 

displays. 

CONCLUSION 

We have emphasized the spatial orientation of clinically relevant 

anatomical structures within the middle and inner ear 

that previously have been seen only with standard microscopic 

techniques. 

KEY WORDS: MR microscopy, temporal bone, endoscopy 


High-Fidelity Simulation of Interventional 

Neuroradiology Procedures 

Cotin, S. 1 · Luboz, V. 1 · Pegoraro, V. 2 · Neumann, P. F. 1 · Wu, 

X. 1 · Dawson, S. L. 1 

1 Simulation Group at the Center for the Integration of 

Medicine and Innovative Technology, Boston, MA, 

2 University of Utah, Salt Lake City, UT 

PURPOSE 

In August 2004, the FDA mandated that physicians performing 

carotid stenting must complete a multistep educational 

program including simulator-based training to proficiency, 

regardless of physician specialty. The unacceptable risks of 

learning on patients have placed a new emphasis on creating 

realistic and accurate simulations which can meet the FDA’s 

mandate. The development of a realistic computer-based 

interventional neuroradiology simulator would provide procedural 

and skill training in an integrated educational curriculum 

permitting a safer and repeatable learning without 

putting patients at risk. 


Our prototype interventional neuroradiology procedural simulation 

system utilizes new approaches in synthetic fluoroscopy, 

virtual catheter/guidewire modeling, contrast agent 

propagation, fluid dynamics and interface design. Synthetic 

fluoroscopy uses a volumetric approach which directly 

incorporates a CT dataset while cerebrovascular anatomical 

models are segmented directly from a CTA dataset. Realtime 

catheter/guidewire models are based upon incremental 

linear Finite Element Method with integrated collision 

detection, permitting multiple contacts and sliding. Stents 

are implemented as a mass-spring system which responds to 

internal forces and collides with vessel boundaries. Laminar 

blood flow model is computed through a simplified Navier- 

Stokes equation of the cerebrovascular network while contrast 

agent propagation is controlled by a first-order partial 

differential advection equation. Interactivity is provided 

through the use of actual catheters and guidewires tracked 

through an interface that can be embedded within a patient 

mannequin. Designed with an educational curriculum of 

case scenarios, the current system delivers high fidelity simulation 

on a cost-effective platform. 

443 

FINDINGS 

The integrated real-time procedure simulation system has 

embedded new approaches in fluoroscopic rendering, 

catheter/guidewire motion modeling, contrast agent propagation 

modeling, and 3D anatomical representation. This 

system is implemented on a modern PC workstation with 

commercially available components to enhance cost-effectiveness. 

Each computationally intense component is scalable 

such that increasing the resolution of each component 

requires linearly more time to compute. To assess face and 

content validity and evaluate our system for future research 

directions, we will demonstrate the system on-site and survey 

neuroradiology experts. Users will be able to manipulate 

directly real catheters and guidewires into a patient mannequin, 

monitor a dynamic internal fluoroscopic view, perform 

diagnostic injections of contrast agent, and place virtual 

stents. We will present a questionnaire survey to each user 

and collect their feedback. Our survey is based on a modified 

Cooper-Harper characteristic scale as used by pilots to evaluate 

flight simulators. This ten-point scale is a subjective 

measurement which asks a set of refined questions to determine 

whether the components of simulator are unsatisfactory, 

need improvement, or are sufficient. The on-site survey is 

critical for us to improve the computational resource distribution 

and further optimize this procedural and skill training 

simulator. 

CONCLUSION 

In summary, a set of interventional radiology procedure simulation 

components have been developed and integrated into 

a procedural simulation and skill training platform for the 

treatment of stroke. This prototype emphasizes high fidelity 

visual feedback and physically accurate instrument motion 

and contrast agent propagation under virtual fluoroscopic 

imaging. Because of its cost-effective design and system 

compactness, such a system would facilitate interventional 

training. 

KEY WORDS: Simulation, computation, interventional radiology 




Interactive Atlas of Pediatric Brain CT/MR Imaging 

Kott, B. 1 · de Regt, D. 2 · Mayock, R. 3 · Weinberger, E. 1,3 

1University of Washington School of Medicine, Seattle, WA, 

2 3 Reed College, Portland, OR, Children’s Hospital and 

Regional Medical Center, Seattle, WA 

PURPOSE 

The large datasets that now comprise single MR examinations 

do not lend themselves easily to static representation in 

a teaching file. Moreover, with the advent of PACS, there has 

been a gradual switch to being able to view images in stack 

mode, where one scrolls through all images in a given series, 

rather than looking at each image as if printed on a page of 

film. To simulate this environment, we created a program 

that easily allows the importing of large datasets (i.e., all 

images in an MR examination), and has many of the linking 

and scrolling options offered in current PACS systems. 

Selected MR and CT cases of the pediatric brain were used 

to populate our database, to demonstrate how the program 

can be used as an atlas, a self-taught learning module, or for 

teaching case presentation. 


The atlas comprises over 10,000 images derived from brain 

CT and MR examinations of pediatric patients. Viewing of 

the images is controlled by a program written in C# and runs 

under Microsoft Windows with the .NET Framework 

installed. 

FINDINGS 

The interactive atlas allows for easy simultaneous scrolling 

of different sequences in the same plane for the same patient 

as well as the same sequence for different patients. This 

allows for both a dedicated evaluation of a single patient as 

well as the ability to compare the same region over many different 

patients. Direct comparison of normal with abnormal 

can thus be made. A comment field for each image can be 

accessed selectively allowing one to view cases as knowns 

or unknowns. The software design allows the user to further 

customize the database by easily removing or adding additional 

images to the provided data set. 

CONCLUSION 

We demonstrate an intuitive interactive digital atlas to facilitate 

learning about normal and abnormal MR appearance of 

the pediatric brain. The program allows for easy modification 

of the database to create other teaching modules. 

KEY WORDS: Brain 

444 


Congenital Disorders of the Sellar and Parasellar Region 

Spampinato, M. V. · Grimme, J. D. · Camacho, D. L. · 

Castillo, M. 



PURPOSE 

The pituitary gland derives from the combination of two 

events occurring during the fourth week of gestation. A 

diverticulum from the ectodermal lining of the stomodeum, 

called Rathke’s pouch, develops into the adenohypophysis. 

At the same time, the posterior lobe, functionally linked to 

the hypothalamus, originates from a neuroectodermal evagination 

of tissue from the floor of the diencephalon. Here we 

illustrate how congenital pathology of the sellar and parasellar 

regions may originate from abnormalities of these coordinated 

embryologic events. 


Our presentation will be a didactic electronic exhibit, consisting 

of a background section on the embryology of the 

pituitary gland, and a pictorial gallery, illustrating the imaging 

features of common and rare congenital disorders of the 

sellar and parasellar regions. 

FINDINGS 

We will present cases of congenital disorders of the sellar 

and parasellar region chosen from our teaching file. Selected 

MR and CT images will illustrate pathologic processes related 

to 1) embryogenesis of the adenohypophysis, including 

ectopic anterior pituitary gland, pituitary hypoplasia, persistence 

of the craniopharingeal canal, craniopharingioma, 

Rathke’s cleft cyst, pars intermedia cyst; 2) development of 

the hypothalamo-pituitary axis, such as ectopia of the neurohypophysis, 

optic-infundibular dysplasia, duplicated pituitary 

gland, and hamartoma of tuber cinereum; 3) disturbance 

of the separation of surface ectoderm and neurectoderm 

of the neural folds (transsphenoidal encephalocele and 

associated findings). 

CONCLUSION 

The teaching point of the exhibit is to demonstrate the imaging 

appearances of congenital diseases arising in the sellar 

and parasellar regions. Furthermore, to demonstrate how the 

knowledge of the anatomy of the region of the sella and of 

its embryologic correlations with the neighboring structures 

enables for the correct interpretation and diagnosis of congenital 

pituitary disorders. 

KEY WORDS: Congenital disorders, pituitary gland, MR 

imaging


The Pediatric Brain Wiz 

Muro, G. J. · Pan, J. 

Bridgeport Hospital 

Bridgeport, CT 

PURPOSE 

The variable MR appearance of the continually maturing 

pediatric brain can be daunting and sometimes problematic 

for radiologists reading them. The purpose of this project 

was to create a widely accessible database and image library 

of normal pediatric brain MR images for reference purposes. 


An authoring program was designed in the Visual Basic programming 

language. The authoring program gathers normal 

pediatric MR images in a DICOM format from a PACS 

workstation. The program organized the study information 

and images by age, gender, sequence, and plane. Each item 

from a list of important anatomical structures and regions 

then was matched to specific sequences and images for each 

study in the image library. 

FINDINGS 

Separate standalone and Internet-based programs were created 

to quickly and easily reference the normal pediatric brain 

MR image database and image library. The programs sort the 

images by anatomical structure/region, age, image sequence, 

and plane. For example, a radiologist concerned about the 

appearance of the corpus callosum on a sagittal T1-weighted 

series of a particular patient, quickly can generate an agematched 

group of sagittal T1-weighted images centered 

through the CC as a reference. The radiologist then can 

decide subjectively if the study is normal. The program is 

educational as well. The program contains linked discussions 

describing normal pediatric brain development. In 

addition, one can gain a better appreciation for the changes 

that occur with normal development by scrolling through 

series of images matched by level, plane, and sequence 

across a range of age-sorted normal MR images. 

CONCLUSION 

The Pediatric Brain Wiz is a very useful normal pediatric 

brain reference tool. It allows a very quick and easy agematched 

comparison to studies within a normal pediatric MR 

image database and image library. There are several educational 

features as well. 

KEY WORDS: Pediatric, brain, reference 

445 



Notes:

A 

A. Boukai, A. S-83 

Aagaard-Kientiz, B. O-19 

Abdalla, A. O-286 

Abdel Razek, A. A. O-75, O-102, O-103, 

O-286, P-43 

Abdu, W. A. O-308 

Abduljalil, A. P-53 

Abe, T. E-26, O-46, P-126, P-138 

Abrahams, J. J. I-136 

Abrams, K. J. O-375 

Abrantes, Y. O-214, O-377, O-384 

Acker, M. S-1 

Ackerman, A. S-18 

Ackerman, R. H. P-9 

Adachi, Y. P-85, S-13 

Adams, R. O-451 

Adzick, N. S. O-318 

Afchani, O. P-95 

Affel, M. E. P-9 

Agid, R. O-419 

Ahmed, H. O-228 

Ahn, K-J. E-11, P-78 

Aho, T. S-67, S-69 

Ai, Sr., L. P-59 

Aiken, A. H. O-190 

Akalan, N. P-48 

Akbas, T. O-468 

Akmangit, I. O-40 

Al-Ali, F. O-160 

Albadrawey, A. O-102 

Albuquerque, F. C. O-151, O-208, O-211, 

O-212 

Alegret, M. O-243, O-244 

Alexander, A. L. O-229 

Alkan, Sr., A. O-476 

Allen, G. S. O-387 

Alley, M. T. E-8, O-379, O-383 

Almli, R. P-166 

Al-Nakshabandi, N. S-83, S-84 

Alonso, J. S-9 

Alorainy, I. A. S-75, S-77, S-83, S-84 

Alsop, D. O-239 

Altinok, D. O-453 

Altinok, G. O-453 

Altinok, Jr., M. T. O-476 

Alvarez, H. O-448, O-449 

Alvarez-Sabín, J. O-498 

Anand, A. E-2 

Anand, P. O-509 

Ances, B. M. O-378, P-8, P-29, P-77 

Anderson, D. O-38 

Anderson, J. C, O-111, O-283, O-307, 

O-354, P-55 

Anderson, P. O-471 

Andersson, T. O-177 

Ando, Y. S-12 

Angtuaco, E. J. C. O-468 

Anselmetti, G. O-427 

Antonietti, L. O-231, O-233 

Anzai, Y. O-248 

Aoyagi, Y. S-60 

Appignani, B. O-239 

Apuzzo, M. L. P-50 

Araki, T. P-85 

Araki, Y. P-69 

Arat, A. O-163 

Araujo, D. P-88 

Araujo, D. B. P-20 

Arends, J. P-68 

Arenillas, J. F. O-498 

Arthur, A. O-171 

Arya, V. O-34 

Asao, C. P-171 

Aschenbach, R. O-104 

Ashwal, S. P-170 

Atieh, J. M. P-148 

Aton, E. O-374 

Auger, Sr., C. E-21 

Aulino, J. M. O-81, O-387 

Avignone, S. O-174 

Aviv, R. I. O-143, O-296, O-508, S-5 

Ayaz, M. O-418 

Aymerich, X. O-369 

Aziz, A. O-437 

B 

447 

Index of Program Participants 

Numbers refer to session and presentation numbers, not to page numbers. 

Index Key: E = Electronic Scientific Exhibit; I = Invited Speaker; O = Oral Paper; P = Poster; S = Scientific Exhibit 

