Technical Note: Preferred dosimeter size and associated correction factors in

commissioning high dose per pulse, flattening filter free x-ray beams
A. Sudhyadhom, N. Kirby, B. Faddegon, and C. F. Chuang

Citation: Medical Physics 43, 1507 (2016); doi: 10.1118/1.4941691

View online: http://dx.doi.org/10.1118/1.4941691
View Table of Contents: http://scitation.aip.org/content/aapm/journal/medphys/43/3?ver=pdfcov
Published by the American Association of Physicists in Medicine

Articles you may be interested in

Technical Note: A direct ray-tracing method to compute integral depth dose in pencil beam proton
radiography with a multilayer ionization chamber
Med. Phys. 43, 6405 (2016); 10.1118/1.4966703

Technical Note: Correction for intra-chamber dose gradients in reference dosimetry of flattening-filter-free MV
photon beams
Med. Phys. 43, 4729 (2016); 10.1118/1.4958960

Conception and realization of a parallel-plate free-air ionization chamber for the absolute dosimetry of an
ultrasoft X-ray beam
Rev. Sci. Instrum. 85, 083304 (2014); 10.1063/1.4890817

The Italian affair: The employment of parallel-plate ionization chambers for dose measurements in high dose-
per-pulse IORT electron beams
Med. Phys. 37, 2918 (2010); 10.1118/1.3432601

Ion recombination correction for very high dose-per-pulse high-energy electron beams
Med. Phys. 32, 2204 (2005); 10.1118/1.1940167
Technical Note: Preferred dosimeter size and associated
correction factors in commissioning high dose per pulse,
flattening filter free x-ray beams
A. Sudhyadhoma)
Department of Radiation Oncology, University of California, San Francisco, San Francisco, California 94115

N. Kirby
Department of Radiation Oncology, University of California, San Francisco, San Francisco, California 94115
and Department of Radiation Oncology and Radiology, UTHSCSA, San Antonio, San Antonio, Texas 78229

B. Faddegon and C. F. Chuang

Department of Radiation Oncology, University of California, San Francisco, San Francisco, California 94115

(Received 1 December 2014; revised 5 January 2016; accepted for publication 19 January 2016;
published 26 February 2016)
Purpose: High dose rate flattening filter free (FFF) beams pose new challenges and considerations
for accurate reference and relative dosimetry. The authors report errors associated with commonly
used ion chambers and introduce simple methods to mitigate them.
Methods: Dosimetric errors due to (1) ion recombination effects of high dose per pulse (DPP)
FFF beams and (2) volume-averaging effects of the radial profile were examined on a TrueBeam
STx. Four commonly used cylindrical ion chambers spanning a range of lengths (0.29–2.3 cm)
and volumes (0.016–0.6 cm3) were used to determine the magnitude of these effects for 6 and
10 MV unflattened x-ray beams (6XFFF and 10XFFF, respectively). Two methods were used to
determine the magnitude of ion collection efficiency: (1) direct measurement of the percent depth
dose (PDD) for the clinical, high DPP beam in comparison to that obtained after reducing the
DPP and (2) measurement of Pion as a function of depth. Two methods were used to quantify the
magnitude of volume-averaging: (1) direct measurement of volume-averaging via cross-calibration
and (2) calculation of volume-averaging from radial profiles of the beam. Finally, a simple analytical
expression for the radial profile volume-averaging correction factor, Prp = [OAR(0.29L)]−1, or the
inverse of the off-axis ratio of dose at 0.29L, where L is the length of the chamber’s sensitive volume,
is introduced to mitigate the volume-averaging effect in Farmer-type chambers.
Results: Errors in measured PDD for the clinical beams were 1.3% ± 0.07% and 1.6% ± 0.07% at
35 cm depth for the 6XFFF and 10XFFF beam, respectively, using an IBA CC13 ion chamber, due to
charge recombination with a high DPP. Volume-averaging effects were 0.4% and 0.7% for the 6XFFF
and 10XFFF beam, respectively, when measured with a Farmer-type chamber. For the application of
TG-51, these errors combine when using a CC13 to measure the PDD and a Farmer for absolute
output dosimetry for a total error of up to 2% at d max for the 10XFFF beam.
Conclusions: Relative and absolute dosimetry in high DPP, unflattened x-ray beams of 10 MV
or higher requires corrections for charge recombination and/or volume-averaging when dosimeters
with certain geometries are used. Chambers used for PDD measurement are available that do not
require a correction for charge recombination. A simple analytical expression of the correction
factor Prp was introduced in this work to account for volume-averaging effects in Farmer chambers.
Choice of an appropriate dosimeter coupled with application of the established correction factors
Pion and Prp reduces the uncertainty in the PDD measurement and the reference dose measurement.
C 2016 American Association of Physicists in Medicine. [http://dx.doi.org/10.1118/1.4941691]

