Dosimetry for IMRT

Dosimetry for IMRT
Thomas Rockwell Mackie
Department of Medical Physics
University y of Wisconsin
Madison WI
trmackie@wisc.edu

Financial Disclosure
I am a founder and Chairman of TomoTherapy
Inc. (Madison, ( , WI) ) which is participating p p g in the
commercial development of helical tomotherapy.

Outline
• Dosimetry Issues Relevant to IMRT
• Charged Particle Equilibrium
• Temporal Non-Constancy
• Partial Volume Effects
• Non-Standard Fields
• IMRT as a Large Collection of Small Non-Standard Fields
• IAEA Calibration Initiative for Non-Standard Non Standard Fields
• Patient-Specific Methods To Verify Delivery of IMRT
• Independent Calculations
• Delivery of Individual Radiation Fields
• Delivery of All Fields into a Phantom
• Metrics for Patient-Specific Patient Specific Dosimetry QA

Dosimetry Issues
Relevant to IMRT
• Charged particle equilibrium
– Different spectrum for collection of small fields
– NNon-uniform if dose d
• Temporal non-constancy
– A very small effect for ion chambers
– May not be true for other dosimeters
• Partial volume effect
– Most important effect especially when measuring
output p
factors for small fields

Non-Uniform Intensity Changes the
Energy gy Spectrum p
Photon
Photon
Beamlet
Beamlet
More High
Energy gy
Electrons
More Low
Energy
Electrons

Change in Stopping Power Ratio
Comparison of Spencer-Attix (∆=10 keV) restricted mass collisional stopping
powers ratios (water/air) at 5 cm depth in water for various 6 MV beams with
stereotactic and MLC beams.
6 MV beams Beam quality,
TPR(20,10)
Andreo,
(1994)
Sánchez-
Doblado,
(2003)
El Elekta kt SL-18 SL 18
10 x 10 cm
0 690 1 1187 1 1188 1 000
2
0.690 1.1187 1.1188 1.000
1.0 cm diameter
stereo field
0.3 cm diameter
stereo field
Siemens Primus
10 x 10 cm 2
2 x 2 cm 2 irregular
on-axis beamlet
2 x 2 cm 2 irregular
8 cm off-axis
1.1155 0.997
1.1153 0.997
0.677 1.1213 1.1221 1.001
1.1203 0.999
1.1250 1.003
Column4/Column3

Signal (fC)
12000
10000
8000
6000
4000
2000
0
-2000
0.16 0
Temporal Delivery of IMRT
46.9 46
93.7 93
Delivery of Dose to a Single Voxel
141 1
188 1
235 2
282 2
329 3
376 3
423 4
IMRT Delivery Signatures
470 4
517 5
Step & Shoot HIART
564 5
613 6
658 6
702 7
747 7
Elapsed Time (sec)
From Tim Holmes, St. Agnes Baltimore
792 7
836 8
881 8
926 9
971 9
10015
10060
11105
11149
11194
12239

Cumulative
Dose
Normalized
1
0.9
0.8
0.7
0.6
0.5
04 0.4
0.3
0.2
01 0.1
0
Temporal Delivery of IMRT
DDelivery li of f DDose tto a Si Single l Voxel V l
5 %
90 %
5 %
Step and Shoot Helical Tomo
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Elapsed Time (minutes)
From Tim Holmes, St. Agnes Baltimore

Outpuut
Factorr
0.9
0.8
00.7
0.6
05 0.5
0.4
03 0.3
Partial Volume Effect
Dose to Water For Small Fields
MMonte t CCarlo l
Diamond
Diode
LAC+film
Film
PTW 31014 (Pinpoint)
PTW 31010 (Semiflex)
PTW 30006 (Farmer)
0.2
0.5 1.0 1.5 2.0 2.5 3.0
Side of Square Field (cm)
From Roberto Capote, IAEA

output o factor
rel to 4x4
1.10
100 1.00
0.90
0.80
0.70
0.60
0.50
0.40
Partial Volume Effect
output factors relative to 4 cm x 4 cm
standard data
photon diode
di diamond d
pinpoint
0.125 cc chamber
electron diode
0 1 2 3 4 5 6 7 8 9 10
equivalent field lenght (cm)
From Karen Rosser, Royal Marsden