Babb, J. S. O-28, O-183, O-488, P-10 

Babic, D. O-177 

Backes, W. H. I-346, O-213, P-135 

Bae, H. O-433 

Baek, C-H. P-112 

Bahar, N. P-132 

Bahl, G. O-368 

Baik, J. H. S-61 

Bailey, L. L. P-151 

Baimouratov, R. E-32 

Bajwa, Z. E-12 

Bakaya, S. O-101 

Baker, J. P-141 

Baker, K. B. O-178, O-226, P-82 

Baleriaux, D. P-39 

Balkanci, F. O-163 

Ball, T. O-296 

Bammer, R. I-458, O-352 

Bapuraj, J. R. O-26, O-34 

Barbaro, N. M. O-422 

Bardo, D. M. E. S-78 

Barest, G. O-108, S-58 

Barger, A. O-77 

Barker, P. B. E-35, O-182, O-293 

Barkovich, A. J. I-262, P-162 

Barnard, J. J. S-73 

Barnes, P. D. P-149, P-153 

Barnwell, S. O-170, O-173 

Barr, J. D. I-135, O-171 

Barry, C. O-35 

Barry, K. S-27 

Barth, R. F. O-186 

Bartlett, E. S. O-56, O-57, O-58, O-59 

Bartolomei, L. O-326 

Bartynski, W. S. O-88, O-240, O-420 

Baruzzi, F. O-427 

Baryshnik, D. O-390 

Basche, S. O-104 

Basiratnia, R. P-63, P-66, P-109, P-110 

Basmaji, C. N. S-49 

Bateman, B. T. O-16, O-279, P-173 

Bates, D. S-8 

Batjer, H. H. O-423, O-505 

Batra, K. S-89 

Bauknecht, H-C. P-104 

Baumgarten, D. A. O-120 

Bavaria, J. S-1 

Baytion, M. P-2 

Beaton, R. O-311 

Beauchamp, N. E-2 

Beck, J. O-468 

Becker, H. O-287 

Becker, T. E-23 

Becker, W. P. O-119 

Begelman, K. S-58 

Behr-Ventura, D. S-18 

Bell, T. O-167 

Bello, J. A. O-80, O-421 

Bellon-Guardia, M. P-57 

Beltramello, A. O-428 

Ben Bashat, D. O-456 

Benndorf, G. P-72 

Ben-Sira, L. O-456 

Benzinger, T. L. S. P-166 

Berger, R. M. I-32, I-63 

Berghaus, A. P-104 

Berkowitz, E. A. P-172 

Berman, M. F. O-16, O-279 

Bermudez, P. S-16, S-31 

Bernardi, B. O-284, P-101, P-154 

Bernstein, M. O-77 

Bernstein, R. O-321 

Besenski, N. O-220, P-22 

Beversdorf, D. P-52 

BhanuPrakash, K. N. E-2 

Bhatia, R. G. O-395, P-178 

Bhattacharyya, P. O-327 

Bianchi, F. P-11 

Biega, T. J. P-38 

Biegel, J. A. O-497 

Bigler, E. O-440 

Bilal, M. O-75 

Bilaniuk, L. T. O-318, O-398, O-486 

Bilsky, M. H. O-472, S-85 

Biondi, A. O-97, P-124 

Birchall, D. S-8 

Bishopric, N. H. O-17 

Blackband, S. J. O-363 

Blackmore, C. C. O-277 

Blake, M. I-39, I-71 

Blaser, S. I. I-403, O-30, O-392, O-437, 

O-494, P-182 

Blechschmid, N. O-119 

Bluml, S. O-442, O-450, O-493 

Boada, M. O-301 

Boardman, J. F. O-240, O-420 

Bock, M. O-417 

Boesiger, P. O-229, O-353, P-103 

Bogomolny, D. O-237 

Bohner, Jr., G. O-413 

Program Participants

Program Participants 

Bohning, D. O-220 

Boland, P. S-85 

Bonaldi, G. O-427 

Bonekamp, D. E-35 

Bonetti, M. O-432 

Bonfante, E. P-74 

Bongioanni, P. P-11 

Booth, T. O-478 

Borg, B. O-327, P-82 

Borthakur, A. O-247 

Bose, A. O-118 

Bottomley, P. A. O-178 

Bourekas, E. C. P-147, P-148 

Bourgeois, D. O-487 

Bowen, B. C. I-68 

Boyle, J. O-105 

Bracard, S. P-124 

Bracco, S. E-1 

Bradbury, M. S. O-105 

Bradley, Jr., W. G. I-142, I-206, O-368 

Brady, P. S. P-7 

Branca, V. O-174 

Branson, H. P-182 

Branstetter, IV, B. F. I-15 , O-101, O-166, 

S-59 

Brant, L. J. O-182 

Breitwieser, C. O-413 

Breslau, J. I-70 

Brinkman, W. J. O-248 

Britz, G. O-319 

Broadbent, P. O-171 

Brockmann, M. A. O-365 

Bronov, O. O-89 

Brook, A. L. I-256 

Brooks, M. L. S-22, S-55 

Brown, P. P-55 

Brugger, P. C. O-290 

Bruhn, H. P-67 

Bryan, R. N. I-269, P-84 

Bub, L. D. O-277 

Budai, R. S-26 

Budisavljevic, M. P-22 

Bulakbasi, N. P-35, P-37 

Bullitt, E. P-36 

Burdette, J. H. O-221, O-222 

Burger, I. M. O-446, P-129 

Burgess, J. E. O-36 

Burn, J. S-8 

Burnet, N. G. O-492 

Burrows, P. E. I-65 

Burton, E. C. O-372 

Busa, E. P-167 

Bussiere, M. O-162 

Butman, J. A. O-445, P-38 

Butowski, N. O-231, O-233 

Butters, M. A. O-225 

Byrne, J. V. O-145 

Byun, H-S. P-112, S-61 

Byun, J-Y. E-11, P-78, P-89 

C 

Cacayorin, E. P-74 

Cadavid, D. O-324 

Calabresi, P. A. O-191 

448 

Calderwood, G. O-305 

Calhoun, V. O-223 

Callen, P. W. P-162 

Camacho, A. C. O-107 

Camacho, D. L. A. E-22, E-40, O-230, 

O-236, O-443, S-51 

Camargo, E. O-112 

Camp, J. E-37 

Campeau, N. G. O-400A 

Canapicchi, R. P-11 

Canive, J. M. P-94 

Canna, S. O-445 

Canyigit, M. O-163 

Carfrae, M. S-70 

Carlotti, C. G. P-20 

Carlson, D. O-78 

Carpegiani, P. O-427 

Carr, J. O-54, O-423 

Carrascoso Arranz, J. P-134 

Carr-Brindle, V. O-149 

Carriero, A. O-475 

Carroll, T. J. O-54, O-82, O-157, O-321, 

O-423, O-505 

Carvajal, A. S-16, S-31, S-76 

Casasco, A. O-97 

Case, R. S. O-385, S-3 

Cashen, T. O-54, O-157, O-321, O-423, 

O-505 

Castaneda-Guerrero, M. P-57 

Castillo, M. E-4, E-22, E-40, I-336, 

O-230, O-236, O-391, 

O-443, P-44, S-51 

Castro, J. D. V. P-20 

Castro, M. A. O-36 

Caulo, M. P-45, S-19 

Causin, F. O-326, S-26 

Cavedon, C. O-326, S-26 

Caviness, V. S. P-167 

Cebral, J. R. O-36 

Cekirge, H. S. O-40, O-163, O-484 

Celik, H. P-177 

Cerase, A. E-1, P-45, S-19 

Cha, I. H. S-38 

Cha, S. O-190, O-231, O-233 

Chagnon, M. O-144 

Chakeres, D. W. P-53 

Chakraborty, S. P-64 

Chaljub, G. O-107 

Chaloupka, J. C. O-18, O-164, P-139, 

P-142 

Chambers, M. R. O-387 

Chan, E. H. Y. S-68 

Chan, K. E-36 

Chan, L. L. S-48 

Chanelles, M. O-375 

Chaney, K. A. S-25 

Chang, J. S. O-231, O-233 

Chang, P. C. T. S-2 

Chang, S. O-190 

Chang, S. D. O-91 

Chao, C. P. E-29 

Chapman, J. C. O-250 

Charletta, D. A. I-65 

Chason, D. P. S-73 

Chen, C. C. C. S-2 

Chen, P. P-141 

Chen, V. J. O-434, P-100 

Chen, W. S. S-2 

Chen, W-L. P-25 

Chen, X. P-123 

Chenevert, T. L. O-238 

Cheyne, D. P-98 

Chhabra, A. S-89 

Chia, J. S-43 

Chiang, C. M. S-2 

Childs, A. M. O-24 

Chin, C. T. O-362, O-422 

Chinnery, P. S-8 

Chiolero, R. O-376 

Chiras, J. O-97 

Choi, C. G. O-100 

Choi, J. J. P-79 

Choi, S. H. O-242, P-161 

Chong, W. K. O-291 

Chow, M. O-381 

Christoforidis, G. A. O-186, P-52, P-147, 

P-148 

Chuang, S. H. O-494, O-497 

Chugani, D. O-284 

Chugani, H. O-284 

Chung, E. P-14 

Chung, T. W. S-64 

Cil, B. O-163 

Cila, A. E-25, P-48 

Cinnante, C. O-174 

Cirillo, Jr., R. L. P-143 

Cirillo, L. O-428 

Clark, C. O-247 

Clark, H. P-146 

Clark, W. O-170 

Clary, R. P-107 

Cleary, K. R. P-79 

Cloft, H. J. O-39, O-41, O-42, O-43, 

O-148, O-150 

Coelho, D. O-35 

Cognard, C. P-124 

Cognitive and Neurobiological Research 

Consortium-Traumatic Brain Injury 

O-347 

Cohen, J. E. O-53, O-161 

Cohen, W. A. O-248, O-319, O-357 

Colangelo, V. P-56 

Coley, S. C. O-94, O-216 

Collqhoun, I. O-143 

Colosimo, C. P-45, S-19 

Comi, A. M. O-293 

Conley, M. E. O-397 

Connor, S. E. J. P-119, S-50 

Connors, III, J. J. I-199, O-160 

Consigny, D. O-149 

Constantini, S. O-456 

Cook, S. D. O-324 

Cooper, E. O-358 

Cooperman, S. O-472, S-85 

Coopersmith, H. O-80 

Corbera, K. O-299 

Corkill, R. O-145 

Corliss, R. F. E-16 

Corrigan, J. S-69 

Corrigan, K. O-153 

Cortes, M. D. P. O-373, P-179 

Costa, A. O-174 

Cotin, S. E-38 

Coulthard, A. S-8 


Index Key: E = Electronic Scientific Exhibit; I = Invited Speaker; O = Oral Paper; P = Poster; S = Scientific Exhibit

