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
Patient Specific Dosimetry QA
• Metrics for Patient-Specific
Dosimetry Issues
Relevant to IMRT
• Charged particle equilibrium
– Different spectrum for collection of small fields
– Non-uniform
N
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
p factors for small fields
output
Non-Uniform Intensity Changes the
gy Spectrum
p
Energy
More High
gy
Energy
Electrons
Photon
Beamlet
Photon
Beamlet
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)
SánchezDoblado,
(2003)
Column4/Column3
0 690
0.690
1 1187
1.1187
1 1188
1.1188
1 000
1.000
1.0 cm diameter
stereo field
1.1155
0.997
0.3 cm diameter
stereo field
1.1153
0.997
1.1221
1.001
2 x 2 cm2 irregular
on-axis beamlet
1.1203
0.999
2 x 2 cm2 irregular
8 cm off-axis
1.1250
1.003
Elekta
El
kt SL-18
SL 18
10 x 10 cm2
Siemens Primus
10 x 10 cm2
0.677
1.1213
Temporal Delivery of IMRT
Delivery of Dose to a Single Voxel
IMRT Delivery Signatures
Step & Shoot
HIART
12000
10000
6000
4000
-2000
Elapsed Time (sec)
From Tim Holmes, St. Agnes Baltimore
1239
1194
1149
1105
1060
1015
971
926
881
836
792
747
702
658
613
564
517
470
423
376
329
282
235
188
141
93.7
0
46.9
2000
0.16
Signal (fC)
8000
Temporal Delivery of IMRT
D li
Delivery
off D
Dose tto a Single
Si l Voxel
V
l
Helical Tomo
5%
90 %
1
0.9
0.8
0.7
0.6
0.5
04
0.4
0.3
0.2
01
0.1
0
5%
Normalized Cumulative Dose
Step and Shoot
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
Partial Volume Effect
Dose to Water For Small Fields
0.9
Outpu
ut Factorr
0.8
Monte
M
t C
Carlo
l
Diamond
Diode
LAC+film
Film
0
0.7
0.6
05
0.5
PTW 31014 (Pinpoint)
PTW 31010 (Semiflex)
PTW 30006 (Farmer)
0.4
03
0.3
0.2
0.5
1.0
1.5
2.0
2.5
Side of Square Field (cm)
From Roberto Capote, IAEA
3.0
Partial Volume Effect
output factors relative to 4 cm x 4 cm
output factor rel to 4x4
o
1.10
1 00
1.00
0.90
0.80
0.70
standard data
photon diode
di
diamond
d
pinpoint
0.125 cc chamber
electron diode
0.60
0.50
0.40
0
1
2
3
4
5
6
7
equivalent field lenght (cm)
8
From Karen Rosser, Royal Marsden
9
10
High Uncertainty in Output Factors
E
Example:
l St
Statistics
ti ti off 45 O
Output
t t Factors
F t
for
f 6 mm and
d 18 mm square fields
fi 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
• 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.
Standard Reference
Conditions
Small Field
Conditions
Scan Phantom Across a Narrow Beam
• 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.
Jaw field width (FW)
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
b i at one
begins
side of the field
and is scanned
across the
phantom. FW is
defined by the
50% to 50%
dose.
FW is
i the fan
f
beam
be
width
idth
The ffan b
Th
beam
field is always
scanned the
same distance
(10 cm) at the
same speed.
speed
Beam’s eye view
Scanning at Velocity v
Assume that the source is a point source
Jaw field width (FW)