Key words: absolute dosimetry, relative dosimetry, ionization chambers and flattening filter free

1. INTRODUCTION or required corrections be applied. Two main effects were

High dose rate flattening filter free (FFF) beams bring
their own challenges and considerations for accurate output
calibration and commissioning. In this paper, we examine
some of these issues in common cylindrical ionization
chambers and consider techniques for mitigating errors. The
detectors preferred for measurements on common flattened
Linac beams were either less appropriate for FFF beams
investigated for ion chambers of different sizes: (1) ion
recombination effects due to the higher dose per pulse (DPP)
in the absence of attenuation from the removed flattening
filter and (2) volume-averaging in absolute dose calibration
resulting from the forward peaked bremsstrahlung of the
unflattened beam. Ion recombination effects in FFF beams
have been observed by other groups,1–3 but guidance on
expected errors and how to mitigate them have been limited.

1507 Med. Phys. 43 (3), March 2016 0094-2405/2016/43(3)/1507/7/$30.00 © 2016 Am. Assoc. Phys. Med. 1507
1508 Sudhyadhom et al.: Absolute and relative measurement errors in FFF beams
T I. Ion chamber and diode specifications, measured values of Ppol, and k Q values calculated using the
PDDs in Table II and the work of McEwen (Ref. 9). Ppol values were independent of depth to 0.1% and the
average values across depths are shown.
Farmer CC13
Model PTW 30013 IBA CC13
Volume 0.6 cm3 0.13 cm3
Length 23.0 mm 5.8 mm
Radius 3.05 mm 3 mm
Ppol(6XFFF) 1.001 1.000
Ppol(10XFFF) 1.001 1.000
k Q (6XFFF) 0.995 0.997
k Q (10XFFF) 0.982 0.987
In addition, while volume-averaging is known to be present in
FFF beams,2 FFF commissioning would benefit from further
quantification of this effect and the availability of a simple
correction method.
These two effects may be accounted for with correction
factors.4,5 Volume-averaging effects can be determined
through 2D integration of the radial profile over the length and
width of an ion chamber.6 A simple, one-point measurement
correction formalism is evaluated here for the case of
the Farmer chamber in a FFF beam. Depth-dependent ion
recombination can also be corrected, by acquiring two scans
with different bias voltages on the field detector to produce
a depth-dependent Pion curve. However, this increases the
time needed to collect percent depth dose (PDD) data during
machine commissioning and impacts the accuracy of the
measurement. An addendum5 to TG-51 includes additional
guidance on the handling of ion recombination and volume-
averaging effects for physicists during the calibration of FFF
beams which underscores the need to determine the magnitude
of these effects. In this work, we investigated two sources of
systematic uncertainty found in commissioning FFF beams:
(1) high DPP induced increases in ion recombination and
(2) volume-averaging over the radial profile of a FFF beam for
larger length chambers. An analysis of depth-dependent ion
recombination and volume-averaging effects was performed
here for four ion chambers. The data presented in this work
offer simple ways to mitigate these errors or correct for
these effects and are intended for use in conjunction with
published resources in the commissioning of a TrueBeam
(Refs. 1 and 7) (Varian, CA) and other FFF machines such as
the Elekta Versa HD, CyberKnife, and TomoHD.
2. METHODS AND MATERIALS
A Varian TrueBeam STx with five different x-ray beams
(6X, 6XFFF, 10X, 10XFFF, and 15X) was used for these
measurements. With the exception of the cross-calibration,