Cowsill, C. O-94 

Craig, E. O-38 

Cramer, D. P-121 

Criqui, G. O-121 

Cristaudo, C. O-432 

Crompton, D. S-8 

Crone, K. O-477 

Cross, B. J. O-378, P-8, P-29 

Cross, D. O-248 

Cross, III, D. T. O-168 

Crowder, C. O-306 

Cruise, G. O-43 

Cruz, Jr., L. C. H. S-40 

Cullen, S. P. O-448, O-449 

Curtis, A. S-8 

Czervionke, L. F. O-386 

D 

da Cruz, Jr., L. C. H. P-15, S-20, S-23, 

S-30 

Dai, D. O-39, O-41, O-42, O-44, 

O-148, O-155 

Dai, Jr., J. O-328, P-59 

Daitzchman, M. O-214, O-377, O-384 

Daneman, A. O-437 

Danielson, M. A. O-39, O-41, O-44, 

O-148, O-150 

Danzer, E. O-318 

Darabi, D. P-5 

Dardzinski, B. O-477 

Darnell, A. S-76 

Das, A. M. O-287 

Das, K. P-64 

Dash, S. O-155 

Dashnaw, S. P-2 

Datir, Sr., A. P. P-27 

Dattani, M. T. O-291 

David, A. P-90 

Davidson, H. C. I-8, I-125, I-27, 

I-50, O-24 

Dawson, S. L. E-38 

Day, A. P-141 

de Araujo, D. B. P-88, P-90 

De Falco, R. O-428 

de Regt, D. E-39 

Dean, A. F. O-492 

DeBaun, M. R. P-152 

Debus, J. O-417 

Decarie, J-C. P-157 

Defatta, R. P-116 

deGuzman, R. P-180 

Deinzer, F. P-123 

DeLaPaz, R. L. O-299 

DeLeon, M. J. P-13 

Delgado, J. E. O-375 

DelGaudio, J. O-32 

Delman, B. N. E-20, E-24, O-375, 

O-506, S-45, S-52 

DeLone, D. R. O-77, O-109 

den Heeten, G. J. O-411 

Derdeyn, C. P. I-133, I-268, O-168 

Desai, S. V. O-507 

Dhaliwal, S. G. O-416 

Di Lella, G. M. P-45, S-19 

DiDomenico, P. P-153 

Diehn, F. E. O-31 

Dillon, W. P. I-66, O-121, O-231, O-233, 

O-320, O-374 

Ding, Y. O-39, O-41, O-42, 

O-44, O-148, O-150 

Diren, B. P-46, P-96 

Dirisamer, A. O-352 

Dixit, S. O-171 

Do, H. M. O-91, O-160 

Doerfler, A. O-119 

Domenici, R. P-155 

Domingues, R. C. P-15, S-20, S-23, 

S-30, S-40 

Donati, M. A. P-155 

Dong, Q. O-238 

Donner, E. J. P-99 

Donnerstag, F. G. F. O-287 

Douville, C. O-319 

Drachenberg, C. B. O-306 

Drake, J. P-182 

Driscoll, C. L. W. E-37, P-118 

Du, J. O-471 

Duckwiler, G. R. I-7, I-264, O-93, 

O-123, O-416, P-144 

Duda, J. O-358 

Duhaime, A-C. O-479 

Dunfee, B. L. O-108, S-58 

Dunkel, I. J. O-472, P-102 

Durán, C. S-76 

Durick, N. O-382 

Durst, A. P-91 

Dutton, R. O-297, P-23 

Dydak, U. O-229, O-353 

E 

449 

Eames, F. S-70 

Ebara, M. E-26, O-46, P-126, P-138 

Eckel, L. J. O-400A 

Ederies, M. A. S-5 

Egashira, R. S-4, S-11 

Egnor, M. R. O-480, O-482 

Ehrich, J. O-287 

Ekholm, S. S-21, S-46 

El Gammal, T. O-314 

Eldevik, P. S-81 

El-Feky, W. O-372 

Elhawarey, G. O-102 

Elijovich, L. S-33 

Elliott, M. A. P-84 

Elmogy, S. O-75, P-43 

Elsamaloty, H. O-501 

Elserougy, L. O-102 

Elshenshawy, H. O-103 

Emerson, J. S-82 

Endo, Y. O-83, P-105 

Engel, Jr., J. O-297, P-23 

Engelhorn, T. O-119 

Engellandt, K. P-42 

Engstrom, J. W. O-422 

Enochs, W. S. O-79 

Enterline, D. S-37 

Enzmann, D. R. I-204 

Epelman, M. O-437 

Eran, A. O-214, O-377, O-384 

Eraso, A. S-63 

Erb, G. O-234 

Erberich, S. G. O-434, P-100 

Erdem, E. O-468 

Erdogan, C. E-6, O-40, P-32 

Erly, W. K. I-457, O-311 

Ervine, S. L. M. O-388 

Escofet, C. S-76 

Escorsi-Rosset, S. P-90 

Escott, E. J. S-59 

Esposito, F. J. S-22 

Esser, D. O-104 

Essig, M. O-249, O-417, O-490, P-47 

Esteves, F. O-99 

Ethridge, K. P. O-107 

Evans, A. L. O-216 

Evans, J. M. S-87 

Evans, J. W. P-99 

Ewend, M. P-36 

Ezz, M. O-286 

F 

Faerber, E. S-89 

Fain, S. O-481 

Falcone, S. P-178 

Falkai, P. O-169 

Fallon, B. O-299 

Fan, X. Y. O-238 

Fanning, N. F. O-55 

Farb, R. O-113 

Farb, R. I. O-152, O-390, O-419 

Farina, D. O-76 

Faris, G. W. S-78 

Farkas, J. O-98, O-155 

Faro, S. H. O-228, O-370 

Fasano, N. P-178 

Fassihi, A. A. O-122, P-73, P-76 

Fatterpekar, G. M. E-24, O-506, S-45 

Fayad, Z. A. O-82 

Fehlings, D. L. P-98 

Fei, G. O-252 

Feng, L. P-2 

Ferrari, A. O-475 

Ferrario, A. O-48, O-175 

Ferreira, F. B. P-15, S-20, S-23 

Ferrer, I. O-301 

Ferris, R. O-101 

Fertikh, D. S-22, S-55 

Feygin, T. O-398, O-486 

Fiehler, J. O-115 

Field, A. S. I-128 

Fields, T. M. P-183, S-24 

Fikackova, H. P-106 

Finden, S. G. O-79 

Fink, J. R. P-144 

Finlay, J. O-442 

Fiorella, D. O-151, O-208, O-211, O-212 

Firat, A. K. O-476 

Firat, Z. Y. O-476 

Fischbein, N. J. O-25 

Fischer, S. O-209 

Fischl, B. P-167 

Flanders, A. E. O-111, O-303, O-356 





Flannery, M. P-95 

Flis, C. M. P-119, S-50 

Floris, R. P-6, P-17, P-19, P-56, P-156, 

P-168 

Flowers, D. L. O-221, O-222 

Floyd, T. S-1 

Foerster, B. P-41, P-83, S-81 

Foltz, C. O-396 

Fonda, C. P-155 

Forbes, G. S. I-265 

Forbes, K. P. O-292 

Ford, III, K. L. S-87 

Fornari, E. O-435 

Forsting, M. O-119 

Foster, M. A. O-84 

Fountain, A. J. O-120 

Fowler, M. O-394 

Fox, A. J. O-55, O-56, O-57, O-58, O-59 

Francescon, P. O-326, S-26 

Frank, D. O-395 

Frazee, J. O-416, P-81 

Frerichs, K. O-20, P-141 

Friedberg, J. O-30 

Friedlich, P. O-450 

Friedman, A. O-224 

Friedman, D. P. O-303 

Frijia, F. P-11 

Frim, D. M. S-78 

Fujikawa, A. O-215, O-359, O-361 

Fujimoto, H. O-21, O-22, O-23, O-179 

Fujimoto, T. O-21, O-22, O-23, O-179 

Fujita, N. S-56 

Fujitsuka, III, M. O-164, P-139, P-142 

Fukuba, E. P-93 

Fukuoka, H. P-31 

Fulbright, R. K. O-250 

Furukawa, S. S-53 

Futterer, S. O-54 

G 

Gada, V. S. S-5 

Gaetz, W. C. P-98 

Gaikwad, S. B. O-380 

Gailloud, P. O-446, O-483, P-129 

Gajjar, A. O-455 

Gallo, B. O-395 

Gandhi, D. O-210, S-69 

Gangadhar, B. N. O-507 

Ganly, I. O-78 

Garaci, F. G. P-6, P-17, P-19, P-56, 

P-156, P-168 

Garbern, J. Y. P-154 

Garcia-Rivas, J. V. P-57 

Garfinkle, W. B. P-7 

Garg, A. O-380 

Gargan, L. O-496 

Gasparini, F. F. O-30, P-182 

Gass, A. O-327 

Gaudiello, F. P-6, P-17, P-19, P-56, 

P-156, P-168 

Gauvrit, J-Y. P-80 

Ge, Y. O-294, O-313, P-10 

Gean, A. D. I-126 

Gedzdman, E. O-395 

450 

Gee, J. C. O-358 

Gelmez, S. P-46, P-96 

Gemmete, J. O-210 

Geng, D-Y. O-241 

Gennari, P. E-1 

Gentili, M. E-1 

Gentry, L. R. E-12, E-13, E-14, E-16, 

E-17, E-31, S-65, S-68 

George, T. O-224 

Georgy, B. A. I-513, P-181 

Gessaroli, M. P-60 

Gevry, G. O-144 

Ghadamgahi, M. P-63, P-66, P-109, 

P-110 

Ghajar, J. O-374 

Ghossein, R. O-78 

Giannotta, S. L. O-122, P-50 

Giesel, F. L. O-249, O-491 

Gifford, W. O-416 

Gilbert, S. C. S-87 

Gilbertson, J. R. O-109 

Giles, B. O-21, O-22, O-23, O-179 

Gillard, J. H. O-492 

Gilles, F. H. I-342, O-434, O-493 

Gimbel, J. R. I-24 

Ginsberg, L. E. I-201 

Given, II, C. A. S-24, S-183 

Gkogkas, C. P-141 

Glenn, O. A. P-162 

Go, J. L. E-19, E-23, I-6, I-25, P-50, 

P-73, P-76 

Goddard, J. O-14, O-431, P-136 

Goericke, S. L. O-119 

Goh, J. S-65, S-68 

Gokan, T. S-46 

Golay, X. I-260 

Goldbach-Mansky, R. O-445 

Goldsher, D. O-111, O-214, O-377, 

O-384 

Gomez, C. E. O-372 

Gomez-Anson, B. M. O-243, O-244, 

O-301 

Gomez-Rio, M. P-57 

Gomori, J. M. O-53, O-161 

Gonzalez Toledo, E. O-394 

Gonzalez, C. F. O-370, S-89 

Gonzalez, R. G. I-274, O-112 

Gonzalez-Ares, S. O-375 

Gonzalez-Gomez, I. O-493 

Goo, H. W. P-161 

Gor, D. M. O-89, O-398 

Goradia, D. O-357 

Gordon, N. O-228 

Gostenik, K. S-7 

Goto, Y. P-171 

Gounis, M. J. O-17, O-95, O-147, P-132 

Govindarajan, S. P-169 

Govindarajulu, S. P-169 

Goyal, M. O-156, S-32, S-36 

Graham, M. M. S-47 

Grahovac, S. Z. O-240 

Grant, P. E. I-402, O-436, P-164, P-165 

P-167 

Gray, L. O-425 

Gray, L. A. O-429 

Green, E. B. P-183, S-24 

Greenberg, B. D. O-226, P-82 

Griffith, G. M. S-34 

Griffiths, P. D. O-216, O-249, O-316, 

O-317, O-447, O-491 

Grimaldi, G. E-7 

Grimme, J. D. E-22, E-40, O-230, 

O-391, S-51 

Grinde, J. O-149 

Grissom, L. S-55 

Grivé, E. O-369, O-498, S-9 

Gropman, A. O-292 

Grossman, R. I. I-333, O-294, O-313, 

O-367, P-10 

Grum, K. O-149 

Grunwald, I. Q. O-45, O-169 

Gu, T. O-382 

Gu, X. P-65 

Guarnieri, G. O-428 

Guilbert, F. O-92, O-146, O-165 

Gül, G. O-45 

Gulati, S. O-288 

Gulka, I. O-162 

Gullane, P. I-202 

Gultekin, S. P-177 

Gunnarsson, T. O-47, O-504, P-70 

Guo, G. O-217 

Guo, Q. Y. P-163 

Guo, X. O-250 

Gupta, A. P-160 

Gupta, A. K. O-26, O-34, O-288, P-159 

Gupta, G. O-90, O-98, O-155, P-2 

Gupta, V. O-380 

Gururangan, S. O-393 

Gutin, P. O-237 

Guvenc, I. P-35, P-37 

H 

Haacke, E. M. O-418, P-170 

Haas, L. J. O-48, O-175 

Hackney, D. B. O-239 

Hadley, W. L. O-300 

Hafez, M. O-75 

Haghighi, M. H. O-324 

Hahm, C. P-111, P-113, P-115 

Hahn, H. O-249 

Hahn, J. P-153 

Haider, J. O-440 

Hakim, M. P-21 

Hakyemez, B. E-6, P-32 

Hall, M. J. O-38 

Hallam, D. K. O-357 

Halliday, W. O-437 

Halovanic, C. O-451, S-43 

Hamedi, M. P-63, P-66, P-109, P-110 

Hamilton, B. E. O-27, O-283, O-307, 

O-322, P-55 

Han, E. T. O-219, O-329 

Han, M. H. P-127, P-140 

Hanlon, F. M. P-94 

Harari, P. M. E-16 

Harder, S. L. O-441 

Hardjasudarma, M. O-394 

Harkey, P. P-22 

Harnsberger, H. R. I-9, O-24, O-27 

Haroon, H. P-39 



Harrington, D. L. P-94 

Harris, F. S-1 

Harris, K. O-224 

Hart, Jr., J. O-223 

Hartig, G. K. E-16 

Hartker, III, R. O-309, O-323 

Hartmann, H. O-287 

Hartmann, M. O-118, O-180, O-234 

Haruo, I. O-383 

Hashimoto, M. P-158 

Hashimoto, T. S-46 

Hasso, A. N. I-63 

Hatakeyama, Y. O-414, P-158 

Hatfield, L. A. O-293 

Hattingen, E. O-235 

Haubitz, B. O-287 

Haughton, V. I-1, I-271, O-471, O-481 

Hawkins, C. O-494 

Hayakawa, K. P-28, P-184 

Hayakawa, M. O-18, P-75 

Hayashida, Y. O-181, O-444, P-31, P-34, 

P-71, S-12 

Hayek, R. A. O-308, O-479 

Haynor, D. R. O-355 

Hecht, E. M. O-366, O-500 

Hecox, K. E. S-78 

Hedlund, G.. O-27 

Heidenreich, J. O. P-67 

Hekmatnia, A. P-109, P-110, P-63, P-66 

Helton, K. J. O-397, O-455 

Helwig, J. A. O-283 

Henkes, H. O-118, O-172, O-209 

Henry, B. P-68 

Heran, M. K. O-438 

Herbert, J. O-294, P-10 

Herial, N. A. O-501 

Herm, R. P-146 

Herskovits, E. H. O-295 

Hess, C. P. O-219, O-329 

Hesseltine, S. O-312, O-313 

Heverhagen, J. O-186 

Higashi, M. S-35, S-39 

Higgins, R. J. P-68 

Hirabuki, N. S-56 

Hirai, T. O-181, O-444, P-31, P-34, P-69, 

P-71, P-171, S-12 

Hirano, M. E-8, O-379, O-383 

Hirsch, J. O-327 

Hiwatashi, A. P-180, S-21 

Ho, T-L. P-25 

Ho, Y. P-175 

Hoang, P. E-19 

Hochhauser, L. P-74 

Hoff, P. R. S-45 

Hoffman, J. M. O-185 

Hoffmann, C. O-250 

Hogan, M. O-156 

Holiday, R. A. E-12, E-18 

Holland, E. C. I-349 

Hollingworth, W. O-277, O-439 

Holloway, K. P-68 

Holodny, A. I. I-196, O-237, O-485, 

P-102 

Holshouser, B. P-151 

Holtzman, A. W. S-70 

Honya, K. O-215, O-359, O-361 

Hoon, Jr., A. H. O-454 

Hooshi, P. E-10 

Hopkin, J. R. O-399 

Hopkins, J. K. O-54, O-423, O-505 

Hori, M. P-85 

Horia, M. S-69 

Horowitz, S. W. S-41, S-42 

Horribine, L. P-130 

Horska, A. O-182 

Hosoki, T. P-24, P-26, P-30 

Hou, B. L. O-237 

Houck, P. R. O-225 

Hourani, R. O-182 

Howieson, J. O-322 

Hsieh, W-J. P-16 

Hsu, D. P. O-116 

Hsu, L. P-18 

Hsu, S-W. P-139 

Htaik, T. P-7 

Hu, L. S. P-116 

Hu, P. O-355 

Hu, Q. E-32 

Huang, D. O-381 

Huang, M. X. P-94 

Huang, S. E-32 

Huang, W. P-102 

Hudgins, P. A. IO-27, O-32, O-99 

Huebsch, T. P-54 

Huerga, E. O-369 

Huh, B. O-425 

Hui, F. O-509, P-61 

Hum, B. O-370 

Hung, H. C. S-2 

Hunter, J. V. O-440, S-64 

Hurcan, C. P-4 

Hurst, R. W. O-89, O-378, P-8, P-29 

Huston, III, J. O-51 

Hutchinson, P. J. A. O-492 

Hwang, J. C. O-242 

Hynynen, K. O-176 

I 

451 

Iancu-Gontard, D. O-92, O-146 

Iannucci, G. O-326, S-26 

Ibrahim, M. P-182 

Ida, Y. P-75 

Ifthikharuddin, S. F. S-90 

Iihara, K. S-35, S-39 

Ikonen, S. P-1 

Ilkbahar, S. P-37 

Imakita, S. S-35, S-39 

Imbesi, S. G. O-37 

IMS Study Group O-158 

Imuta, M. P-171 

Inagawa, S. E-8, O-379, O-383 

Incesu, L. P-46, P-96 

Inglese, M. O-488 

Inoue, T. P-3 

Ionete, C. O-49 

Irie, K. E-26, O-46, P-75, P-126, P-138 

Isalberti, M. O-174 

Ishibashi, T. E-26, O-46, P-126, P-138 

Ishigame, K. P-85 

Ishikura, R. S-56 

Isoda, H. E-8, O-379 

Itoh, H. O-285, S-79 

Iuliano, E. M. P-79 

Ivanov, N. E-2 

Izbudak, I. O-502 

J 

Jackson, A. O-325, P-39 

Jacobson, J. P. P-151 

Jafri, Z. O-508 

Jahan, R. O-123, O-416, P-81 

Jain, R. O-424, O-474, S-67, S-69 

Jallo, G. I. O-315 

James, A. O-437 

Jan, S. O-430 

Jansen, O. O-365, P-42 

Jarvik, J. G. O-277, O-355, O-439 

Jawad, A. O-295 

Jayakumar, P. N. O-304, O-507 

Jayaraman, M. V. O-91, O-160 

Jayaraman, S. E-20, S-52 

Jellineck, D. A. O-491 

Jena, R. O-492 

Jennings, J. E. E-17 

Jeon, S-S. P-89 

Jeong, A. K. O-242 

Jeong, H-S. P-112 

Jerosch-Herold, M. O-185 

Jiang, H. O-454 

Jiarakongmun, P. O-154, P-125 

Jiménez de la Peña, M. D. P-134 

Jin, L. O-252 

Johnson, A. J. O-314 

Johnson, B. A. I-515 

Johnson, G. O-183, O-184, O-232, O-294, 

O-312, O-313, O-488, P-10 

Johnson, J. O-25 

Johnson, M. D. O-387 

Johnson, M. H. I-10, O-35, S-71 

Johnson, M. P. O-318 

Johnston, B. D. O-439 

Johnston, M. V. O-454 