 (r)
D( r )
10 cm
Animation of Scanning Path
From the Point of View of the Phantom
The fan beam
field 10 cm long
by FW wide
b i at one
begins
side of the field
and is scanned
across the
phantom. FW is
defined by the
50% to 50%
dose.
FW is
i the fan
f
beam
be
width
idth
The ffan b
Th
beam
field is always
scanned the
same distance
(10 cm) at the
same speed.
speed
Beam’s eye view
Scanning at Velocity v
Jaw field width (FW)
 (r )
FW
D
v
D( r )
10 cm
Impact of Source Occlusion for Variable
Jaw Widths
Dashed line is
expected
t dWater Tank Dose Ratios result for a
point source
20
18
16
Dos
se
Dose Ratio
D
14
•
12
10
8
6
4
•
2
0
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 and Dose/FW as a
Function of Field Width (FW)
Divide by Field Width and Normalize
Water Tank Dose Ratios Water Tank Dose/FW Ratios 20
1.2
18
1
Normalized
Do
ose/FW
Dose/FW
W Ratio
16
Dose
Dose Ratio
14
12
10
8
6
4
2
Basically a measure
of the clipping the
photon source
so rce seen
for the smaller field
widths.
0.8
0.6
0.4
0.2
0
0
0
1
2
3
4
5
0
Measured Field Width, FW (cm)
Measured
Field Width, FW (cm)
Plot of dose D(FW)
From Brian Hundertmark
1
2
3
Measured Field Width, FW (cm)
4
Measured Field Width, FW (cm)
Plot of D/FW
5
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 off
beam sizes are done.
no
Enough?
yes
no
Agreement
between calculation and
measurement?
Done
yes
IMRT Is Far from Reference Conditions
0.05
0.30
0.55
0.80
0.10
0.35
0.60
0.85
0.15
0.40
0.65
0.90
0.20
0.45
0.70
0.95
0.25
0.50
0.75
1.00
From Art Boyer, Stanford
Relative D
Deviatio
on (%)
Audit for TRS 398 Reference Dosimetry
10
8
6
4
2
0
-2
-4
-6
6
-8
-10
Farmer
Semiflex
PinPoint
1
2
3
4
5
CASE NUMBER
From Roberto Capote, IAEA
6
7
IMRT Audit
R
Relative
e devia
ation (%
%)
10
8
6
4
2
0
-2
2
-4
-6
-8
-10
Farmer
Semiflex
PinPoint
Step-and-shoot
1
2
3
4
5
Dynamic
6
7
8
9
CASE NUMBER
Sánchez-Doblado F, Hartmann G., Pena J., Capote R. et al
IJROBP 68 (2007) 301-310
10 11 12 13
IMRT Is About Using Small Fields
Intensity pattern
possibly
p
y unconstrained
intensity levels
What is delivered
Intensity limited to a few
discrete intensityy levels
Adapted from Michael Sharpe, U. of Toronto
Leaf Penumbra is Important
Lines Defining
One Half Value
Layer of Beam
Attenuation
Lines Defining
Light Field
Boundary
Central
Axis
•
•
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
Qmsr
w
D
kQmsr
 M N
Co  60
D ,w
Qmsr
w
Q
w
(D / M )

(D / M )
 kQ  kQmsr
Qmsr
w
D
fmsr
Corrects for the differences between the conditions of field size,
ggeometry,
y, and beam q
quality
y of the conventional reference field
and the machine-specific reference field.
Calculate Using MC
Using method of Sempau et al 2004 PMB 49;4427-44
kQmsr 
( Dw / Da )
Qmsr
( Dw / Da )
Q
Dw is the dose at the reference point in water
Da is the average dose in the ion chamber
Dw
Adapted from Edmond Sterpin
Da
Composite
p
Field Calibration
Uses a plan-class specific reference field, fpcsr
Q pcsr
Dw
 M N
Co  60
D ,w
Q pcsr
kQ pcsr
( D / M )w

Q
( D / M )w
 kQ  kQ pcsr
fpcsr
Corrects for the differences between the conditions of field size,
geometry,
t and
db
beam quality
lit off the
th conventional
ti
l reference
f
field
fi ld
and the plan-class specific reference field.
Dose Calibration Correction Factor to
G tD
Get
Dose iin a Cli
Clinical
i l Field
Fi ld
Qclin
w
D
kQclin
D
Qmsr
w

Qclin
Qmsr
D
Qmsr
w

M Qclin
M Qmsr
 kQclin
Qclin
w
Qmsr
w
(D / M )

