ELECTRONIC POSTER - ismrm

14:00 4283. Differentiation of Radiation-Injuries and Tumor Recurrence Using Perfusion-
Weighted Imaging
Yu Lin Wang 1 , Lin Ma 2
1 department of radiology,PLA gerneral hospital, China; 2 department of radiology,PLA gerneral hospital,
beijing, China
PWI made it possible to obtain measurements of vascularity within brain lesions. The vascularity of malignant tumor differs
dramatically from that of radiation necrosis. Thus, tumor recurrence within irradiated lesions may be differentiated from regions of
radiation necrosis with PWI. 15 patients were prospectively entered into the study on the basis of the following criteria: previous
treatment with radiation therapy after surgical resection for intraaxial tumors; new development of enhancing lesions within the
radiation field. The final determination of the new development of enhancing lesions was decided either histologically or
clinicoradiologically.
14:30 4284. Non-Negative Matrix Factorization for Differentiation of Brain Metastasis and
Glioblastoma Multiforme, and Visualization of Tumor Infiltration
Jan Luts 1 , Teresa Laudadio, 1,2 , M. Carmen Martinez-Bisbal 3,4 , Sofie Van Cauter 1 ,
Enrique Molla 5 , Jose Piquer 5 , Johan Suykens 1 , Uwe Himmelreich 1 , Bernardo Celda 3,4 ,
Sabine Van Huffel 1
1 Katholieke Universiteit Leuven, Leuven, Belgium; 2 Istituto Applicazioni Calcolo, CNR, Bari, Italy; 3 University
of Valencia, Valencia, Spain; 4 Networking Research Center on Bioengineering, Biomaterials and Nanomedicine
(CIBER-BBN), Valencia, Spain; 5 Hospital de La Ribera, Valencia, Spain
This study focuses on the differentiation between solitary brain metastasis and glioblastoma multiforme based on conventional
magnetic resonance imaging (MRI) and long TE two-dimensional turbo spectroscopic imaging (2D-TSI) data. Fifteen patients with a
brain tumor, nine affected by glioblastoma multiforme and six by metastasis, were considered. Non-negative matrix factorization
(NNMF) results in a clear separation of glioblastomas and metastases. The methods allows visualizing the abundances of the normal
tissue component, which indicate tumor infiltration. In conclusion, automated processing with NNMF of 2D-TSI enables to visualize
metabolic differences between glioblastomas and metastases and enables to visualize tumor infiltration.
15:00 4285. The Effect of Bevacizumab on Normal Appearing White Matter Fibers: A Diffusion
Tensor Imaging (Dti) Study
Moran Artzi 1,2 , Deborah T. Blumenthal 3,4 , Felix Bokstein 3,4 , Benjamin W. Corn, 4,5 ,
Palmon Mika 1 , Orna Aizenstein 6 , Dafna Ben Bashat 1
1 The Wohl Institute for Advanced Imaging, Brain Imaging Center, Tel Aviv Sourasky Medical Center, Tel-
Aviv, Israel; 2 Sackler Faculty of Medicine, Tel-Aviv University , Tel-Aviv, Israel; 3 Neuro-Oncology Service,
Oncology Division, Tel Aviv Sourasky Medical Center, Tel-Aviv, Israel; 4 Sackler Faculty of Medicine, Tel-
Aviv University, Tel-Aviv, Israel; 5 Radiation Oncology Unit, Oncology Division, Tel Aviv Sourasky Medical
Center, Tel-Aviv, Israel; 6 Department of Radiology, Tel Aviv Sourasky Medical Center, Tel-Aviv, Israel
Combined chemo-radiation therapy (RT) is the standard first-line treatment for glioblastoma (GB). Recently, antiangiogenic-drugs
such as bevacizumab have become routine second-line-therapy for patients with recurrent-GB; however the effects of this agent on the
normal appearing white-matter (NAWM) are yet unknown. In this work we scanned patients with primary brain tumors before and
after RT and during bevacizumab therapy. DTI was used to evaluate NAWM changes compared to healthy controls. Changes in
diffusivity values were detected following RT but not during bevacizumab-therapy. Those results support that post-radiation changes
occur, without evidence for additional WM toxicity from bevacizumab-therapy.
Wednesday 13:30-15:30 Computer 74
13:30 4286. Support Vector Machines in DSC-Based Glioma Imaging – Suggestions for Optimal
Characterization
Frank G. Zöllner 1 , Kyrre Eeg Emblem 2 , Lothar R. Schad 1
1 Heidelberg University, Mannheim, Germany; 2 Oslo University Hospital, Oslo, Norway
Dynamic susceptibility contrast magnetic resonance perfusion imaging (DSC-MRI) is a method of choice to characterize gliomas.
Recently, support vector machines (SVM) have been introduced as means to prospectively characterize new patients based on
information from previous patients. Based on features derived from automatically segmented tumor volumes from 101 DSC-MR
examinations, four different SVM models were compared. All SVM models achieved high prediction accuracies (>82%) after
rebalancing the training data sets to equal amounts of samples per class. Best discrimination were obtained using a SVM model with a
radial basis function kernel allowing for a correct prediction of low-grade glioma at 83% and high-grade glioma at 91%.