Jolesz, F. O-176 

Jones, B. V. O-477 

Jones, C. K. O-191 

Jones, J. O-445 

Jordan, L. O-446, O-483 

Joseph, G. O-324 

Joseph, G. J. O-167 

Jradi, H. A. P-147 

Judaš, M. O-290 

Juhasz, C. O-284 

Juncosa, A. E-10 

Jung, S-L. E-11, P-78, P-89 

K 

Kaakaji, R. S-42 

Kabra, M. O-288 

Kademian, J. C. E-27 

Kadirvel, R. O-39, O-41, O-42, O-44, 

O-148, O-150 

Kadkhodayan, Y. O-168 





Kahana, E. O-250 

Kajitani, T. P-93 

Kakeda, S. O-414, P-158 

Kakimoto, N. S-53 

Kalapos, P. O-162 

Kale, H. A. O-400B 

Kalina, P. O-110 

Kallenberg, K. O-251 

Kallmes, D. F. O-39, O-41, O-42, O-43, 

O-44, O-51, O-148, O-150, 

O-429, O-463, O-465, O-475A 

Kalra, V. O-288 

Kamel, Y. O-102, O-103 

Kamholz, J. P-154 

Kan, S. P-28, P-184 

Kanal, E. I-24, O-278 

Kandel, A. O-103, O-286 

Kane, A. G. O-52 

Kanekar, S. G. E-13, E-27, S-86 

Kang, B. S. O-242 

Kang, H. S. P-127, P-140 

Kanno, T. P-75 

Kapilamoorthy, T. R. P-159 

Karakas, Sr., H. M. O-476 

Karim, A. O-394 

Karimi, S. I-272, O-105, O-485, P-102 

Karli Oguz, K. P-48 

Karmarker, P. O-178 

Karmonik, C. P-123 

Kashiwagi, N. S-56 

Kasotakis, M. S-41 

Kasprian, G. O-290 

Kassel, E. E. I-137 

Kassner, A. O-113, O-114, O-117 

Katada, K. P-75 

Katsube, T. P-93 

Katzman, G. L. I-3, I-69, I-59 

Kaufmann, T. J. O-51, O-463,O-465 

Kawamura, Y. O-285, S-79 

Kazmi, K. S-33 

Kelley, E. L. S-71 

Kemp, W. L. S-73 

Ken, L. P-129 

Kendall, B. E. O-508 

Kennedy, D. N. P-167 

Kennedy, R. J. O-24 

Kentenich, M. O-413 

Kepka, A. P-51 

Kerber, C. W. O-37 

Kesavadas, C. P-159, P-160 

Keshavan, M. S. O-507 

Keston, P. P-130 

Ketkar, M. O-394 

Keyserling, H. O-425 

Khademian, Z. P. O-497 

Khan, A. O-418 

Khan, R. B. O-397, O-455 

Khan, V. O-162 

Khandelwal, N. O-26, O-34 

Khatri, P. O-158 

Khawar, S. O-157, O-321, O-423, O-505 

Khosla, A. E-15 

Khoury, J. O-158 

Kido, D. K. P-151, P-170 

Kikuyama, A. S-56 

Kim, A. K. Y. O-122, P-73 

Kim, B-S. E-11, P-78, P-89 

452 

Kim, D. I. P-97, P-133 

Kim, D. J. P-133 

Kim, E. P-150 

Kim, H. J. O-100 

Kim, H-J. P-112, P-150 

Kim, J. O-159, P-97, P-133 

Kim, J. H. P-114, S-38 

Kim, J. K. O-388 

Kim, J-S. P-89 

Kim, M. J. J. O-237 

Kim, P. E. O-122, P-73, P-76 

Kim, R. Y. O-314 

Kim, S. P-111, P-113, P-115 

Kim, S. H. P-127, P-140 

Kim, S. J. O-100 

Kim, S. S. P-150 

Kim, Sr., H-J. S-61 

Kim, S-T. P-112, S-61 

Kim, T. P-2 

Kim, T-K. S-38 

Kim, W. O-362 

Kim, Y. P-111, P-113, P-115 

Kim, Y. P-111, P-113, P-115 

Kim, Y. M. S-61, P-97 

Kim, Y-J. E-11 

Kimiwada, T. O-284 

Kincaid, J. S-27 

Kingsley, P. B. O-250 

Kinney, H. C. O-450 

Kirchin, M. P-47 

Kirsch, W. O-418 

Kish, K. O-86, S-7 

Kitagaki, H. P-93 

Kitajima, M. O-181, O-444, P-31, P-34, 

P-71, S-12 

Kizilkilic, O. P-4 

Klingebiel, R. O-413, P-104 

Klisch, G. P-143 

Klotz, E. O-180, O-234 

Klucznik, R. P-72, P-123 

Klurfan, P. O-47, O-504, P-70 

Knauth, M. O-251 

Knopp, E. E-33, O-488 

Knopp, M. V. O-186, O-490 

Knox, K. O-37 

Knyazeva, M. G. O-435 

Ko, N. U. O-121 

Ko, Y-H. P-112 

Kobayashi, N. O-466, S-91 

Kocaoglu, M. P-35, P-37 

Kodama, N. O-83, P-105 

Kodama, T. P-120 

Koenigsberg, R. S-89 

Koff, M. O-399 

Kohara, N. O-302 

Koktzoglou, I. O-82 

Kollias, S. I-405, O-353, O-360 

Kong, K. M. O-85 

Konrad, F. O-187 

Kopell, B. O-226, P-82 

Koral, K. O-496, S-43 

Kornegay, A. L. P-77 

Korogi, Y. E-28, O-181, O-414, O-444, 

P-31, P-34, P-71, P-158, P-171, S-12 

Korones, D. N. O-497 

Koroshetz, W. J. O-112 

Kost, G. J. S-82 

Kostic, V. O-245 

Kostovic, I. O-290 

Kosugi, T. E-8, O-379 

Kotsenas, A. L. O-386 

Kott, B. E-39 

Kowal, D. J. P-7 

Kozerke, S. P-103 

Kozic, D. B. O-245 

Kraft, R. A. O-221, O-222 

Kragha, K. K. P-21, S-27 

Kraut, M. A. O-223 

Krejza, J. O-295 

Kress, B. P-91 

Kretzschmar, H. A. O-251 

Krick, C. O-169 

Krieger, M. D. O-442, O-493 

Krishnamoorthy, A. P-169 

Krishnamoorthy, K. S. O-436, P-164, 

P-165 

Krishnamoorthy, T. P-159 

Krol, G. S. O-472, P-137, S-85 

Kroma, G. S-63 

Kruger, A. Y. O-400 

Ku, A. I-198 

Kucinski, T. O-115 

Kudo, S. S-4, S-11 

Kuehne, D. O-172, O-209 

Kueker, W. O-145 

Kuhl, C. K. I-404 

Kuhtz-Buschbeck, J. P. O-365 

Kumagai, H. P-85 

Kun, L. O-455 

Kuo, J. S. P-50 

Kuratsu, J. O-181, P-31, P-34 

Kuruoglu, E. P-96 

Kwock, L. O-230, P-44 

Kwon, B. J. P-127, P-140 

Kwon, O. K. P-127, P-140 

Kwon, T. H. S-38 

L 

La Fata, V. O-32 

Lake, D. R. O-443 

Lalwani, A. K. O-25, O-28 

Landis, D. M. D. O-116 

Lane, J. I. E-37, O-31, O-110, P-118 

Lang, Z. J. O-85 

Lange, T. O-229 

Langer, D. J. P-145 

Lanzieri, C. F. O-116 

Laothamatas, J. P-176 

Larheim, T. A. S-53 

Larsen, D. W. O-122, P-73 

Larson, D. O-190 

Lasjaunias, P. O-448, O-449 

Laub, G. E-3 

Laukka, J. P-154 

Laurienti, P. J. O-221, O-222 

Lavini, C. O-411 

Law, E. M. O-366, O-500 

Law, M. O-183, O-184, O-232, O-294, 

O-312, O-313, O-488, O-489, 

P-10, S-33 

Lawrence, J. O-360 



Layton, K. F. O-475A 

Layzer, R. O-422 

Lazo, C. O-246, O-298 

Le, T. H. O-374 

Leblanc, P. O-165 

Lechter, M. P-131 

Leclerc, X. P-80 

Lee, A. O-297, P-23, S-54 

Lee, B. C. E-24, O-506 

Lee, B. C. P. O-452, P-107 

Lee, C. P-183, S-24 

Lee, D. O-100, O-162 

Lee, H. K. E-27, O-100, S-6, S-15, S-47, 

S-54, S-68, S-86 

Lee, H-J. O-106, P-25 

Lee, J. O-367 

Lee, J. D. P-97 

Lee, J. H. O-100 

Lee, J. S. E-20, P-117 

Lee, J-H. O-242 

Lee, J-W. P-78 

Lee, K-W. P-25 

Lee, N. J. P-114, S-38 

Lee, R. K. P-180 

Lee, R. R. P-94 

Lee, S. P-111, P-113, P-115 

Lee, S. K. O-18, O-47 

Lee, S-K. P-97 

Lee, T. S-2 

Lee, Y. P-114 

Lee, Y. H. S-38 

Lee, Y. J. P-133 

Lees, J. O-309, O-323 

Lefton, D. R. O-315, P-145 

Le-Huu, M. P-47 

Leite, J. P. P-88 

LeJeune, J. J. P-5 

Lekht, I. O-120 

Lemaire, L. P-5 

Leonard, J. R. O-452 

Leone, M. O-475 

Lertvananurak, R. P-153 

Lesley, W. S. O-385 

Lev, M. H. O-112 

Lev, S. S-18, S-88 

Levey, E. O-454 

Levin, H. S. O-440 

Leviton, A. I-341 

Levy, D. O-491 

Levy, R. A. P-146 

Levy, R. M. P-51 

Lewis, A. M. O-422 

Lewis, D. O-319 

Lewis, D. A. O-39, O-41, O-42, O-44, 

O-148, O-150 

Lewis, R. P-154 

Lexa, VII, F. I-71 

Leypold, B. O-356 

Li, C-S. O-455 

Li, D. O-82 

Li, J. O-223, P-13 

Li, L. O-328 

Li, L. P. P-58 

Li, Sr., J. P-59 

Li, X. O-231, O-233 

Li, Y. P-13 

Li, Y-X. O-241 

Liao, C. Ta. P-108 

Licitra, R. P-11 

Lieber, B. B. O-17, O-147, P-132 

Liebeskind, D. S. O-378, P-8, P-29, P-77 

Liebig, T. O-172, O-209 

Liess, C. P-42 

Lignelli, A. O-299 

Lim, W. S-48 

Lin, C-C. P-92 

Lin, D. D. O-293 

Lin, E. O-280, P-2 

Lin, F. S-71 

Lin, J. O-297, P-23 

Lin, Q. O-189 

Lin, R. O-85, O-217 

Lindell, E. P. O-31, P-118 

Lindisch, D. J. P-79 

Linfante, I. O-95, O-147 

Linnau, K. F. O-357, O-439 

Linovitz, R. O-433 

Lis, E. O-472, O-485, P-137, S-85 

Litt, B. P-77 

Littenberg, B. O-314 

Liu, C-S. J. P-84 

Liu, G. R. O-85, P-58 

Liu, G. T. P-84 

Liu, H-M. P-16, P-128 

Liu, J. O-324 

Liu, S. O-121, O-320 

Liu, X. O-328, O-364, O-450, O-493, 

S-57 

Livshiz, J. O-184, O-489 

Llamas-Elvira, J.M. P-57 

Lodemann, K-P. P-47 

Logan, W. J. P-98, P-99 

Lombardo, F. P-11 

Long, M. P-65 

Lonser, R. L. P-38 

López Pino, M. A. P-134 

Lopez, C. J. O-436, P-164, P-165 

Lowe, M. J. O-178, O-226, O-327, P-82 

Lowens, S. O-172 

Lownie, S. P. E-9, O-162 

Loya, A. O-298 

Lu, Y. O-320 

Luboz, V. E-38 

Lucaya, Sr., J. E-21 

Lüdemann, L. P-67 

Ludovici, A. P-6, P-17, P-19, P-56, 

P-156, P-1-68 

Luecke, T. O-287 

Lui, Y. W. O-232 

Lum, C. O-156, S-32, S-36 

Luo, S. E-32 

Lury, K. M. E-4 

Lutsep, H. I-6, I-132, O-170 

Lylyk, P. O-48, O-175 

M 

453 

Madison, M. T. O-14, O-431, P-136 

Maeda, M. P-49, S-14, S-44 

Maeder, P. P. O-376, O-435 

Magalhaes, A. C. A. O-86, S-7 

Magalhães, F. V. S-20, S-40 

Magee, T. O-309, O-323 

Maia, R. P-15 

Maier, S. E. P-49, P-93, S-44 

Maira, G. P-45, S-19 

Majoie, C. B. O-411 

Makki, M. O-284, P-101, P-154 

Makris, N. P-167 

Maldjian, J. A. I-195, O-221, O-222 

Maldonado, T. O-367 

Maley, J. S-6, S-15, S-47 

Malik, O. O-296 

Malin, D. R. O-81 

Mallesh, A. P-21 

Malone, D. O-226, P-82 

Malone, S. A. P-99 

Maly, P. P-41, S-81 

Mamourian, A. C. O-308, O-399, O-479, 

S-10 

Mandrekar, J. N. O-51, O-148, O-150 

Manfrè, L. O-427, O-432 

Mani, V. O-82 

Manley, G. T. O-374 

Manninger, S. P. O-185 

Mannon, L. O-294, P-10 

Mansour, K. O-99 

Mantilla-Martin, Sr., T. O-499 

Manzo, L. P-60 

Marc, G. P-5 

Marcellus, M. L. O-91, O-160 

Marcuccilli, C. J. S-78 

Marin, H. O-424, S-67 

Markl, M. E-8, O-379, O-383 

Marks, M. P. I-134, O-91, O-160 

Maroldi, R. O-76 

Maroney, T. P. P-143 

Marotta, T. O-152 

Marrie, R. A. O-327 

Marsot-Dupuch, K. O-391 

Martel, M. O-35 

Martín, A. S-31 

Martín, C. S-76 

Martin, D. S. I-25, I-52 

Martin, N. P-81 

Martínez de Vega, V. P-134 

Martinez, M. S-31 

Martuzzi, R. O-435 

Maruyama, N. O-444 

Marziali, S. P-6, P-17, P-19, P-56, 

P-156, P-168 

Masaryk, T. J. O-208, O-381 

Mas-Caradec, M-C. P-5 

Matheus, M. G. E-4 

Mathew, A. P-160 

Mathis, C. A. O-225 

Mathur, A. P-166 

Matsuda, K. M. O-436, P-164, P-165 

Matsumoto, M. O-83, P-105 

Matsunaga, K. P-28 

Matsusako, M. O-466, S-91 

Maurer, P. O-433 

Mawad, K. O-95 

Mawad, M. E. P-72, P-123 

Mayo, M. C. O-231 

Mayock, R. E-39 

Mayumi, M. O-285 

Mazanek, J. P-106 

Mazumdar, S. O-225 





McAnulty, A. L. S-73 

McCollough, C. H. O-31 

McComb, J. G. O-493 

McCormack, E. J. O-480, O-482 

McCue, J. E-16, O-382 

McDannold, N. O-176 

McDougall, C. G. O-151, O-208, O-211, 

O-212 

McGarvey, M. L. P-77, S-1 

McGehee, B. E. O-107 

McGraw, C. P-9 

McIntyre, J. J. O-479 

McKinstry, R. C. P-152, P-166 

McLean, N. O-32 

McNicholl, S. T. P-77 

Meads, D. O-159 

Meder, J-F. P-80 

Medhkour, A. O-501 

Meeker, M. O-374 

Mehta, A. O-112, O-291 

Mehta, H. O-291 

Melancon, D. O-373, P-131, P-179 

Melhem, E. R. O-247, O-295, O-378, S-1 

Melis, M. P-6, P-17, P-168 

Meltzer, C. C. O-225 

Men, S. O-162 

Menda, Y. S-47 

Mendelsohn, D. B. P-116 

Meshberg, E. O-357 

Messé, S. R. P-77 

Meszarosova, M. P-106 

Metcalfe, A. O-144, O-165 

Meuli, R. A. O-111, O-376, O-435 

Meyers, S. P. O-497 

Michals, E. S-25 

Midia, M. P-175 

Mihmanli, I. P-4 

Mikulis, D. J. O-113, O-114, O-117, 

O-217, O-419 

Milgrom, L. I-16, I-57 

Miller, A. O-387 

Miller, B. P-121 

Miller, B. A. O-491 

Miller, D. P-86 

Miller, E. O-456 

Miller, G. A. P-94 

Miller, G. M. P-62 

Miller, J. O-99 

Miller, S. O-21, O-22, O-23, O-179 

Miller, W. O-156, S-36 

Mills, S. P-39 

Milner, L. D. O-221, O-222 

Miloslavski, E. O-172, O-209 

Minichilli, F. P-11 

Minoshima, S. O-248 

Miranda, C. O-48, O-175 

Mironov, A. E-5 

Mirsky, D. M. O-421 

Mirza, S. O-355, O-357 

Mishra, N. K. O-380 

Mistretta, C. O-382, O-471 

Mithalal, R. S-22 

Mitomo, M. P-24, P-26, P-30 

Mitra, S. O-34 

Mitrabhakdi, E. P-176 

Mitsias, P. D. O-424 

Miyakoshi, A. P-145 

454 

Miyamoto, S. S-35, S-39 

Miyamoto, W. O-414 

Miyata, Y. P-120 

Mizsei, G. O-82 

Mody, S. O-453 

Moeller, F. O-365 

Moftakhar, R. O-19, O-94 

Mohamed, F. B. O-228, O-370 

Mohammad, Y. M. P-147, P-148 

Mohr, G. P-131 

Moinuddin, A. P-152, P-166 

Molyneux, A. O-145 

Molyneux, A. J. O-94 

Monaco, F. O-475 

Monajati, A. S-90 

Mondani, M. S-26 

Monier, S. O-75 

Montalban, X. O-369 

Montanaro, D. P-11 

Montanera, W. J. O-152 

Montoya Bordón, J. P-134 

Moodie, R. O-437 

Moon, W-J. P-14 

Moonis, G. O-247, O-295 

Moore, K. R. O-442 

Morales, R. E. O-306 

Morales, T. O-416 

Moran, C. J. O-168 

Morasch, M. D. O-82 

Moret, J. O-96 

Mori, S. O-191, O-452, O-454 

Mori, Y. E-26 

Morimoto, A. K. O-27 

Morino, H. S-56 

Morishita, S. P-34 

Moritani, T. E-27, S-6, S-15, S-47, S-54, 

S-80 

Moriya, J. O-414, P-158 

Morris, C. S-8 

Morris, D. O-171 

Morris, D. E. O-236 

Morris, J. M. P-62 

Morris, P. P. O-159 

Morsi, H. P-72 

Mortilla, M. P-155 

Moskowitz, C. O-237 

Moss, R. L. O-35 

Mounayer, C. O-96 

Mount, D. L. E-14 

Mrazek, D. A. O-227 

Mühler, M. O-413 

Muilli, D. P-121 

Mukherjee, P. O-219, O-329, O-362, 

O-374 

Mukherji, S. K. O-27 

Mukundan, S. O-393 

Mull, M. O-213, P-135 

Mun, S. K. P-79 

Munechika, H. S-46 

Munoz, D. O-311 

Muñoz, E. O-243, O-244 

Murakami, R. O-444 

Murakami, S. S-53 

Murao, K. S-35 

Murayama, Y. E-26, O-46, P-126, P-138 

Muro, G. J. E-36, E-41, I-9 

Murphy, K. J. O-446 

Musacchio, A. O-48, O-175 

Muto, M. O-427, O-428 

Muzik, O. O-284 

Myers, M. E. O-14, O-431, P-136 

Myers, T. V. O-14, O-431 

Mylett, K. A. P-77 

Myseros, J. O-477 

N 

Na, D. G. P-150 

Naeini, R. P-72 

Nag, S. O-152 

Nagae-Poetscher, L. M. O-454 

Nahser, H. C. P-64 

Naidich, T. P. E-24, O-375, O-506, P-12, 

P-117, S-45, S-52 