dose measurements (and scans) were completed in an
IBA Blue Phantom scanning water tank (IBA dosimetry,
Germany). Cross-calibration was completed in CIRS Plastic
Water. The ion chambers used in this study were, in decreasing
volume, Farmer PTW TN30013 (0.6 cm3), IBA CC13
(0.13 cm3), IBA CC04 (0.04 cm3), and Pinpoint 3D (PP3D)
PTW TN31016 (0.016 cm3). Additional parameters of the
Medical Physics, Vol. 43, No. 3, March 2016
1508
CC04 PP3D Diode
IBA CC04 PTW 31016 PTW 60017
0.04 cm3 0.016 cm3 0.03 mm3
3.6 mm 2.9 mm 30 µm (thick)
2 mm 1.45 mm 0.6 mm
0.997 1.003
0.998 1.006
0.996 0.997
0.987 0.984
devices are summarized in Table I. They were selected to span
a range of sizes to measure the effect of volume-averaging on
the unflattened beams. Ion recombination effects across these
chambers also varied. In addition, a diode (PTW 60017) with
a sensitive volume of 0.03 mm3 and a radius of 0.6 mm was
used for profile scans in the FFF beams. While the CC13
is the only small volume chamber studied that fits the new
addendum to TG-51’s criteria as a reference chamber, other
small chambers may fit alternative criteria for a reference
chamber (such as those of the LNE-LNHB).8 The CC04, for
example, has been previously determined to be reasonable as
a reference chamber.8
Depth-dependent changes in ion recombination efficiency
result in a systematic skew in collected PDD data. These
errors were assessed in two manners: (1) indirectly through
measurement of Pion and (2) directly by measuring the PDD in
the clinical beam and with a nonclinical lower DPP beam. Pion
was measured for four depths (d max of 1.3 cm for 6XFFF and
2.2 cm for 10XFFF, 10, 20, and 35 cm in some cases) in the
two FFF beams using the two-voltage technique4 with +300
and +150 V bias. Lower DPP versions of the FFF beams were
created, with the dose servo turned off, by lowering the grid
voltage to reduce the beam current while repeaking the beam
with the RF driver voltage to compensate for reduced beam
loading. Once the DPP was reduced to approximately one-
tenth of the clinical value, large field (40×40 cm2) cross-plane
and in-plane profiles at d max were acquired and compared to
those of the clinical beam to ensure the beam energy remained
the same, a published technique.10 These profiles and PDDs
(for both the low DPP and clinical DPP beams) were taken
within a single session to minimize potential drift in the
machine that might occur over multiple days/sessions. Finally,
Pion and PDDs were acquired with the lower DPP beam for
the CC04 and CC13 ion chambers. These chambers were
chosen for the lower DPP measurements as they showed the
best and worst ion recombination characteristics, respectively,
of the chambers studied. For tabulated data collection (as
in the tables), the PDD data were acquired by taking
multiple discrete measurements with the ion chamber at a
fixed position. We acquired a relatively high number of MU
per measurement (1000 MU) and multiple measurements to
minimize the uncertainty due to measurement reproducibility
as opposed to scanning the beam. For illustrative purposes
alone, a scanned PDD was taken of both the clinical and low
1509 Sudhyadhom et al.: Absolute and relative measurement errors in FFF beams 1509

F. 1. Measured PDDs for the (A) 6XFFF and (B) 10XFFF beam using the CC13 and CC04 in the clinical and lower DPP beams for a 10 × 10 cm2 field. The
inset panel in the upper right shows the overprediction in the PDD at deep depth for the CC13 in both clinical beams. This effect was minimal in the low DPP
beams and for measurements involving the CC04.