High Uncertainty in Output Factors
EExample: l St Statistics ti ti of f 45 OOutput t t Factors F t for f 6 mm and d 18 mm square fi fields ld
(Novalis, SSD = 1000 mm, depth = 50 mm, various detectors)
Factor of Two in
Beam Calibration!
From Wolfgang Ullrich, BrainLab

Reasons for Drop in Output
with Small Field Size
• Backscatter into monitor unit from
beam defining jaws
• Reduced scatter (phantom and head)
• Electronic disequilibrium
• Obscuration of the source

Problems with Conventional Output
Factors
Standard Reference
Conditions
Small Field
Conditions
• Small field
openings p g obscure
the source which is
difficult to measure
and error prone.
• Amount of phantom
scatter changes as
well as lateral
disequilibrium
disequilibrium.
• Partial volume
effects can mask
machine output
factor.

Scan Phantom Across a Narrow Beam
Jaw field width (FW)
• If the source were a point source, source the
measured primary dose would be
directly proportional to the jaw field
width and inversely proportional to the
scan speed.
• Nearly equal phantom scatter conditions
for all jaw settings.
• No partial volume effects once scanning
is complete.
Beam’s eye view

Animation of Scanning Path
From the Point of View of the Phantom
The fan beam
field 10 cm long
by FW wide
FW is i the fan f beam be width idth
bbegins i at one
Th The ffan bbeam
side of the field
field is always
and is scanned
scanned the
across the
same distance
phantom. FW is
(10 cm) at the
defined by the
50% to 50%
same speed. speed
dose.
Beam’s eye view

Scanning at Velocity v
Assume that the source is a point source
Jaw field width (FW)
� �
( r)
Dr ( )
10 cm

Animation of Scanning Path
From the Point of View of the Phantom
The fan beam
field 10 cm long
by FW wide
FW is i the fan f beam be width idth
bbegins i at one
Th The ffan bbeam
side of the field
field is always
and is scanned
scanned the
across the
same distance
phantom. FW is
(10 cm) at the
defined by the
50% to 50%
same speed. speed
dose.
Beam’s eye view

Scanning at Velocity v
Jaw field width (FW)
� ( r)
Dr ( )
10 cm
D ��
FW
v

Dose D Ratio
Dosse
Impact of Source Occlusion for Variable
Jaw Widths
20
18
16
14
12
10
8
6
4
2
0
Dashed line is
expected
result for a
point source
t dWater Tank Dose Ratios
0 1 2 3 4 5
Measured Field Width, FW (cm)
Measured Field Width, FW (cm)
Mainly Due to Source Obscuration
• Only the field
width, FW for
each
measurement
was altered.
• The scan distance
(10 cm) and scan
velocity were
identical.

Dose Ratio
Dose
20
18
16
14
12
10
8
6
4
2
Dose and Dose/FW as a
Function of Field Width (FW)
Divide by Field Width and Normalize
Water Tank Dose Ratios
Normalized
Doose/FW
Dose/FWW
Ratio
1.2
1
0.8
0.6
0.4
0.2
Water Tank Dose/FW Ratios
Plot of dose D(FW) Plot of D/FW
Basically a measure
of the clipping the
photon source so rce seen
for the smaller field
widths.
0
0
0 1 2 3 4 5
0 1 2 3 4 5
Measured Measured Field Field Width, FW FW (cm) (cm) Measured Field Width, FW (cm)
Measured Field Width, FW (cm)
From Brian Hundertmark

No Difference Between Water and
Solid Water
From Brian Hundertmark

Little Difference With Depth
From Brian Hundertmark

Small Vs. Farmer Chambers
A1SL: 0.057 0 057 cc
PTW: 0.6 cc
From Brian Hundertmark

Helical Delivery
From Brian Hundertmark

The Same Technique Could Be Applied for
Other Small Beams
Now for a point source, the measured primary dose
would be proportional to the area of the beam and
inversely proportional to the scan velocity.
A circular beam
is scanned in a
raster pattern to
create a broad
beam using the
same overall
irradiation time.
Beam’s eye view
You would
actually scan the
phantom
instead of
scanning the
beam.