Naito, H. S-35, S-39 

Nakajima, M. O-215, O-359, O-361 

Nakamoto, Y. O-302 

Nakamura, H. O-181, O-444, P-31 

Nakata, Y. P-85 

Nalini, A. O-304 

Nallainathan, S. K. E-41 

Nan, B. P-41 

Nanda, A. O-394 

Natarajan, V. I-21 

Near, L. O-98 

Nedzi, L. A. O-396, P-172 

Neff, J. O-158 

Negoro, M. P-75 

Neil, J. J. P-166 

Nelson, Jr., M. D. O-434, O-442, O-450, 

O-493, P-100 

Nelson, K. L. P-40 

Nelson, S. O-231, O-233 

Nennig, E. P-87, P-91 

Nesbit, G. M. I-257, O-111, O-170, 

O-173, O-185, O-283, 

O-307, O-354, P-55 

Neumann, P. F. E-38 

Neuwelt, E. A. O-185 

Newell, D. O-319 

Ng, S. H. P-108 

Nguyen, D. T. D. P-79 

Nguyen, H. P-74 

Nguyen, T. O-156 

Nicholas, J. O-220, P-22 

Nicol, E. O-210 

Nielsen, J. F. P-100 

Niemann, D. O-19, O-94 

Niitatori, T. O-215 

Nijenhuis, R. J. O-213, P-135 

Nishida, M. P-167 

Nishino, K. O-414 

Nitatori, T. O-359 

Niven, S. P-64 

Nixon, T. E. P-64 

Nizan, Z. O-250 

Nogueira, J. S-20, S-23 

Nomiyama, K. S-4, S-11 

Norbash, A. O-20 

Nos, C. O-369 

Noujaim, S. E. P-21, S-27, S-49 



Nouri-Mahdavi, K. P-63, P-66, P-109, 

P-110 

Nowinski, W. L. E-2, E-32 

Numaguchi, Y. O-466, S-91 

Nuzzi, N. P. O-174 

Nyenhuis, J. O-178 

O 

O’Connor, E. S-42 

Oda, N. O-414 

O’Donovan, C. A. O-159 

Ogasawara, K. P-3 

Ogawa, A. P-3 

Ogawa, T. P-126, P-138 

Ogg, R. O-455 

Ogoudikpe, C. O-144 

Oguzkurt, L. P-4 

Oh, S. O-184, O-488 

Ohgiya, Y. S-21, S-46 

Ohnari, N. O-414, P-158 

Oka, M. O-483, S-21, S-46 

Okada, T. O-302 

Oklu, R. P-2 

Okuda, T. P-171 

Oleaga, J. P-7 

Omojola, M. F. O-49 

Omura, M. C. O-412 

Onishi, Y. S-35, S-39 

Onoue, H. P-126, P-138 

Onyx Brain AVM Trial Investigators O-93 

Ootaka, H. P-28 

Opatowsky, M. J. O-372, S-87 

Oppenheim, C. P-80 

Oral, R. S-80 

Orbach, D. B. O-367 

Origitano, T. O-38 

Oshima, M. P-75 

O’Sullivan, B. I-203 

Otto, N. P-42 

Ozanne, A. O-448, O-449 

Ozarslan, E. O-363 

Ozelame, R. O-494 

Ozgen, N. E-25 

Ozturk, A. O-40, O-484, P-48 

Ozturk, H. O-484 

P 

Pabon, B. O-175, O-48 

Pace, M. T. O-424, O-474 

Padfield, N. O-30 

Padmanabhan, R. O-145 

Palacios, E. S-63 

Pamuk, A. G. O-40 

Pan, J. E-41 

Panday, A. K. O-26 

Pang, Y. P-83 

Panigrahy, A. I-343, O-434, O-442, 

O-450, O-493, P-100 

Papa, M. S-22 

Papanagiotou, P. O-45 

Papon, X. P-5 

Papsin, B. C. O-30, O-392 

455 

Parer, J. T. P-162 

Parikh, A. H. P-36 

Park, C. P-111, P-113, P-115 

Park, D. P-111, P-113, P-115 

Park, J. K. O-242 

Park, S-W. S-61 

Parkinson, R. J. O-54, O-157, O-505 

Parlak, M. E-6, P-32 

Parmar, H. A. O-392, O-494, O-509, P-61 

Parnes, S. S-70 

Parrish, T. B. I-35, O-503, P-51 

Pasco, A. P-5 

Pasquini, E. P-155 

Patankar, T. P-39 

Patel, A. O-25 

Patel, M. O-143 

Patel, N. A. O-430 

Patel, P. S-69 

Patel, S. A. O-370 

Patel, S. C. O-424, S-67, S-69 

Patel, S. G. O-78 

Patton, A. E-29 

Peck, K. K. O-237 

Pectasides, M. O-436, P-164, P-165 

Pedraza, S. O-111 

Pegoraro, V. E-38 

Pekala, J. S. O-224, O-393, O-425, S-10, 

S-37 

Pelc, N. J. E-8, O-379, O-383 

Pelz, D. M. E-9, O-162 

Peppers, T. O-433 

Perez, C. L. S-73 

Perez-Garcia, M. P-57 

Pergolizzi, R. M. O-36 

Perlman, S. O-246 

Perlow, A. O-95, O-147 

Perry, J. O-471 

Persiva, O. O-498 

Peterova, V. P-106 

Petersen, B. O-170, O-173 

Petrella, J. R. I-273, O-224 

Petrou, M. P-41, P-83, S-81 

Petrovic, I. O-245 

Phillips, C. D. E-13 

Phillips, M. D. O-178, O-226, O-327, 

P-82 

Phillips, N. O-455 

Piacentino, M. O-326 

Piantini, R. E. O-441 

Piazzalunga, B. O-76 

Pickard, J. D. O-492 

Piekut, D. T. P-180 

Pienaar, R. O-436, P-164, P-165, P-167 

Pikus, L. O-295 

Pile-Spellman, J. O-16, O-279, O-280, 

P-2, P-173 

Pillai, J. J. I-409 

Pinna, V. O-326, S-26 

Pinto, M. H. S-22, S-55 

Pinto, R. O-315 

Piotin, M. O-96 

Piovan, E. O-428 

Piqueras, Sr., J. E-21 

Pistey, R. O-108 

Pitcock, J. A. O-223 

Pivawer, G. E-12 

Place, C. S-32 

Platek, S. O-228 

Platnick, J. O-280 

Plosker, A. D. S-73 

Poch, Sr., J. M. E-21 

Pochettino, A. S-1 

Polasani, R. S. P-7 

Politi, M. O-45 

Pollock, A. N. O-398, O-486 

Pollock, J. A. I-197 

Polyakov, I. O-149 

Pongpech, S. O-154, P-125 

Pontes-Neto, O. M. P-88 

Ponzo, J. A. O-305, O-464 

Port, J. D. O-227 

Poskitt, K. J. O-438 

Post, M. J. D. P-178 

Pourmoghaddas, A. P-109, P-110 

Powell, N. S-29 

Pramanik, B. O-366, O-367, O-500 

Prayer, D. O-290, O-371 

Prayson, R. O-381 

Prenafeta, M. S-16 

Prescod, K. O-510 

Prestigiacomo, C. J. O-90, O-98, O-155 

Preul, C. P-54 

Price, J. C. O-225 

Price, S. J. O-492 

Primak, A. N. O-31 

Procopio, E. P-155 

Prohovnik, I. O-250 

Provenzale, J. M. I-270, I-347 

Pruessmann, K. P. I-460 

Pujol, S. O-20 

Pulfer, K. A. E-14, E-31 

Puppi, A. M. O-475 

Purcell, D. D. O-25 

Purdy, P. O-21, O-22, O-23, O-179 

Putman, C. M. O-36 

Q 

Quebada, P. B. O-479 

Quest, R. O-296 

R 

Rabah, R. O-453 

Rabe, A. O-453 

Rabeah, O. O-286 

Rad, M. O-183, O-313 

Radhakrishnan, K. P-160 

Radhakrishnan, V. V. P-160 

Radoš, M. O-290 

Ragoowansi, A. O-400 

Railo, M. P-1 

Raizer, J. S-41, S-42 

Ramnarine, D. O-510 

Ramnath, R. O-309, O-323 

Ramos, E. O-306 

Ramos-Font, C. P-57 

Ranganathan, L. N. P-169 

Rao, V. O-79 

Rappe, A. O-149 





Rapps, N. O-187 

Rasey, J. S. I-2, I-20 

Rasmussen, P. A. O-208, O-381 

Raymond, J. O-92, O-144, O-146, O-165 

Raymond, S. O-176 

Razack, N. O-210 

Recchia, F. M. O-81 

Recio Rodriguez, M. P-134 

Reddy, R. O-247 

Reede, D. L. E-12, E-13, E-14, E-16, 

E-18, E-31, I-138 

Regli, L. O-376 

Reich, D. S. O-191 

Reinartz, J. O-172, O-209 

Reith, W. O-45, O-169, P-47 

Remmele, C. O-115 

Renowden, S. O-510 

Replogle, R. O-21, O-23, O-179 

Revzin, M. V. S-88 

Reyes, M. O-396 

Reyes, T. O-478, S-43 

Reynolds, III, C. F. O-225 

Rezaei, M. P-63, P-66, P-109, P-110 

Rezai, A. R. O-178, O-226, P-82 

Riccelli, L. P. P-55 

Richard, O. O-143 

Riles, T. O-367 

Riley, G. T. O-308 

Ringer, A. O-153 

Ringleb, P. O-118 

Rio, J. S-9 

Ririe, D. O-307 

Robb, R. E-37 

Robert, R. O-22 

Roberts, T. P. L. I-259, I-53, I-408, 

I-461, O-114, O-229 

Roberts, W. O-173 

Robertson, H. J. S-63 

Robinson, R. A. O-33, S-66 

Robledo, O. O-144 

Robson, C. D. I-511 

Roche, P. E. O-480, O-482 

Rodesch, G. O-448, O-449 

Rodriguez, D. P. O-495 

Rodriguez, J. O-394 

Rodriguez-Fernandez, A. P-57 

Roh, H. G. P-150 

Rollins, N. K.O-451, O-478, O-496, S-43 

Roncaroli, F. O-296 

Rorke-Adams, L. B. O-389 

Rosel, P. O-396, P-172 

Rosenbaum, K. N. O-292 

Rosenblum, M. O-489 

Ross, B. D. I-348 

Ross, J. S. I-345, P-40 

Rosset, S. P-88 

Rossi, B. P-11 

Rotblat, M. O-106 

Roth, C. O-45 

Rother, J. O-115 

Rothfus, W. E. O-88 

Rothman, S. L. G. E-30, O-310 

Roughley, S. P-64 

Rovira, A. O-369, O-498, S-9 

Rovira-Gols, A. S-9, S-16, S-31, S-76 

Rowan, S. O-113, O-117 

Rowley, H. A. I-127, O-19, O-229, O-382 

Roy, D. O-92, O-146, O-165 

Rudwan, W. O-75 

Rumboldt, Z. O-220, O-443, O-487, P-22 

Runge, V. M. E-32, P-40, S-3 

Russell, E. J. I-332 

Rutka, J. I-276, O-494 

Rutz, A. P-103 

Ryoo, J. W. P-150 

S 

456 

Saatci, I. E-25, O-40, O-163, O-484 

Sabir, S. S-75, S-77 

Saboori, M. P-63, P-66 

Sacco, M. O-475 

Sacher, M. O-506 

Sachin, S. P-22 

Sackeim, H. O-299 

Sadasivan, C. O-147 

Sadek, A. O-286 

Saguchi, T. E-26, O-46, P-126, P-138, 

P-184 

Said, A. H. M. E-9 

Sainz, A. O-243, O-244 

Sakahara, H. E-8, O-379, O-383 

Sakai, M. P-24, P-26, P-30 

Sakai, O. O-108, S-58 

Sakaie, K. O-321 

Sakamoto, A. C. P-88, P-90 

Sakamoto, S. O-302 

Sakuma, H. S-44 

Sakuma, J. O-83, P-105 

Salamon, G. E-3, E-10 

Salamon, N. E-3, E-10, O-246, O-297, 

O-298, P-23 

Salazkin, I. O-144 

Saleem, S. N. E-9 

Salzman, K. L. O-24 

Sánchez, E. O-498 

Sanchez-Guerra, M. O-301 

Sandhu, J. S. O-147 

Sanei, H. P-63, P-110 

Sanghvi, A. N. O-400 

Sano, H. P-75 

Santi, S. D. P-13 

Santos, A. C. P-20, P-88, P-90 

Santos, M. B. M. P-20 

Sarac, S. E-25 

Sarafoglu, T. O-375 

Saraiva, L. A. L. P-20 

Sarcone, A. P-79 

Sartor, K. O-187, P-91 

Sasaki, C. O-35 

Sasaki, T. O-83, O-183, P-105 

Sato, Y. S-6, S-15, S-54, S-80 

Sattenberg, R. J. P-7 

Satti, S. S-89 

Sawanyawisuth, K. P-174 

Sawlani, V. S-29 

Sayre, J. O-123, O-416, P-81 

Schaefer, P. W. I-258, O-112 

Schellhas, K. P. I-514 

Scherer, H. P-104 

Schilling, A. M. P-67 

Schirf, B. O-157 

Schleck, C. O-51 

Schlieter, M. P-91 

Schmalbrock, P. O-186, P-52, P-53 

Schmid, M. O-371 

Schmidler, J. O-250 

Schnyder, P. O-376 

Schoenberg, A. O-116 

Schoenberg, S. P-47 

Schorlemmer, C. O-498, S-9 

Schramm, P. O-180, O-234, P-91 

Schrey, M. O-234 

Schröder, J. O-249 

Schrom, T. P-104 

Schultze-Haakh, H. E-3 

Schumacher, H. C. O-279 

Schumacher, J. O-395 

Schwamm, L. H. O-112 

Schwartz, E. D. I-459, O-358 

Schwartz, M. B. P-12 

Schwartzmann, R. O-370 

Scott, B. E. S-90 

Scott, J. N. O-152 

Scott, M. L. J. O-325 

Scoutt, L. S-71 

Segal, S. P-13 

Segall, H. D. I-3 

Seifert, P. P-68 

Sellar, R. J. P-130, S-72 

Selva, L. O-298 

Semnic, R. O-245 

Sen, A. O-503 

Senda, M. O-302 

Sennaroglu, L. E-25 

Seol, H. Y. S-38 

Seri, I. O-434, O-450 

Sethi, N. O-324 

Setton, A. O-50, O-218 

Sevdalis, E. O-324 

Shah, J. P. O-78 

Shah, L. M. O-224, O-425, S-37 

Shah, P. N. S-1 

Shah, T. C. E-7, O-218 

Shah, V. S-55 

Shaibani, A. O-157, O-321, O-423, O-505 

Shapiro, M. D. O-309, O-323 

Sharma, D. K. O-356 

Sharma, M. O-156 

Sharma, R. O-288 

Shatzkes, D. R. I-124, O-28 

Sheah, K. S-48 

Sheedy, S. P. O-109 

Sheikov, N. O-176 

Shekdar, K. V. O-389, O-398 

Shelef, I. O-504, P-70 

Shelton, C. O-24 

Shen, C. E-36 

Shen, P. O-280 

Shepherd, T. M. O-363 

Sheridan, C. O-441 

Sherman, P. M. O-84 

Shetty, R. S. P-18 

Shibata, D. K. O-248, O-289 

Shimizu, S. S-60 

Shin, W. O-157, O-321, O-423 

Shirasaka, T. P-24, P-26, P-30 

Shounakh, P. S-89 

Shownkeen, H. O-38 



Shreiber, R. O-214, O-377, O-384 

Shroff, M. O-30, P-98, P-182 

Shumsky, J. S. O-358 

Shy, C. G. S-2 

Shy, M. P-154 

Sibtain, N. A. P-119, S-50 

Siddiqi, J. P-121 

Sidhu, R. P-180 

Silbergleit, R. P-21, S-27 

Silvaggio, J. A. O-92, O-146 

Silvennoinen, H. M. P-1 

Silver, F. O-113, O-114, O-117 

Silvera, V. M. O-315 

Simko, M. O-467 

Simon, E. M. I-512, O-318, O-398, 

O-486 

Simonetta, A. O-305 

Simonetti, G. P-6, P-17, P-19, P-56, 

P-156, P-168 

Simonetti, O. P. O-82 

Sina, C. O-174 

Singel, S. A. O-171 

Singla, R. O-86, S-7 

Sinha, A. O-288 

Sinha, U. E-3, O-246, O-298 

Sintini, M. P-60 

Sit, S. P. O-118 

Sitoh, Y-Y. O-509, P-61 

Sitton, C. W. P-74 

Sivakumar, D. O-304 

Sklar, E. O-400B 

Skrap, M. S-26 

Slivka, A. P. P-147, P-148 

Sluzewski, M. O-411 

Smith, J. K. E-4, O-230, P-36, P-44 

Smith, S. A. O-191 

Smith, W. S. O-121 

Smoker, W. R. K. E-12, E-13, E-14, 

E-16, E-18, E-27, E-31, 

I-67, O-391, S-6, S-15, 

S-47, S-54, S-65, S-68, 

S-80, S-86 

Smyth, M. O-452 

Snead, III, O. C. I-275 

Sneade, M. O-94 

Snyder, A. Z. P-152 

Söderman, M. O-177 

Soellinger, M. P-103 

Soinne, L. P-1 

Soliman, N. O-103 

Solle, M. O-236 

Solymosi, D. O-185 

Som, P. M. E-20, E-24, P-117, S-45, S-52 

Somuncu, I. P-35 

Son, Y-I. P-112 

Sonnier, H. L. S-3 

Soohoo, S. P-10 

Sorensen, A. G. I-407 

Soto, S. O-498 

Sourour, N. O-97 

Souweidane, M. M. O-472 

Spampinato, M. V. E-22, E-40, O-230, 

S-51 

Spearman, M. P. O-400 

Speck, L. M. P-9 

Speck, U. O-119 

Spelle, L. O-96 

Spies, S. S-41, S-42 

Spiga, M. G. O-17 

Spitzer, E. M. S-90 

Sprengers, M. E. O-411 

Srinivas, J. S. O-507 

Srinivasan, A. O-156, S-32, S-36 

Srinivasan, A. V. P-169 

Stambuk, H. E. O-78 

Stancanello, J. S-26 

Stantoun, M. P-175 

Stanzione, P. P-19 

Stashinko, E. O-454 

Stecco, A. O-475 

Stefani, M. A. O-96 

Steinberg, G. K. O-91 

Steiner, C. O-178 

Stephani, U. O-365 

Sternau, L. O-400B 

Stevens, A. O-185 

Stevens, S. I-205 

Stieg, P. E. I-8 

Stippich, C. O-187, P-87, P-91 

Stockmann, T. M. O-123 

Stone, L. O-327 

Storm, Jr., P. B. O-389 

Stroman, P. W. O-360 

Strother, C. M. P-72, P-123 

Struffert, T. O-45, O-169 

Stys, P. O-156 

Su, A. E-33 

Suarez, J. I. O-116 

Sudhir, K. S-34 

Suh, D. C. O-100 

Suh, S. H. P-133 

Suh, S. I. S-38 

Summers, P. E. O-145 

Sundgren, P. C. O-238, P-41, P-83, S-81 

Sunenshine, P. J. O-366, O-500 

Sunshine, J. L. I-5, O-116 

Supprian, T. O-169 

Suri, S. O-26, O-34 

Sussman, M. S. I-129 

Sutcliffe, T. L. P-98 

Sutton, L. N. O-318 

Suzuki, IV, Y. O-164, P-142 

Suzuki, K. O-83, P-105, P-139, P-69 

Svetel, M. O-245 

Sviri, G. O-319 

Syed, M. O-430 

Symons, S. P. O-55, O-56, O-57, O-58, 

O-59 

T 

457 

Tabar, V. O-485 

Tae, K. P-111 

Tai, A. W-H. O-375 

Takahama, S. P-30 

Takao, H E-26, O-46, P-126, P-138 

Takase, Y. S-4, S-11 

Takeda, K. P-49, S-14, S-44 