(D / M )
In principle, the dose calibration correction factor would be
different for each clinical measurement. Eventually, Monte Carlo
treatment planning systems could compute kQclin for both the
patient and a QA phantom.
Overview of Static Field Dosimetry
REFERENCE DOSIMETRY
D wQ m sr  M  N DC o,w 60  k Q  k Q m sr
10 cm x 10 cm
Reference Field
D wQ clin  D wQ m sr   QQmclinsr
Linac Stereotactic Field
k Q m sr
N DC o,w 60  k Q
BrainLab MiniMLC
(10 cm x 10 cm)
Hypothetical 10 cm x 10 cm
Reference Field
 QQ
m sr
CyberKnife (6 cm Diameter)
k Q m sr
Ion Chamber =
clin
TomoTherapy
(5 cm x 10 cm )
“Effective kQ” for Tomo
1. Determine the %dd(10)x[HT
( ) [
Ref]] for a tomotherapy
py unit 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 in
the IAEA protocol.
Q
Thomas et al (2005)
Qmsr
Static and Composite Field
C l l ti
Calculations
for
f Tomo
T
Static Field Calibration
(Section III.A.1)
Jeraj et al 2005
Composite Field Calibration
(Section III.A.2)
Duane et al 2006
kQ pcsr
kQmsr
5 cm x 10 cm
0.997
Unmodulated Helical Delivery
5 cm Slice Width
1.000
2 cm x 10 cm
0.993
Unmodulated Helical Delivery
2.5 cm Slice Width
1.000
2 cm x 2 cm
0.990
Unmodulated Helical Delivery
1 cm Slice Width
0.997
ICRU Committee on IMRT
Sponsors
• André Wambersie MD, PhD, Brussels, Belgium
• Paul DeLuca PhD, Madison, USA
• Reinhard
R i h d Gahbaue
G hb
MD,
MD Ph.D,
Ph D Cleveland,
Cl
l d USA
• Gordon Whitmore Ph.D, Toronto, Canada
Chairmen
• Vincent
Vi
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.
• Report
R
td
dose to
t small
ll mass off water,
t nott dose
d
to tissue.
QA for IMRT
• Appropriate QA
Q off TPS
S and delivery
equipment
• Patient-specific QA:
• Delivery of individual fields into a
dosimeter
• Delivery
D li
off all
ll off 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.
Gap Error
G error
Gap
Dose
D
error
20.0
%D
Dose errror
Range of gap width
15.0
Gap error (mm)
10.0
2.0
1.0
5.0
0.2
0.5
0.0
0
1
2
3
4
5
Nominal g
gap
p (cm)
( )
From Tom Losasso, Memorial Sloan Kettering
Leaf Positioning Errors
1 mm Bands
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
(SMLC) Delivery
D li
System
S t
Tolerance
T
l
Limit
Action
A
ti
Level
1 mm
0.2 mm
0 2 mm
0.2
2 mm
0.5 mm
0 5 mm
0.5
0.75 mm
radius
di
1.00 mm
radius
di
2%
2%
3%
3%
MLC*
Leaf position accuracy
Leaf position reproducibility
Gap width reproducibility
* Measured at all four cardinal gantry angles
Gantry, MLC, and Table
Isocenter
Beam Stability
Low MU Output (<2MU)
Low MU Symmetry (<2MU)
Jatinder Palta, AAPM Summer School 2003
Dynamic Multileaf Collimator
(DMLC) Delivery System
Tolerance
Li it
Limit
Action
L
Level
l
0.5
0
5 mm
0.2 mm
0.2 mm
0.1 mm/s
1 mm
0.5 mm
0.5 mm
0.2 mm/s
MLC
Leaf position accuracy
Leaf position reproducibility
Gap width reproducibility
Leaf speed
Gantry, MLC, and Table
Isocenter
Beam Stability
Low MU Output
p ((<2MU))
Low MU Symmetry (<2MU)
0.75 mm
radius
1.00 mm
radius
3%
2%
Jatinder Palta, AAPM Summer School 2003
5%
3%
Fail
Fail
Fail
Leaf open time histograms (15 sec gantry)
Leaf opening time
Leaf #
Leaf Usage
Pass
Higher Avg. Leaf
Opening Time
TomoTherapy
py Leaf Latency
y
• TomoTherapy
py uses
linear fit of
measured data to
model leaf latency
latenc
• Based on small
lead opening times
we expect leaf
latency as the
cause
patient 1
original plan
PTV 70
PTV 64
PTV 54
PTV 50
larynx
brainstem
r. parotid
replanned
pitch = 0.143
pitch = 0.287
pitch = 0.215
pitch = 0.287
pitch = 0.215
pitch = 0.287
patient 2
PTV 54
esophagus
p g
lungs
cord
l. ventricle
patient 3
PTV 60
PTV 50
esophagus
larynx
cord
r. parotid
b i t
brainstem
End-to-End QA
QA Measurements
“Cheese”
Phantom
used
d ffor Q
QA
measurements
Film Plane
Phantom can be rotated
or turned to acquire any
orthogonal plane
Measure
plane and
point dose
at the same
time
On and Off-Axis Results
Film and Ion Chamber Absolute Dose
Delivered Dose: 2
2.5cm
5cm Treatment Beam
2.5
On-Axis Tumor
Off-Axis Tumor
2.0
Dose (G
Gy)
1.5
1.0
0.5
0.0
-20
-15
-10
-5
0
Distance (cm)
5
10
15
20
QA for All
of the
Fields
Tomotherapy
py
Example
Delivery QA Panel
Delivery QA Panel
Comparison of Phantom Plan and Verification Film
Plan
Note
N
t th
the
High
Gradients
Film
Film
Plan
From Chet Ramsey, Thompson Cancer Survival Center
QA of Individual Fields
Q
External diode/ion-chamber arrays
• MapCheck
• PTW Octavius phantom
• IBA Matrix
Integrated detector systems
• EPID p
portal dosimetry
y
Independent Calculation
Dose Accuracy and Distance to
Agreement
Dose
Dose accuracy for low gradients
 ( rm , rc )   D( rm )  D( rc ) / D50%
Measured
Low gradients < 20%/cm
r ( rm , rc )  rm  rc
Computed
Distance to agreement
for high gradients
Radius
Gamma Function
 ( rm , rc )  {[ ( rm , rc ) /  DM ]2  [ r ( rm , r c ) / d M ]2 }1/ 2
Dose accuracy axis
 DM
Pass
 ( rm , rc )
Distance to
agreement
axis
r ( rm , rc )
d M
Gamma Function
 ( rm , rc )  {[ ( rm , rc ) /  DM ]2  [ r ( rm , r c ) / d M ]2 }1/ 2
Dose accuracy axis
 DM
Fail
 ( rm , rc )
Distance to
agreement
axis
r ( rm , rc )
d M
IMRT Evaluation using
Anthropomorphic Phantoms
Molineu et al IJROBP 2005
Ibott et al Tech in Ca RT 2006
Followill et al Med Phys 2007
For H&N, using a criteria of 5% or 4mm, the
passing
i rate
t drops
d
f
from 75% tto 58%
Courtesy Ibott, RPC
Gortec IMRT Test Phantom
TLDs are placed at seven locations.