Avoids Problems with
Measuring Output Factors for
Small Fields
•Charged g particle p equilibrium q
– Nearly same charged particle spectrum
• Temporal non-constancy
non constancy
– Deliberately uses temporal non-constancy
– Mimics how composite small fields become
larger irradiated fields
• Partial volume effect
– No partial volume effects

Measurements Would Benchmark Source Model
Model the source
distribution
Calculate a delivery to a
large area using a
superposition of small
beams
Measure the dose in the
center of the field
Repeat for another set of
small ll beams b until til a range of f
beam sizes are done.
no
Enough?
no
yes
Agreement
between calculation and
measurement? yes
Done

IMRT Is Far from Reference Conditions
0.05 0.30 0.55 0.80
0.10 0.35 0.60
0.15 0.40 0.65
0.20 0.45 0.70 0.95
0.25 0.50 0.75 1.00
From Art Boyer, Stanford
0.85
0.90

Audit for TRS 398 Reference Dosimetry
Relative
Deviatio D on (%)
10
8
6
4
2
0
-2
-4
-66
-8
-10
CASE NUMBER
Farmer
Semiflex
PinPoint
1 2 3 4 5 6 7
From Roberto Capote, IAEA

Relative R e deviaation
(% %)
10
8
6
4
2
0
-22
-4
-6
-8
-10
IMRT Audit
Farmer Semiflex PinPoint
Step-and-shoot
Dynamic
1 2 3 4 5 6 7 8 9 10 11 12 13
CASE NUMBER
Sánchez-Doblado F, Hartmann G., Pena J., Capote R. et al
IJROBP 68 (2007) 301-310

IMRT Is About Using Small Fields
Intensity pattern
possibly p y unconstrained
intensity levels
Intensity limited to a few
discrete intensity y levels
What is delivered
Adapted from Michael Sharpe, U. of Toronto

Leaf Penumbra is Important
Central
Axis
Lines Defining
One Half Value
Layer of Beam
Attenuation
Lines Defining
Light Field
Boundary
• It is important to have an accurate model of
the curved leaf ends. ends
• More important for summation of small fields.

Meeting for the Dosimetry Code of Practice:
Small Fields and Novel Beams 3-5/12/07
Outside Participants
• Jan Seuntjens j
• Hugo Palmans
• Karen Rosser
• Saiful Huq
• Wolfgang Ullrich
• Warren Kilby
• Rock Mackie
IAEA
• Pedro Andreo
• Ken R. Shortt
• Stanislav Vatnitskiyy
• Roberto Capote
• Joanna Joa a Izewska e s a
• Ahmed Meghzifene
• Rodolfo Alfonso

• My Pictures\IAEA_2008\IAEA_2008
006.jpg jpg

Examples of Small
and Novel Fields
• GammaKnife (1 (1.8 8 cm diameter collimator is the
largest collimator – intrinsically composite field –
not 100 cm SSD)
• Linac SRS beams (extrapolate to small field
conditions)
• Accuray (6 cm diameter di collimator lli is i the h largest l
collimator)
• TomoTherapy (5 cm is the largest slice width –
no field flattening filter)
• IMRT (made up of numerous small fields)

Static Field Calibration
Uses a machine-specific reference field, fmsr
Q Co�60
w D,w Q Q
D msr � M �N �k �k
k
(D/ M)
Q
�
w
Qmsr (D/ M)
Q
w
msr
D
Q
w
msr
msr
fmsr
Corrects for the differences between the conditions of field size,
geometry, g y, and beam quality q y of the conventional reference field
and the machine-specific reference field.

Calculate Using MC
Using method of Sempau et al 2004 PMB PMB 49;4427-44
Dw
D
D Dw
a
k
�
( D / D )
w a
Q
msr
Qmsr ( D / D )
Q
w a
is the dose at the reference point in water
is the average dose in the ion chamber
Adapted from Edmond Sterpin
Da

Composite p Field Calibration
Uses a plan-class specific reference field, fpcsr
Q Co�60
w D,w Q Q
D
pcsr
� M �N �k �k
k
pcsr
�
(D/ M)
Q
pcsr
w
Q Q
(D/ M) w
pcsr
fpcsr
Corrects for the differences between the conditions of field size,
geometry, t and d bbeam quality lit of f the th conventional ti l reference f field fi ld
and the plan-class specific reference field.