Takehara, Y. O-383 

Takemoto, S. O-361 

Takhtani, D. O-84, O-502 

Taki, T. S-56 

Tali, T. E. P-177 

Talmi, D. O-306 

Tampieri, D. O-373, P-131, P-179 

Tamura, S. P-120 

Tan, D. S-48 

Tan, T. Y. S-65, S-68 

Tanabe, J. L. P-86 

Tanaka, H. P-24 

Tanaka, R. S-35, S-39 

Tanenbaum, L. N. P-33 

Tang, P. H. S-65, S-68 

Tantivatana, J. O-13 

Tarantini, S. P-60 

Tarr, R. W. O-116 

Tartaglione, T. O-490 

Tartaro, A. O-490, P-45, S-19 

Tasbas, B. O-484 

Taschner, C. A. P-80 

Tay, A. S-48 

Tayfun, C. P-35, P-37 

Taylor, K. O-113, O-114 

Taylor, M. J. P-99 

Taylor, R. A. O-18 

Tech, K. S-27 

Tegeler, C. O-159 

Teitelbaum, G. P. O-122, P-73 

Tejada, J. G. O-18 

Telian, S. A. O-27 

Ter Minassian, A. P-5 

Terada, H. S-14 

terBrugge, K. G. O-47, O-152, O-217, 

O-504, P-70 

Tercan, F. P-4 

Thacker, N. A. O-325 

Thakkar, C. O-508 

Thakur, S. P-102 

Theng, J. S-48 

Thennarasu, K. O-304 

Thielen, K. R. O-51, O-475A 

Thirunavuukarasuu, A. E-32 

Thoma, R. J. P-94 

Thomann, P. O-249 

Thomas, A. K. P-151 

Thomas, B. P-159, P-160 

Thomas, E. J. P-101 

Thomas, S. O-278 

Thompson, C. O-291 

Thompson, P. O-297, P-23 

Thorell, W. O-381 

Thron, A. O-213, P-135 

Thulborn, K. R. P-95 

Thurber, A. B. P-77 

Thurnher, M. M. O-352 

Tian, W. O-364 

Timmerman, R. D. O-314 

Ting, G. P-175 

Tintoré, M. O-369 

Tirschwell, D. L. O-439 

Tisnado, J. P-143 

Tittgemeyer, M. P-54 

Tjauw, I. O-351, O-354 

Tkach, J. A. P-82 

Toga, A. O-297, P-23 

Tokgoz, N. P-177 

Tolosa, E. O-243, O-244 

Tomitaka, E. P-69 





Tomlin, H. A. O-371 

Tomlinson, G. O-152, O-419 

Tomsick, T. A. O-153, O-158 

Tong, K. A. O-441, P-170 

Torii, R. P-75 

Toyoda, K. S-60 

Traina, J. O-375 

Tran, T. E-19 

Tronnier, V. O-180, O-187, P-91 

Tropine, A. O-188 

Trout, A. T. O-429, O-463, O-465, 

O-475A 

Tsai, F. Y. S-2 

Tsai, Y-H. P-128 

Tschampa, H. J. O-251 

Tsuchiya, K. O-215, O-359, O-361 

Tsui, W. P-13 

Tsukahara, H. O-285, S-14 

Tulloch, K. O-324 

Türe, U. P-46, P-96 

Turk, III, A. S. O-19, O-149, O-382 

Turski, P. A. I-263, O-19, O-382 

Tuvia, K. O-294 

Tyszka, J. M. I-335 

U 

Ucar, M. P-177 

Uchida, K. P-93 

Uchida, N. P-93 

Uchino, A. S-4, S-11 

Ucoz, T. P-35, P-37 

Uemura, A. O-466, S-91 

Ugurel, M. S. O-18 

Ulmer, J. L. I-194 

Ulmer, S. O-365, P-42 

Unal, S. S. O-227 

Unger, S. O-392 

Urbach, H. O-251 

Urena, R. J. O-99 

V 

Vachha, B. O-451 

Valanne, L. K. P-1 

Valentine, A. O-508 

Valentino, D. E-10 

Vallee, J-N. P-124 

Vallone, I. M. E-1 

Vallone, S. O-427 

Van Effenterre, R. O-97 

van Leeuwen, M. O-111 

van Melle, G. O-376 

van Rijn, J. C. O-411 

van Rooij, W. J. J. O-411 

van Zijl, P. C. M. O-454 

Vandenberg, S. O-231 

Vavilala, M. S. O-439 

Vazquez, Sr., E. E-21 

Velthuis, B. O-111 

Veltkamp, D. O-478 

Venkatasubramanian, G. O-507 

Venturi, C. E-1 

Vezina, G. O-292 

Viaño López, J. P-134 

Vibhute, P. G. P-117 

Vigneron, D. B. O-219, O-329, O-374 

Vilar-Lopez, R. P-57 

Villablanca, J. P. O-123, O-412, O-416, 

P-81 

Vincent, D. O-220 

Vinuela, F. E-26, O-416 

Viohl, I. O-178 

Volkau, I. E-2, E-32 

von Cramon, D. Y. P-54 

von Tengg-Kobligk, H. O-249 

Vosburgh, K. O-20 

Voyvodic, J. O-224 

Vykhodtseva, N. O-176 

W 

458 

Wada, A. P-93 

Wagner, A. L. O-426, O-469 

Wagshul, M. E. O-480, O-482 

Wakana, S. O-191, O-454 

Wakhloo, A. K. O-17, O-95, O-147, 

P-132 

Walker, M. T. O-54, O-82, O-321, 

O-423, O-503, P-51 

Walker, S. A. P-86 

Wallace, E. E-23 

Wallace, M. A. O-88, O-166 

Walters, T. D. O-55, O-56, O-57, O-59 

Walz, D. M. O-50, S-18 

Wang, A-M. P-21, S-27 

Wang, E. O-488 

Wang, G. W. O-116 

Wang, H. S-34 

Wang, M. C. O-439 

Wang, P. Y. O-283, O-307, O-322, 

O-351, O-354 

Wang, X. M. P-163 

Wang, Y-H. P-128 

Warren, H. G. O-386 

Wasudev, N. O-309, O-323 

Watanabe, A. T. I-344, P-181 

Watanabe, K. O-285, S-79 

Watanabe, Y. P-24, P-26, P-30 

Watson, T. D. E-18 

Weber, C. O-99 

Weber, W. O-209 

Webster, K. A. O-17 

Weeks, Jr., J. K. O-397, O-455 

Wei, Sr., C. P-65 

Weigele, J. B. O-89, O-378, P-8, P-29 

Weill, A. O-92, O-146, O-165 

Weiller, C. O-115 

Weinberger, E. E-39 

Weingart, J. O-182 

Weir, R. P-74 

Weisend, M. P. P-94 

Weissman, J. L. P-55 

Welker, K. M. O-77, O-109 

Welsh, C. T. O-487 

Welsh, R. C. O-238 

Wenig, B. M. I-200 

Weon, Y-C. P-112, P-161, S-61 

Weprin, B. O-496 

Westerlund, E. O-433 

Westesson, P-L. O-328, O-364, P-180, 

S-21, S-46, S-53, S-57 

Westin, C-F. O-20 

Westman, D. A. O-120 

Weybreight, P. O-238, P-41 

Whitby, E. O-316, O-317 

White, D. A. P-152 

White, M. L. O-33, O-300, S-6, S-15, 

S-66 

White, M. M. S-47 

White, P. M. P-130, S-72 

Whitlow, C. O-159 

Widjaja, E. O-316, O-317 

Wiggins, III, R. H. I-11, I-8, I-14, 

I-27, I-34, I-50, 

I-58, O-24, O-27 

Wignall, E. O-249 

Wilde, E. O-440 

Wilkinson, I. D. O-216, O-249, O-491 

Williams, D. O-309, O-323 

Williams, M. O-228 

Williams, S. P-68 

Williams, T. O-399 

Willinsky, R. A. O-47, O-113, O-152, 

O-419, O-504, P-70 

Wilmink, J. T. O-213, P-135 

Wilson, G. J. O-351 

Wilson, J. S-27 

Wintermark, M. O-111, O-121, O-320, 

O-376 

Wirth, III, E. D. I-406 

Witte, R. J. E-37, O-31, P-118 

Wojak, J. C. O-160 

Wolansky, L. J. O-324 

Woldenberg, R. F. E-7, O-50, O-218 

Wolf, K. J. P-67 

Wolf, R. L. O-295 

Wong, A. M-C. P-108 

Wong, E. T. O-239 

Wong, W. H. M. I-255 

Woo, H. O-381 

Woo, J. H. O-295, P-84, S-1 

Wood, F. B. O-221, O-222 

Woodhams, R. P-28 

Wu, J. O-239 

Wu, K-D. P-16 

Wu, R. H. P-58, O-85, O-217 

Wu, V. P-16 

Wu, X. E-38 

Wu, Y. O-471, O-481 

Wuppalapati, S. O-448 

X 

Xavier, A. R. O-90, O-98, O-155 

Xiao, Z. W. O-85 

Xu, D. O-219, O-329 

Xu, W-J. O-189 

Y 

Yachnis, A. T. O-363 

Yagishita, A. S-13 



Yakes, W. F. I-64, O-15, O-87 

Yakinci, Sr., C. O-476 

Yamada, J. S-85 

Yamada, M. S-35, S-39 

Yamashita, S. E-8, O-379, O-383, 

O-466, S-91 

Yamashita, T. S-12 

Yamashita, Y. O-181, P-31, P-34, 

P-69, P-71, P-171, S-12 

Yamura, M. O-181, O-444, P-31, 

P-34, P-71 

Yang, M. O-186, P-52 

Yange, W. O-186 

Yano, T. P-120 

Yen, P-S. P-92 

Yen, T. C. P-108 

Yeshraj, G. O-304 

Yetkin, Z. P-116 

Yildirim, N. P-32 

Yildirim, T. P-4 

Yim, C. O-246 

Ying, J. O-314 

Ymashita, Y. O-444 

Yokota, A. P-158 

Yoo, W-J. P-78 

Yoon, H-K. P-150 

Yoshimatsu, S. P-69 

Young, A. B. S-24 

Young, R. J. O-28, O-183 

Yu, C. P-50 

Yuan, Jr., F. P-65 

Yuhan, J. P-181 

Z 

Zagzag, D. O-184, O-488, O-489 

Zaidat, O. Z. O-116 

Zak, I. S-7 

Zakaris, E. P-172 

Zarnow, D. M. O-318 

Zauner, M. S-16, S-31, S-76 

Zeccolini, F. O-428 

Zee, C. S. E-19, O-241, P-50, P-73, P-76 

Zerin, M. J. O-453 

Zerr, I. O-251 

Zeumer, H. O-115 

Zhan, J. P-13 

Zhang, H. P-2 

Zhang, Y. O-33, O-300, S-66 

Zhao, Y. P-50 

Zheng, W. B. P-58 

Zhong, C. O-252 

Ziegert, A. J. E-14 

Zimmerman, R. A. I-337, O-318, O-389, 

O-398, O-486, O-497 

Zimmerman, R. D. I-334 

Ziolko, S. K. O-225 

Zou, P. O-455 

459 





Notes: 



ASNR 43 rd Annual Meeting 


Metro Toronto Convention Center, Toronto, Ontario, Canada 

SESSION 12 - NEURONEWS SCIENTIFIC PAPERS 

Paper NN1 Starting at 1:00 PM, Ending at 1:08 PM 

Monday, May 23 1:00 pm – 2:40 pm 

ABSTRACTS 

A Systematic Literature Review of Magnetic Resonance Spectroscopy (MRS) for the Characterization of Brain 

Tumors 

Hollingworth, W. 1 ·Medina, L. S. 2 ·Shibata, D. K. 1 ·Lenkinski, R. E. 3 ·Bernal, B. 2 ·Zurakowski, D. 4 ·Comstock, B. 1 ·King, 

A. 1 ·Jarvik, J. G. 1 

1 University of Washington, Seattle, WA, 2 Miami Children's Hospital, Miami, FL, 3 Beth Israel Deaconess Medical Center, 

Boston, MA, 4 Harvard Medical School, Boston, MA, 5 University of Washington, Seattle, WA. 

Purpose: In January 2004, the Centers for Medicare & Medicaid services (CMS) decided to continue its national nonreimbursement 

policy for Magnetic Resonance Spectroscopy (MRS) for the characterization of brain tumors. This 

determination was based, in large part, on two reviews summarizing the literature in 2002. The purpose of this study is to 

provide an updated systematic literature review and develop methodological guidelines for future research. 

Materials & Methods: We developed a structured search strategy for the MEDLINE, EMBASE and Cochrane databases 

to identify articles published between 1/1/2002 and 12/31/2004. Two reviewers independently judged abstracts based on 

five exclusion criteria: 1) not 1 H MRS; 2) not focused on brain tumors; 3)


3.0-T MRI of the Pituitary : Lessons Learnt From Our 110 First Patients 

Bonneville, J. 1 ·Bonneville, F. 1 ·Chayep, C. 1 ·Oulton, S. 1 ·Calmon, G. 2 ·Cattin, F. 1 

1 University Hospital of Besançon, Besançon, FRANCE, 2 GE Healthcare Technologies, Velizy, FRANCE. 

Purpose : Magnetic resonance imaging has emerged as the method of choice for diagnostic imaging of the pituitary gland 

and sellar region. Compared with MRI at 1.5T, high-field MRI scanners operating at 3.0T provide higher SNR with higher 

spatial resolution, better image quality and acceptable acquisition times. Notable limitations of high-field MRI are reduced 

T1 contrast, increased magnetic susceptibility and chemical shift and quadrupled SAR. The goal of this study is to 

evaluate the advantages and drawbacks of 3.0T versus1.5T MRI in the diagnosis of pituitary lesions. 

Materials and methods : 110 patients with a pituitary lesion have been examined on a Signa Excite 3.0T MR scanner 

(GE Healthcare) equipped with the 8-channel head coil. There were 86 pituitary adenomas (35 prolactinomas, 18 

gonadotropic adenomas, 16 GH-secreting, 8 corticotropic, 7 non-secreting and 2 TSH-secreting adenomas), 8 Rathke 

cleft cysts, 5 craniopharyngiomas, 4 meningiomas of the sellar region, 1 intracavernous internal carotid artery aneurysm, 

4 empty sellae and 2 granulomas of the infundibulum. In all patients, coronal T1 and T2-weighted images and T1 after 

gadolinium injection were performed. Axial and sagittal sequences were obtained in selected cases. 

Results : Coronal FSE T2-W sequences provide exquisite delineation of the small anatomic structures of the pituitary 

gland and of the cavernous sinus. Coronal T1-weighted SE sequence provides poor image contrast and is sensitive to 

susceptibility artifact. Consequently image distortions are seen mainly in the presence of extensive pneumatization of 

sphenoid sinus. T1-weighted Fast FLAIR or 3D gradient echo sequences can be alternative techniques depending on the 

sphenoid anatomy. Magnetic susceptibility is particularly cumbersome when looking for vasopressine storage on axial SE 

T1 sections. Fat sat sequences are promising here. Finally, SAR issue was not noticeable in this study. 

Figure 1 : 3.0-T coronal Fast SE T2-weighted image. Right intracavernous residual adenomatous tissue after 

removal of a GH-secreting pituitary adenoma. 

Conclusion : 3T imaging provides together advantages and limitations for examination of the pituitary region. 

Optimization of technical parameters is of paramount importance to benefit from this new exciting technology. Figure 1 : 

3.0-T coronal Fast SE T2-weighted image. Right intracavernous residual adenomatous tissue after removal of a GHsecreting 

pituitary adenoma. 

Keyword: Pituitary Gland ; Pituitary tumors ; 3.0T-MR


Intravenous and Intra-arterial Dynamic Susceptibility-weighted Contrast-enhanced Perfusion MR Imaging of 

Meningiomas Before and after Embolization 

Cha, S.·Martin, A.·Weber, O.·Parsa, A. T.·McDermott, M. W.·Berger, M. S.·Dillon, W. P.·Higashida, R.·Saloner, D. 


San Francisco, CA, 

Purpose: Meningiomas are the most common primary extra-axial brain tumors that are amenable for cure when complete 

surgical resection is possible. Transarterial embolization of meningiomas is widely accepted as a standard of care to 

reduce intraoperative blood loss and perioperative bleeding complication. However, embolization can only be done safely 

in vessels supplied by external carotid artery (ECA) or other dural branches of intracranial vessels. Furthermore, the 

success of embolization can be difficult to assess on conventional anatomical imaging alone. Dynamic susceptibilityweighted 

contrast-enhanced (DSC) perfusion MR imaging (pMRI) has been shown to correlate with tumor vascularity and 

be sensitive to changes in microvascular environment. The purpose of our study was to use perfusion maps derived from 

DSC pMRI to characterize quantitative and spatial distribution of vascular supply to meningiomas before and after 

embolization. We hypothesized that intravenous and intra-arterial DSC pMRI can accurately depict the topographic 

distribution of different vascular supply to meningiomas and serve as arbiter of successful embolization. 