Point 1: Isocenter

P i t2
Point
2: S
Spinal
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
Courtesy M. Tomsej,
Brussels
Audit Results
Dm/Dc=f(CENTER) per meas. pt
1.2
1.15
1.1
isocenter
Dm/Dc
1.05
spinal cord iso
spinal
i l cord
d cranial
i l
1
PTV T D
PTV T D cranial
PTV N G
0.95
PTV N G caudal
0.9
Sample Result
0.85
0.8
0
1
2
3
4
5
6
7
8
9
10
11
CENTER
12
13
14
15
16
17
18
19
20
Sample Tomotherapy Results
Inter-Institution Dose Accuracy
Accuracy Distribution
600
Frequency
500
400
300
200
100
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 = 0
0.025
025
(Updated from Zefkili et al 2004)
Intra-Institution Dose Accuracy
Accuracy Distribution
350
Frequency
300
250
200
150
100
50
0
-10
-8
-6
-4
-2
0
2
4
6
8
Relative Difference Between Measured and Calculated Dose (% )
Number of Measurements = 1591
Mean = 0.45%
Standard Deviation = 2.5%
(Updated from Dong et al 2003)
10
QA Accuracy
y for IMRT
• Previous ICRU 5% point-dose accuracy
specification
ifi ti replaced
l
db
by a volumetric
l
ti d
dose
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
M d lB
Model-Based
dD
Dose C
Calculation
l l ti
0
D
D
1
( A, r ) 
( A, r )
( A, r ) (1  b( A))
MU
MU
0
Beam
Output
d
A
P
Computed
Directly
Correction to
Account for
Backscatter
Into Monitor
Chamber
Monitor Units Calculations for
M d lB
Model-Based
dD
Dose C
Calculation
l l ti
D

( Acal , dcal ) 

0
M

 Measured (1  b( A ))

cal
MU
U D

 ( Acal , dcal ) 
 0
 Calculated
dcal
A
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 att
worst.
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
Varian 2100 – 10 MV. Results
with other jaw completely open
Liu et al.,
Med. Phys 2000;27:737-744
Monitor Backscatter for Square
On Axis Fields
On-Axis
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 p
proposed
p
to eliminate p
partial 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.