Dose Calibration Correction Factor to
GGet t DDose iin a Cli Clinical i l Field Fi ld
M
D
Qclin � D
Qmsr�� Qclin � D
Qmsr�
Qclin
�k
k
w w Q w Q
�
(D/ M)
Q
w
Q Q
(D/ M) w
clin
clin msr
msr M
clin
Q Qmsr
In principle, the dose calibration correction factor would be
different for each clinical measurement. Eventually, Monte Carlo
treatment planning systems could compute kQ
for both the
clin
patient and a QA phantom.

Overview of Static Field Dosimetry
REFERENCE DOSIMETRY
D � M � N �k �k
Qmsr Co�
60 Qclin QmsrQclin D � D ��
w D ,w Q Q
w w Q
msr
msr
10 cm x 10 cm
Reference Field
N � k
Co�
60
D,w Q
k Qmsr
Hypothetical 10 cm x 10 cm
Reference Field
Ion Chamber =
k
Q Qmsr
Linac Stereotactic Field
BrainLab MiniMLC
(10 cm x 10 cm)
CyberKnife (6 cm Diameter)
TomoTherapy
(5 cm x 10 cm )
� clin Q
Q msr

“Effective kQ” for Tomo
1. Determine the %dd(10)x[HT ( ) [ Ref] ] for a tomotherapy pyunit
for a 5 cm x 10
cm field at 85 cm SSD.
2. Using the graph at the right look up the value of %dd(10)x[HT TG-51].
3. Using the graph on the left and the derived value %dd(10)x[HT TG-51]
determine the abscissa and this effective kQ value is then called k � k Q Qmsr
in
the IAEA protocol.
Thomas et al (2005)

Static and Composite Field
CCalculations l l ti for f Tomo T
Static Field Calibration
Composite Field Calibration
(Section III.A.1)
(Section III.A.2)
Jeraj et al 2005 Duane et al 2006
k
Q
msr
5 cm x 10 cm 0.997 Unmodulated Helical Delivery
5 cm Slice Width
2 cm x 10 cm 0.993 Unmodulated Helical Delivery
2.5 cm Slice Width
k
Q
pcsr
1.000
1.000
2 cm x 2 cm 0.990 Unmodulated Helical Delivery 0.997
1 cm Slice Width

ICRU Committee on IMRT
Sponsors
• André Wambersie MD, PhD, Brussels, Belgium
• Paul DeLuca PhD, Madison, USA
• RReinhard i h d Gahbaue G hb MD, MD Ph.D, Ph D Cleveland, Cl l d USA
• Gordon Whitmore Ph.D, Toronto, Canada
Chairmen
• Vi Vincent t Grégoire G é i MD PhD, PhD Brussels, B l Belgium B l i
• T. Rock Mackie PhD, Madison, USA
Members
• Wilfried De Neve MD PhD PhD, Gent Gent, Belgium
• Mary Gospodarowicz MD, Toronto, Canada
• Andrzej Niemierko PhD, Boston, USA
• James A. Purdy y PhD, Davis, USA
• Marcel van Herk PhD, Amsterdam, The Netherla
• Nilendu Gupta PhD, Colombus, USA
Consultants
• Anders Ahnesjö PhD, Uppsala, Sweden
• Michael Goiten PhD, Windisch, Switzerland

Supplement to ICRU-50 and ICRU-62
• More emphasis on statistics.
• Prescription and reporting with dose-volume
specifications.
p
• No longer use ICRU-Reference Point.
• Want median dose D50% reported.
• Use model-based dose calculations.
• Include the effect of tissue heterogeneities.
R t d t ll f t t d
• Report dose to small mass of water, not dose
to tissue.

QA for IMRT
• Appropriate QQA of f TPS S and delivery
equipment
• Patient-specific QA:
• Delivery of individual fields into a
dosimeter
• DDelivery li of f all ll of f the th fields fi ld into i t a phantom h t
• Independent dose calculation algorithms
with similar of better dose calculation
accuracy
• In-vivo dosimetry not limited to a single
point.