Materials and Methods: We prospectively studied 11 patients with meningioma using intravenous (IV) DSC pMRI 

immediately before and after transarterial embolization. In four of 11 patients, we also performed intra-arterial (IA) pMRI 

by injecting dilute Gd-DTPA directly into the ECA, internal carotid artery (ICA), and/or vertebral artery (VA) that supply the 

meningioma. We measured cerebral blood volume (CBV), cerebral blood flow (CBF), and mean transit time (MTT) of the 

tumor before and after embolization. 

Results: Pre- and post-embolization DSC pMRI clearly depicted the regions of devascularization and areas of residual 

tumor hypervascularity following embolization as shown in figure below. The intra-arterial pMRI of ECA, ICA, or VA 

confirmed the spatial distribution of pial and dural arterial supply to the tumor. Among the hemodynamic variables, CBF 

followed by MTT showed the most significant change between pre- and post-EMBO DSC pMRI where CBF increased and 

MTT decreased following embolization. The overall tumor CBV did not show significant change.

Conclusions: Our preliminary results suggest that DSC pMRI provides quantitative and spatial information on different 

arterial supply to meningiomas and hemodynamic changes following embolization. We found that DSC pMRI can clearly 

depict areas of devascularization and residual tumor vascularity following embolization and hence serve as arbiter of 

successful embolization. 

Keyword: Meningioma ; Embolization ; Perfusion MR imaging 


Does High-field MRI have an Influence on the Classification of Patients with Clinically Isolated Syndromes 

suggestive of Multiple Sclerosis according to Imaging and Diagnostic Criteria? 

Wattjes, M. P.·Harzheim, M.·Lutterbey, G.·Gieseke, J.·Kuhl, C. K.·Klotz, L.·Tschampa, H. J.·Klockgether, T.· 

Schild, H. H. 

University Hospital Bonn 

Bonn, GERMANY 

Purpose: To investigate the detection rates of inflammatory lesions in patients with clinically isolated syndromes (CIS) 

suggestive of multiple sclerosis (MS) at 1.5T and high-field MRI operating at 3.0T with respect to the classification 

according to Barkhof MRI and McDonald diagnostic criteria. 

Methods: We performed an intraindivdual comparative study in patients with the clinical diagnosis of CIS suggestive of 

MS. EDSS and cerebrospinal fluid results were obtained from every participant. Imaging at both magnetic field strengths 

were performed after completion of high-dose intravenous corticosteroid therapy in two sessions separated by 24 hours. 

The imaging protocol included contiguous axial slices of the following pulse sequences: T2 Turbo Spinecho (1.5T: TE 

100ms, TR 3500ms; 3.0T: TE 100ms, TR 4100ms), FLAIR (1.5T: TE 110ms, TR 6000ms, TI 2000ms; 3.0T: TE 140ms, 

TR 12000ms, TI 2850ms) und T1 Spinecho (1.5T und 3.0T: TE 12ms, TR 500ms) before and after intravenous injection of 

gadolinium-DTPA. For both field strengths a constant voxel size (0.9x0.9x5mm) was used. Image analysis were 

performed by two neuroradiologists in consensus. High signal white matter lesions with a size of ≥3mm were counted and 

categorized according to their anatomic location in infratentorial, callosal, juxtacortical, periventricular and other white 

matter. Assessment of the fulfilled Barkhof MRI and McDonald diagnostic criteria for both field strengths in every patient. 

Results: 40 patients (10 male, 30 female, 18-55 years, EDSS 0-3.0) were included. The higher detection rate at 3.0T 

resulted in an increased number of fulfilled Barkhof MRI criteria in 11 patients. Among those, 2 patients crossed the 

diagnostic cut off point of 3 out of 4 MRI criteria leading to the diagnosis of dissemination in space according to the 

McDonald diagnostic criteria. 

Conclusion: The higher detection rate of inflammatory lesions at high-field MRI has an substantial influence on the 

classification of patients with CIS according to MRI and diagnostic criteria with considerable consequences for the 

diagnostic work-up and prognostic classification. 

Keyword: high-field MRI ; clinically isolated syndrome ; multiple sclerosis 

Henriette Tschampa, MD, co-author, will be presenting the paper 


fMRI as a Biomarker for Assessing Therapeutic Response in Parkinson’s Disease 

Elsinger, C. L. 1,2 ·Zimbelman, J. 2 ·Jaradeh, S. S. 2 ·Rao, S. M. 2 

1 Neurognostics, Inc., Milwaukee, WI, 2 Medical College of Wisconsin, Milwaukee, WI. 

Purpose: The objective of this study was to test the efficacy of functional magnetic resonance imaging (fMRI) as a 

biomarker for quantifying therapeutic response in Parkinson’s Disease (PD). PD patients have difficulty initiating complex 

sequential movements. Taking advantage of the high temporal resolution of event-related fMRI, we segregated brain 

regions involved in motor planning from those involved in motor execution. By concentrating on premovement brain 

activation patterns, we interrogated brain regions affected by PD. We hypothesized that PD patients off drug would 

demonstrate underactivation of medial premotor circuits, and that underactivation would be partially reversed with

dopamine replacement medications [1;2]. A secondary goal was to measure the sensitivity of fMRI as a biomarker relative 

to standard clinical outcome measures, such as the Unified Parkinson’s Disease Rating Scale (UPDRS). 

Materials & Methods: Six healthy older subjects and 5 PD patients diagnosed with idiopathic PD were scanned twice 

separated by one week. During the OFF session, PD patients refrained from their usual antiparkinsonian medications for 

24 hours prior to scanning. In this event-related fMRI study, subjects were imaged on a 3 Tesla scanner while performing 

a motor sequencing task. Two stimulus conditions were used, a visually-guided and an internally-guided condition. Motor 

sequences were either simple (repetitive) or complex (heterogenous). 

Results: FMRI analyses were performed on a single region of interest, the supplementary/cingulate motor region 

(SMA/CMA), because of its critical role in motor planning. As hypothesized, PD patients off drug demonstrated an 

underactivation of the SMA/CMA during the premovement, planning phase of the trial. Healthy controls exhibited 

activation within the SMA/CMA during the premovement phase of the complex motor sequencing trials. In contrast, PD 

patients off drug exhibit no activated voxels. On drug, PD patients demonstrated a partial normalization of brain activity, 

although the spatial extent of activation was not as great as in healthy controls. Behavioral performance measures and 

UPDRS scores did not differ as a function of drug state. 

Conclusion 

Task-activated fMRI is a sensitive biomarker to disease pathology in early PD. As predicted, fMRI response demonstrated 

maximal sensitivity in discriminating healthy subjects from PD-OFF patients during motor planning. Second, task-activated 

fMRI was sensitive to therapeutic intervention, even with a short ‘off drug’ period of 24 hours. FMRI response, as 

measured by activity in the SMA/CMA, was shown to be more sensitive to therapeutic drug state than the “gold standard” 

clinical outcome measure, the UPDRS. 

Keyword: fMRI ; Parkinson's disease ; therapeutic biomarker 

Supported by NIH/NINDS R43 NS049705-01


INTRA-OPERATIVE REAL-TIME ASSESSMENT OF FLOW DYNAMICS IN INTRACRANIAL ANEURYSM USING 

CONTRAST-ENHANCED ULTRASOUND: PRELIMINARY RESULTS 

Hoelscher, T. 1 ·Singel, S. 1 ·Wilkening, W. G. 2 ·Jiang, H. 3 ·Mattrey, R. F. 1 ·Kerber, C. W. 1 ·Sang U, H. 1 

1 University of California, San Diego, San Diego, CA, 2 Ruhr Center of Excellence for Medical Engineering (KMR), Bochum, 

GERMANY, 3 Siemens Medical Solutions Inc., Issaquah, WA 

Rationale and Objectives: We present data of real-time imaging of flow dynamics of intracranial aneurysms acquired 

during neurosurgical intervention, using intra-operative contrast-enhanced ultrasound. 

Methods: Under local IRB approval, three patients with known aneurysms were studied during the neurosurgical 

intervention. Readings were taken through the skull defect at three different time points: before and after opening of the 

dura mater, and after clipping of each aneurysm. A Siemens Sonoline Antares ultrasound machine equipped with a 

9MHz linear array transducer (VFX13-5SP) was used. Optison (Amersham GE Healthcare, USA) was used as an 

ultrasound contrast agent (UCA). 

Results: In all three cases the aneurysms could be visualized after UCA injection using a phase inversion technique. 

Single microbubbles could be tracked within each aneurysm enabling the visualization of real-time flow dynamics. 

Regions of high- and low-flow as well as turbulent flow within each aneurysm could be identified. The flow velocities in the 

afferent and the downstream vascular segments as well as inside the aneurysm were assessed. After clipping of the 

aneurysms, the same technique was used to monitor successful positioning of the aneurysm clips. 

Figure 1: Transdural insonation of an aneurysm before UCA injection.

Figure 2: Visualization of single microbubbles in the early (left) and a later phase (right) after IV UCA injection. 

Conclusion: For the first time the flow dynamics of intracranial aneurysms can be visualized in real-time during the 

neurosurgical treatment of these lesions. Using an advanced contrast-specific ultrasound technique, the pathological flow 

patterns inside the aneurysm can be studied. The flow velocities and directions can also be determined at specific 

locations within the aneurysm sac, the feeding artery and its downstream supply area. The same technique can also be 

used to monitor immediately the success of aneurysm clipping. Conclusion: For the first time the flow dynamics of 

intracranial aneurysms can be visualized in real-time during the neurosurgical treatment of these lesions. Using an 

advanced contrast-specific ultrasound technique, the pathological flow patterns inside the aneurysm can be studied. The 

flow velocities and directions can also be determined at specific locations within the aneurysm sac, the feeding artery and 

its downstream supply area. The same technique can also be used to monitor immediately the success of aneurysm 

clipping. 

Keyword: intracranial aneurysm ; contrast ultrasound ; flow dynamics 


Neuroimaging Findings Specific for Bipolar Disorder in Children: Quantitative Structural and Diffusion Tensor 

Imaging 

Lipton, M. L. 1,2,3 ·Papolos, D. 4 ·Lombard, J. 5 ·Nierenberg, J. 3 ·Hoptman, M. 3 ·Yhu, S. 2 

1 Montefiore Medical Center, Bronx, NY, 2 Albert Einstein College of Medicine, Bronx, NY, 3 The Nathan S. Kline Institute for 

Psychiatric Research, Orangeburg, NY, 4 Juvenile Bipolar Research Foundation, Maplewood, NJ, 5 The Brain Behavior 

Center, Nyack, NY Bronx, NY. 

Purpose: Bipolar disorder (BD) manifests in children as a severe behavioral disturbance characterized by paroxysmal 

outbursts of disproportionate rage separated by intervening periods of more stable mood with or without depressive 

symptoms. A major clinical problem in child psychiatry is the differentiation of children with BD from those with ADHD. 

Misdiagnosis is common and incorrect treatment of children who in fact have bipolar disorder may lead to poor outcomes. 

We have evaluated a group of patients meeting research diagnostic criteria for the pediatric bipolar core phenotype. The 

aim of this study is to determine if asymmetry of limbic brain structures occurs in children with BD and if these findings are 

related to abnormalities of white matter integrity as shown by diffusion tensor imaging (DTI). 

Materials and Methods: 10 children (ages 8-16) were identified by a behavioral neurologist and referred for imaging at 

1.5 T. Standardized research-clinical evaluations confirmed the diagnosis of Pediatric BD core phenotype. Whole brain 

volumetric imaging was accomplished using a 3D FSPGR acquisition acquired at 1 mm 3 isotropic resolution. Quantitative 

volumetric measurement of amygdala, hippocampus and cingulate gyrus was performed after segmentation of 

grey/white/CSF by experienced raters using validated criteria and 3D SLICER. Whole brain DTI was acquired with b=1000 

and 25 independent diffusion directions. Regions of interest were placed in the frontal white matter for assessment of 

mean fractional anisotropy (FA). Asymmetry was assessed using paired samples T-test.

Results: Initial review of brain images by a CAQ certified Neuroradiologist identified striking asymmetry of mesial 

temporal structures. Quantitative measurements confirmed statistically significant asymmetry of amygdala (p/= 8 points or score 0-1 on NIHSS, and mRS 0-2 and BI 75-100) 

at day 90; and signs of reperfusion on 4-8 hr post-treatment MRI (reduction of the PWI deficit >/= 30% or at least 2 point 

TIMI-scale MRA improvement in a previously occluded artery). 

Results: Acute comprehensive MR examinations were rapidly performed and well tolerated. Only one sICH occurred 

(90µg/kg tier of DIAS), yielding a combined sICH rate of 1.7% (1/59) for doses of 90µg/kg and 125µg/kg. Mortality was 

low, at 5.3% (5/94), without any group difference. The reperfusion rates were: placebo: 23.5% (8/34), 90µg/kg: 34.6% 

(9/26), 125µg/kg: 62.1% (18/29). Reperfusion was a sensitive (65.7%) and specific (77.8%) predictor for positive clinical 

outcome (p


CT and MR Imaging of Carotidynia 

Esfahani, F.·Baer, A. N.·Delbalso, A. M. 

University at Buffalo 

Buffalo, NY, 

Purpose: Carotidynia is an idiopathic neck pain syndrome associated with tenderness over the carotid bifurcation. This is 

a well-known entity in otolaryngology and neurology literature with a distinct histology (1,2,3). Imaging descriptions of this 

disorder remain scarce. (4,5) We are reporting CT and MR features of a case of carotidynia. 

Material and Methods: A 38 year-old woman was suffering from intermittent pulsating left-sided neck pain for several 

months, which worsened with exertion and head movement and radiated to her jaw. A dental x-ray was non-revealing. 

She presented to the emergency department with tenderness over the left carotid bifurcation. Physical, neurological, and 

laboratory examinations were otherwise normal. Contrast-enhanced neck CT showed abnormal density around the left 

carotid bifurcation and the external carotid artery. Subsequently MRA and gadolinium enhanced MR examination of the 

neck were done. MRI showed fluid around the left carotid sheath and the carotid bifurcation with no evidence of 

hematoma. MRA was normal. 

Discussion: Carotidynia is a clinical pain syndrome described as unilateral neck pain aggravated by movement and 

sometimes by swallowing, chewing and coughing (1). Many of these patients may have CT examination. Detection of any 

abnormality around the carotid bifurcation should raise the possibility of carotid artery dissection. MRI reveals the 

presence of fluid around the carotid artery, which shows gadolinium enhancement. Once diagnosed, carotidynia usually 

responds promptly to anti-inflammatory treatment. 

Legends 

1.a. Axial T1 weighted image shows isointense signal surrounding the left carotid sheath. 

1.b. Axial T2 weighted image shows fluid around the left carotid sheath. 

Conclusion: Carotidynia often presents as unilateral neck pain over the carotid artery. Because the differential diagnosis 

of neck pain remains extensive many of these patients may undergo imaging studies. Contrast enhanced CT scan detects 

abnormality around the carotid sheath, which should be further evaluated by MR imaging. MRI detects fluid around the 

carotid sheath, which helps to exclude dissection and reach the proper diagnosis. 

References: 

1. Emmanuelli JL, Gutierrez JR, Chionssone E. Carotidynia: a frequently overlooked and misdiagnosed 

syndrome. Ear Nose Throat J 1998;77:462-464 

2. Upton PD, Walker Smith JG, Charnock DR. Histologic confirmation of carotidyna. Otolaryngol Head Neck 

Surg 2003;129:443-4 

3. Headache Classification Committee of the International Headache Society. Classification and diagnosis 

criteria for headache disorders, cranial neuralgias and facial pain. Cephalgia 1988; 8(suppl 7):48-49 

4. Buetow MP, Delano MC. Carotidynia. AJR 2001; 177:947-9 

5. Burton BS, Syms MJ, Petermann GW, Burgess LPA. MR imaging of carotidynia. AJNR 2000;21:766-769 

Keyword: Carotidynia ; Carotid dissection ; Differential diagnosis


New Development : Head and Neck Mobile Computed Tomography System for Improved Diagnosis and 

Treatment of Acute Stroke and Trauma 

Choi, I. 1 ·Bailey, E. M. 2 ·Caplan, L. R. 3 ·Hreib, K. K. 1 ·Koroshetz, W. J. 3 ·McDonald, C. T. 4 ·Yonas, H. 5 ·Zervas, N. 3 

1 Lahey Clinic, Burlington, MA, 2 NeuroLogica Corporation, Danvers, MA, 3 Massachusetts General Hospital, Boston, MA, 

4 South Shore Hospital, Weymouth, MA, 5 University of New Mexico Medical Center, Albuquerque, NM 

Purpose: Effective stroke/trauma diagnosis and treatment requires the combination of ultra rapid medical imaging and 

specialized clinical examination skills. Stroke and Trauma are acute conditions where time is a critical factor in optimizing 

patient outcome. Noncontrast CT is the most common and well studied initial imaging modality for acute stroke and 

trauma. Traditionally this requires the patient be brought to a radiology suite. We have developed a solution where the CT 

can be brought to the patient. 

Materials and Methods: We have designed a novel Head and Neck Mobile CT system that overcomes many of the 

roadblocks to ultra rapid CT imaging. This system is placed at or delivered to the point of initial patient contact 

(Ambulance, Aircraft, ICU, ER, OR, Angio Suite), and is controlled by a straightforward, menu driven touch-screen which 

can be easily operated with minimal training. The system, which produces images of a quality equal to modern, fixed CTs 

will provide navigation and hemorrhage detection during surgical and interventional procedures. The system also 

incorporates advanced CT capabilities (CT angiography, CT perfusion, CT fluoroscopy, and Xenon quantitative cerebral 

blood flow). 

Results: The first beta systems have been assembled and are being tested and evaluated for efficacy. The presentation 

will include the first images taken with the new device. The following photograph shows the new CT system performing a 

bedside scan. 

Conclusion: A mobile CT that can be brought to the patient point of care will significantly improve diagnosis and 

treatment of acute neurological conditions. Our beta system shows that the system can scan patients at bedside. The 

product is planned to be submitted for FDA 510k clearance in summer 2005. 

The authors of this work have indicated that they will be discussing/presenting an unapproved or investigative use of 

NeuroLogica’s CereTom. 

Keyword: mobile ; computed ; tomography


A New Embolic Device for the Treatment of Intracranial Aneurysms: Preliminary Experience in a Canine 

Aneurysm Model. 

Dowd, C. F. 1 ·Barnwell, S. L. 2 ·Lee, J. A. 3 ·Calabria, M. F. 3 

1 2 

University of California, San Francisco, San Francisco, CA, Oregon Health Sciences University, Portland, OR, 

3 

NeuroVasx, Inc., Maple Grove, MN. 

Purpose: The authors present a new embolic material for the treatment of intracranial aneurysms. The NeuroVasx APEX 

Polymer Aneurysm Filler was developed for the purpose of providing an embolic device of one continuous strand which 

can be detached at any length and which will provide denser aneurysm packing with fewer complications compared with 

current technology. 

Materials and Methods: The Aneurysm Filler is made of a low-density polyethylene strand. Radiodensity is achieved with 

embedded tungsten. The strand has a hollow core and is delivered over a guidewire through a standard microcatheter 

into the target aneurysm. Detachment is achieved by heating a core-wire and causing melt-separation of the filler material. 

The device was tested to demonstrate achievement of volumetric aneurysm filling, percentage occlusion, material 

compaction or migration, acceptable serum tungsten levels, and histologic response. A Good Laboratory Practices (GLP) 

study was undertaken using a canine side-wall aneurysm model. [48 aneurysms in 29 canines] 

Results: In 48 treated aneurysms (classified as wide necked) the average volumetric filling was 35.6% with a maximum 

volumetric filling of 64.9%. This compares favorably with literature reports concerning volumetric filling with platinum coils 

(22.4-38%). At the time of aneurysm treatment, complete occlusion (as determined angiographically) was achieved in 

4.1%. Although this appears less than that reported in the literature for platinum coils (36.2%), the radiopacity of the Filler 

material may permit more detection of small amounts of remnant contrast opacification of an aneurysm in comparison to 

more radiodense platinum coils. Notably, 100% of subtotally-occluded aneurysms showed angiographic improvement in 

percentage occlusion, such that at final follow up (2 weeks, three-months, six-months), 35% of aneurysms showed 

complete occlusion, comparing favorably to current platinum coil technology. Only one aneurysm (of 48) demonstrated 

compaction of Filler material, and there were no instances of material migration after placement. Blood serum Tungsten 

levels were measured in a subset of 18 animals and were within the accepted range of normal prior to treatment and at 

follow up time points. Histologically, the Filler material generated similar tissue response to that of platinum coils and of 

Prolene suture material. Tissue response at the three follow up time points was virtually identical. 

Conclusion: We have developed and tested a new embolic material for aneurysm therapy which compares favorably to 

current technology in aneurysm occlusion and which permits the potential for single-device deployment and detachment 

at any length. Interval improvement in aneurysm occlusion over time was observed. 

Keyword (Complete): aneurysm ; embolization ; endovascular 


NEUROFORM STENT-ASSISTED COILING OF RUPTURED INTRACRANIAL ANEURYSMS: PRELIMINARY SAFETY 

DATA 

Chaloupka, J.·Taylor, R. A.·Hayakawa, M.·Kothimbakam, R. 

University Iowa Hospital & Clinics 

Iowa City, IA. 

Purpose: To assess the potential risk of Neuroform stent-assisted coiling to treat acutely ruptured wide-necked saccular 

and/or fusiform aneurysms. 

Materials and Methods: We retrospectively reviewed all patients at our institution that had an acute subarachnoid 

hemorrhage treated with Neuroform stent-assisted coiling from 2003 to the present. Our primary comparison was the risk 

of thromboembolic complications related to the stent versus bleeding complications related to being on antiplatelet agents.

Results: Sixteen patients were treated within fifteen days of the hemorrhage. All patients were loaded with Plavix 375 mg 

immediately post procedure, all patients were given Plavix 75 mg daily and most patients were also given aspirin 325 mg 

daily, unless there was concern for bleeding risk, such as ventriculostomy placement. Six patients had ventriculostomies 

placed prior to the procedure and none were placed post procedure. Six patients were given a GPIIbIIIa inhibitor 

(tirofiban) bolus dose during the procedure because of platelet aggregates seen on the stent all without complication, 

including three patients with ventriculostomies. There was one symptomatic ischemic stroke (NIHSS 4) that occurred 22 

days after stent placement possibly related to the stent in a patient with a ventriculostomy who was on a single antiplatelet 

agent (1/16, 6.25%). Minor bleeding complications potentially related to antiplatelet therapy included one patient with two 

small asymptomatic right frontal hemorrhages after vasospasm treatment, one patient with a stable epidural hematoma at 

a craniotomy site not requiring treatment, and one patient with transient hematuria (3/16, 18.75%). Major bleeding 

complications included one patient with a groin pseudoaneurysm that required surgical repair (1/16, 6.25%). In the 

Hunt/Hess grade 1-2 patients, there were no ischemic strokes, one minor bleeding complication (stable epidural 

hematoma) (1/10, 10%) and one major bleeding complication (groin pseudoaneurysm) (1/10, 10%),. In Hunt/Hess grade 

3-4 patients, there was one symptomatic ischemic stroke (1/6, 17%), two minor bleeding complications (small 

asymptomatic frontal hemorrhages and hematuria) (2/6, 33%) and no major bleeding complications. 

Conclusion: Our preliminary data suggests that Neuroform stent-assisted coiling with the use of double antiplatelet 

agents in Hunt/Hess grade 1-2 patients can be safe. Higher Hunt/Hess grade patients with ventriculostomies require a 

careful balance of the risk of hemorrhagic complications from placing patients on double antiplatelet agents versus 

ischemic complications from placing patients on a single antiplatelet agent. Tirofiban can be used at the end of the coiling 

procedure if platelet aggregates are seen on the stent to prevent ischemic stroke complications. 

Keyword: Neuroform stent ; Ruptured intracranial aneurysms ; Tirofiban 

The authors of this work have indicated that they will be discussing/presenting an unapproved or investigative use of 

Merck and Co., Inc.’s Tirofiban (Aggrastat) 


Treatment of wide-neck distal bifurcation aneurysms using a kissing balloon technique in 11 patients 

Chapot, R.·Narata, A. P.·Rogopoulos, A.·Drouineau, J.·Herbreteau, D. 

Hôpital Universitaire Dupuytren 

Limoges, FRANCE. 

Purpose: Wide neck aneurysms may be challenging for endovascular treatment, especially when located on small 

arteries. Remodelling balloons enable embolization of wide neck aneurysms, even in distal aneurysms. This treatment 

may fail if the neck is extending on both division branches. In such situation, we used simultaneously two remodelling 

balloons to enable endovascular treatment. 

Materials and Methods: The aneurysms were treated within the last 10 months, their location was at the MCA bifurcation 

(n=10) and A com (n=1). The aneurysms were unruptured and ranged in size from 5 to 12 mm. Two Hyperform® balloons 

were placed in each division branch and inflated simultaneously during coil insertion. 

Results: Complete aneurysmal occlusion was obtained in all attempted cases. There was no ischemic or hemorrhagic 

complication. Thrombus extension without parent artery occlusion occurred in 2 patients and was managed by 

antiGPIIbIIIa. Angiographic follow-up is available in 8 patients, showing a stable result. 

Conclusion: The kissing balloon technique enables endovascular treatment for distal wide neck aneurysms that are 

usually not considered for embolization. This technique appears to be efficient and safe in our preliminary experience. 

Keyword: Aneurysm ; Wide neck ; remodeling


Referred Pain to the Ipsilateral Forehead and Orbit: An Unusual Phenomenon During Bronchial Artery 

Embolization 

Ketkar, M. 1 ·Ramakantan, R. 2 ·Maddali, K. 3 ·Deshmukh, H. 4 

1 Louisiana State University Health Sciences Center, Shreveport, LA, 2 King Edward Memorial Hospital, Mumbai, INDIA, 

3 Lokmanya Tilak Medical College, Bombay, INDIA, 4 King Edward Memorial Hospital, Bombay, INDIA. 

Purpose: We report an unusual pattern of referred pain to the ipsilateral forehead and orbit observed during bronchial 

artery embolization (BAE) for massive hemoptysis due to pulmonary tuberculosis (TB) and postulate possible neural 

mechanisms for its occurrence. 

Methods: Seven men, from a series of 194 patients (171 men, 23 women) undergoing BAE (right bronchial artery 4, left 

3) with gelatin sponge for control of massive hemoptysis due to pulmonary TB form the subject of this report. 

Results: Embolization was successful in achieving control of hemoptysis in these patients and there were no 

complications following the embolization. Transient, moderately severe, ipsilateral supraorbital and/or retroorbital pain 

occurred only during the injection of the gelatin sponge contrast mixture into the bronchial artery. The pain did not occur 

during the injection of heparinized saline or ionic contrast medium.

Conclusions: Referred pain during BAE is an unusual phenomenon. Acute vessel distension triggering visceral 

sensations is probably the causative mechanism. Sympathetic afferents from the bronchi coursing through the posterior 

pulmonary plexus eventually pass to the trigeminal ganglion via the carotid sympathetic chain. The ophthalmic and 

maxillary divisions of the trigeminal nerve then mediate pain sensation to the ipsilateral forehead and orbit. Similarly, 

parasympathetic afferents from the pulmonary plexus crossing the nucleus of the spinal tract of the trigeminal nerve may 

be responsible for interexchange of impulses to the neurons in this nucleus. Sensory fibers of the ophthalmic and 

maxillary nerves relaying in this nucleus are then involved in this pain being referred to the forehead and orbit. 

Keyword: Tuberculosis ; Arteries, bronchial ; Referred pain


PROGRAM ADDENDUM 

As of 4-27-04 

The program changes listed below were received after the 

ASNR 43rd Annual Meeting Proceedings Book went to press. 

For Invited Speaker and Presenter Disclosure Summary, 

refer to printed document inserted in Proceedings Book. 

Session 12 

• Page 36 – NeuroNews Scientific Paper Session - The Abstracts for this session can be found in the 

NeuroNews Addendum. Soonmee Cha, MD and John Chaloupka, MD are the moderators. 

Session 17b 

• Page 44 - Paper 88, Unusual Imaging Findings following Ethanol Ablation of a Glomus Jugulare Tumor 

An author has been added. The author block now reads as follows: 

Wallace, M. A. Kanal, E,· Rothfus, W. E. · Bartynski, W. S. 

University of Pittsburgh Medical Center, Pittsburgh, PA 

Session 17c 

• Page 50 – Paper 99, Ectopic Parathyroid Adenoma: A Retrospective Analysis of Radiographic Findings 

Vito La Fata, MD has been added as an author and will also present the paper. The author 

block now reads as follows: 


Urena, R. J. Hudgins, P. A. · Weber, C. · Miller, J. ·La Fatta, V. Mansour, K.· 

Esteves, F. 

Emory University School of Medicine, Atlanta, GA 

Session 26b 

• Page 81 – Paper 158, Recanalization Versus Perfusion in Acute Stroke Revascularization Therapy 

Pooja Khatri, MD, co-author, will be presenting the paper. 

Session 33d 

• Page 127 – Paper 242, Change of Apparent Diffusion Coefficient in Patients with Parkinson’s Disease 

The institution for co-author Ji Kang Park, MD has changed. Dr. Park is now at Cheju 

National University Hospital , Jeju-City, Jeju-Do, South Korea. The author block now 

reads as follows: 

Park, J. K. 1 · Jeong, A. 2 · Choi, S. 2 · Kang, B 2 . · Hwang, J. 2 · Lee, J. 2 

1 Cheju National University Hospital , Jeju-City, Jeju-Do, South Korea 2 Ulsan 

University Hospital, Ulsan, REPUBLIC OF KOREA

2005 Proceedings Book Addendum 


Session 51a 

• Page 148 – Paper 283, Quantitative and Qualitative Comparison between 3 and 1.5 T Imaging in the 

Evaluation of Epilepsy 

James Anderson, MD, co-author will be presenting the paper. 

Session 51b 

• Page 155 – Paper 295, Optimization of Echo Time Increases Observer Performance in the Detection of 

Artificial Multiple Sclerosis Lesions on Simulated FLAIR Images of the Brain, has been 

withdrawn. 

Session 51c 

• Page 163 – Paper 307, MR Imaging of the Spinal Cord at 3 T: Evaluation of T2 FLAIR and Diffusion- 

Weighted Imaging 

Thursday, May 26 

Gary Nesbit, MD, co-author, will be presenting the paper. 

Session 64d 

• Page 209 – Laurie Loevner, MD will be the second moderator for this session. 

Session 69 

• Page 221 – Lawrence Tanenbaum, MD will be the second moderator for this session 

Session 72d 

• Page 249 – Paper 455, Tractography in Pediatric Brain Stem Glioma Patients: Wallerian Degeneration 

of Corticospinal and Sensory Pathways 

Co-author James Weeks, MD will be presenting the paper. 


Session 75b 

• Page 262 – Pratik Mukherjee, MD has replaced Caroline D. Robson, MB, ChB as moderator. 

Session 75c 

• Page 273 – Paper 495 - Clinical and MR Characteristics of Dysembryoplastic Neuroepithelial Tumors in 

Children 

New authors have been added. The author block now reads as follows: 

Rodriguez, D. P.·Young Poussaint, T.·Zacharoulis, S.·Turner, C. D.·Chi, S. 

N.·Klement, G.·Ullrich, N.·Pomeroy, S.·Marcus, K.·Goumnerova, L.·Scott, 

M.·Kieran, M. W. 

Children's Hospital, Boston, MA. 

2

2005 Proceedings Book Addendum 

Scientific Posters 

Poster 45 

• Page 309 - Multicentric Brain Gliomas and Gliomatosis Cerebri: MR Imaging Findings and Follow Up 

has been withdrawn. 



• Page 393 - MR Imaging of Multicentric Brain Gliomas and Gliomatosis Cerebri has been withdrawn. 


• Page 424 - Radiologic Spectrum of Intraspinal Cystic Lesions and Their Mimics: A Space-Based 

Approach 

Two additional authors have been added. The author block now reads as follows: 

Electronic Scientific Exhibits (eSEs) 

eSE Changes 

eSE 13 

Revzin, M. V. 1 ·Lev, S. 2 ·Lev, M. H. 3 ·Pomerantz, S. 3 

1 North Shore University Medical Center, Manhasset, NY, 2 Nassau University 

Medical Center, East Meadow, NY, 3 Massachusetts General Hospital, Boston, 

MA. 

• Page 432 - “Hey!! I Have a Lump in My Jaw!”: Nonodontogenic Lesions of the Jaw , is now 

Scientific Exhibit 58A. 

eSE 27 

• Page 438 - “Bright and White” Deep Gray Matter Nuclei, is now Scientific Exhibit 81A. 

3

LEARN! 

Anytime. 

Anywhere. 

ASNR eCME 

http://foundation.asnr.org/onlinecme 

When you’re ready to enhance your knowledge, ASNR eCME is there. 

View dozens of online lectures from neuroradiology’s top educators. Free to ASNR members. 

Sponsored by the Neuroradiology Education and Research Foundation

PLATINUM LEVEL 

An educational grant in support of NER Foundation Symposium 2005: 

Head and Neck Neoplasms: Advanced Imaging Applications: 

What Neuroradiologists Need to Know programming; 

Berlex/NER Foundation Fellowship in Basic Sciences Research Award; 

Berlex Best Paper Award in General Neuroradiology 

GOLD LEVEL 

TM 

An educational grant in support of NER 

Foundation/Boston Scientific Fellowship in 

Cerebrovascular Disease Research Award 

PROGRAM CONTRIBUTORS

2005 Proceedings - ASNR

Create successful ePaper yourself

Delete template?

Save as template?