% Dose D errror
20.0
15.0
10.0
5.0
0.0
Gap Error
GGap error Dose D error
0.2
Range of gap width
0.5
1.0
Gap error (mm)
2.0
0 1 2 3 4 5
Nominal gap g p (cm) ( )
From Tom Losasso, Memorial Sloan Kettering

1 mm Bands
Leaf Positioning Errors
Errors Introduced
� -0.5 mm
� -0.2 mm
� + 0.2 mm
� + 0.5 mm

Mechanical Alignment g of
Static MLC Approach
No offset 0.6 mm offset 1.0 mm offset
From Jatinder Palta, University of Florida

Radiation Field Edge Offset
0.7mm offset 0.6mm offset
0.5mm offset
0.3mm offset

Segmental Multileaf Collimator
MLC*
(SMLC) Delivery D li System S t
TTolerance l AAction ti
Limit Level
Leaf position accuracy
1 mm 2 mm
Leaf position reproducibility 0.2 mm 0.5 mm
Gap width reproducibility 02mm 0.2 mm 05mm 0.5 mm
* Measured at all four cardinal gantry angles
Gantry, MLC, and Table
0.75 mm 1.00 mm
Isocenter radius di radius di
Beam Stability
Low MU Output (

Dynamic Multileaf Collimator
MLC
(DMLC) Delivery System
Tolerance Action
Li Limit it LLevell Leaf position accuracy
05 0.5 mm 1 mm
Leaf position reproducibility 0.2 mm 0.5 mm
Gap width reproducibility
0.2 mm 0.5 mm
Leaf speed �0.1 mm/s �0.2 mm/s
Gantry, MLC, and Table
Isocenter
0.75 mm 1.00 mm
Isocenter radius radius
Beam Stability
Low MU Output p ( (

Fail Fail Fail
Pass
Leaf opening time
Leaf #
Leaf open time histograms (15 sec gantry)
Leaf Usage
Higher Avg. Leaf
Opening Time

TomoTherapy py Leaf Latencyy
• TomoTherapy py uses
linear fit of
measured data to
model leaf latenc latency
• Based on small
lead opening times
we expect leaf
latency as the
cause

patient 1
patient 2
patient 3
PTV 70
PTV 64
PTV 54
PTV 50
larynx
brainstem
r. parotid
PTV 54
esophagus p g
lungs
cord
l. ventricle
PTV 60
PTV 50
esophagus
larynx
cord
r. parotid
bbrainstem i t
original plan
replanned
pitch = 0.143 pitch = 0.287
pitch = 0.215
pitch = 0.215
pitch = 0.287
pitch = 0.287

End-to-End QA

“Cheese”
Phantom
used dffor QQA
measurements
Film Plane
Phantom can be rotated
or turned to acquire any
orthogonal plane
QA Measurements
Measure
plane and
point dose
at the same
time

Dose (GGy)
On and Off-Axis Results
Film and Ion Chamber Absolute Dose
Delivered Dose: 22.5cm 5cm Treatment Beam
2.5
On-Axis Tumor Off-Axis Tumor
2.0
1.5
1.0
0.5
0.0
-20 -15 -10 -5 0 5 10 15 20
Distance (cm)

QA for All
of the
Fields
Tomotherapy
py
Example

Delivery QA Panel

Delivery QA Panel

NNote t th the
High
Gradients
Comparison of Phantom Plan and Verification Film
Film
Plan
Plan
Film
From Chet Ramsey, Thompson Cancer Survival Center

QA Q of Individual Fields
External diode/ion-chamber arrays
• MapCheck
• PTW Octavius phantom
• IBA Matrix
Integrated detector systems
• EPID portal p dosimetry y

Independent Calculation

Dose
Computed
Dose Accuracy and Distance to
Agreement
Dose accuracy for low gradients
� �
� ( , ) � ( ) � ( ) /
r r Dr Dr D m c m c 50%
Measured
rr ( , r)�r � r
m c m c
Low gradients < 20%/cm
Distance to agreement
for high gradients
Radius

Gamma Function
� ( r , r ) �{[ �( r , r )/ �D ] �[
r( r , r )/ �d
] }
2 2 1/ 2
m c m c M m c M
Dose accuracy axis
� ( r , r )
m c
�D
rr ( , r)
m c
M
��
d
M
Pass
Distance to
agreement
axis

Gamma Function
� ( r , r ) �{[ �( r , r )/ �D ] �[
r( r , r )/ �d
] }
2 2 1/ 2
m c m c M m c M
Dose accuracy axis
� ( r , r )
m c
�D
M
rr ( , r)
m c
��
d
M
Fail
Distance to
agreement
axis

Molineu et al IJROBP 2005
Ibott et al Tech in Ca RT 2006
Followill et al Med Phys 2007
IMRT Evaluation using
Anthropomorphic Phantoms
For H&N, using a criteria of 5% or 4mm, the
passing i rate t drops d f from 75% t to 58%
Courtesy Ibott, RPC

� Point 1: Isocenter
� PPoint i t 22: SSpinal i l cord d iisocenter t
� Point 3: Spinal cord cranial
� Point 4: PTV T R
� Point 5: PTV T R cranial
� Point 6: PTV N L
� Point 7: PTV N L caudal
Gortec IMRT Test Phantom
TLDs are placed at seven locations.
Courtesy M. Tomsej,
Brussels

Dm/Dc
1.2
1.15
1.1
1.05
1
0.95
0.9
0.85
Audit Results
D m/D c=f(CENTER) per meas. pt
Sample Result
0.8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
CENTER
isocenter
spinal cord iso
spinal i lcord dcranial i l
PTV T D
PTV T D cranial
PTV N G
PTV N G caudal

Sample Tomotherapy Results

Frequency
Inter-Institution Dose Accuracy
600
500
400
300
200
100
Accuracy Distribution
0
0.885 0.905 0.925 0.945 0.965 0.985 1.005 1.025 1.045 1.065 1.085 1.105 1.125
Measured Dose/Computed Dose
Number of Measurements = 2679
Mean = 0.995
Standard Deviation = 00.025 025
(Updated from Zefkili et al 2004)

Intra-Institution Dose Accuracy
Frequency
350
300
250
200
150
100
50
0
Accuracy Distribution
-10 -8 -6 -4 -2 0 2 4 6 8 10
Relative Difference Between Measured and Calculated Dose (%)
Number of Measurements = 1591
Mean = 0.45%
Standard Deviation = 2.5%
(Updated from Dong et al 2003)

QA Accuracy y for IMRT
• Previous ICRU 5% point-dose accuracy
specification ifi ti replaced l d bby a volumetric l t i ddose
accuracy specification.
• Proposed new ICRU volumetric dose
accuracy specification:
- High gradient (≥ 20%/cm): 85% of points
within 5 mm (3.5 mm SD),
- Low gradient (< 20%/cm): 85% of points
within 5% of predicted dose normalized to
the prescribed dose (3.5% SD).

Monitor Units Calculations for
MModel-Based d l B d DDose CCalculation l l ti
D � 0 D
( Ar , ) � ( Ar , ) ( Ar , )(1 �bA
( ))
MU MU �
d A
P
Beam
Output
0
Computed
Directly
�1
Correction to
Account for
Backscatter
Into Monitor
Chamber

Monitor Units Calculations for
MModel-Based d l B d DDose CCalculation l l ti
�� D ��
( Acal , dcal
)
� �
0 M �
Measured
�
� �
(1 �bA
( cal )) dcal
MUU �� D ��
� ( Acal , dcal
)
�
�
� 0 �Calculated Ref
Not including the effect of backscatter into the
monitor it chamber h b will ill result lt in i about b t a 2% error at t
worst.
A

Backscatter into Monitor Chamber
The effect is due to backscattered photons
entering the monitor and resulting in feedback
to the linac to lower its output
Liu et al.,
Med. Phys 2000;27:737-744
Varian 2100 – 10 MV. Results
with other jaw completely open

Monitor Backscatter for Square
On On-Axis Axis Fields
Varian 2100 – 10 MV
Liu et al., Med. Phys 2000;27:737-744

Conclusion
• IMRT uses complex field boundaries and many small
fields to achieve more conformal dose distributions.
• Partial volume effects can result in severe error
especially when measuring the dose in high gradients.
• A new method for determining the effect of field
obscuring g is pproposed p to eliminate ppartial
volume
effects.
• IMRT deliveries are far from the measurement
conditions of calibration.
• IAEA has developed a formalism to account for small
and novel beams.
• Independent calculations or measurements for patient patientspecific
QA are required for IMRT.
• IAEA will recommend that 85% of measurement values
(dose or distance to agreement) are within 5%.
• Model-based monitor unit calculations are simple but
should include scatter into monitor chamber.