DPP beams (as in Fig. 1). These scanned PDDs were not of another chamber, IC1, by
used for quantitative analysis as the signal to noise ratio for Co-60
these measurements was lower (especially for the low DPP ND, w,IC2
Co-60
beam). Pion,IC1 PTP,IC1 Pelec,IC1 Ppol,IC1 Mraw,IC1 kQ,IC1 ND, w,IC1
Volume-averaging due to the unflattened beam profile was = ,
Pion,IC2 PTP,IC2 Pelec,IC2 Ppol,IC2 Mraw,IC2 kQ,IC2
examined using two methods: (1) indirectly by averaging the (1)
profile of the field over the area of the chamber (through full
integration of the profile) and (2) directly by determination where Pion, PTP, Pelec, and Ppol are correction factors described
of the volume-averaged dose to water for the various ion in TG-51;4 Mraw is the uncorrected charge reading; kQ is
chambers studied. Step-by-step profile scans with 1 mm point the given beam’s quality factor from McEwen et al.9 The
spacing were taken for the 6XFFF and 10XFFF beams using cross-calibration was done in CIRS Plastic Water at a SSD
the PTW 60017 shielded diode with a 10 × 10 cm2 field of 100 cm, 10 × 10 cm2 field size, with the chamber axis
at 10 cm depth and at 100 cm SSD (standard conditions positioned at the center of the Plastic Water such that the
for TG-51 calibration). These data were used to create chamber axis was at 10 cm depth (standard TG-51 calibration
an analytical fit model for the 6XFFF and 10XFFF beam conditions). A 2 cm thick plastic slab specific to each chamber
profiles that span the dimensions of the detectors used in was used with the chamber inserted into a cavity of the outer
this study (Table I). A sixth order polynomial was fit to the dimension of the detector. This measured (averaged across
Co-60
average of the cross- and in-plane profiles and only used all flattened energies) ND, w value was used to determine the
to quantify the volume-averaging effect by integration. The absolute dose to water measured for the remaining unflattened
2D volume-averaging effect was estimated by calculating beams (6XFFF and 10XFFF) in all the non-Farmer ion
the double integral of this polynomial over the manufacturer chambers using the TG-51 formalism.
specified nominal length and width of the respective chamber. A simple analytical form for the correction factor,
This method for calculating the volume-averaging effect Prp, was introduced to account for the 1D radial profile
through integration of the profile has been used previously volume-averaging effect specifically in cylindrical Farmer-
by Kawachi et al. to measure this effect in the beam of the type chambers. If the dose, D, as a function of x can be
CyberKnife.6 modeled with the functional form of a quadratic (second
The volume-averaging effect can be directly measured order polynomial), then the dose is of the form
through absolute dose measurements as those intrinsically
D(x) = D0 1 + αx 2 ,


(2)
include volume-averaging effects due to the profile’s flatness
(or lack thereof). The effect of volume-averaging in an ion where D0 and α are constants. Let an ion chamber with length
chamber relative to another ion chamber can be directly L be placed in this region with its center at x = ∆. The average
determined from the absolute dose to water measured with dose, D, experienced by the chamber is then
each chamber. We used the TG-51 formalism to determine 
1 L/2+∆ L2
( )
Co-60
the Co-60 absorbed dose calibration factor, ND, w , in the ion D= D(x)dx = D0 1 + α + α∆ . 2
(3)
chambers studied relative to our ADCL calibrated Farmer L −L/2+∆ 12
chamber (PTW 30013) for the flattened beam energies (6X, For a centered chamber (∆ = 0), the volume-averaging
Co-60
10X, and 15X). The values for ND, w measured for each correction factor is then
flattened beam energy were averaged to create a single value ) −1
L2
Co-60
(
for each chamber. The experimentally measured ND, w for a
D0
Prp = = 1+α . (4)
given chamber, IC2, can be determined from a known value D(∆ = 0) 12

Medical Physics, Vol. 43, No. 3, March 2016

1510 Sudhyadhom et al.: Absolute and relative measurement errors in FFF beams 1510

T II. Ion recombination correction factor, Pion, at various depths for both photon beam energies studied. Clinical beam PDDs measured with the CC04 and
corrected for Pion are shown for reference. Uncertainties are the propagated (in quadrature) coefficient of variation for these measurements into the calculated
value of Pion. The last two columns (on the right) are Pion values measured from the lower dose per pulse FFF beams. The largest changes are shown in bold.

Pion (clinical beam) Pion (lower DPP)

Energy Depth (cm) PDD (%) Farmer CC13 CC04 PP3D CC13 CC04

6XFFF 1.3 100 1.009 ± 0.0004 1.015 ± 0.0005 1.004 ± 0.0005 1.008 ± 0.0006 1.003 ± 0.0005 1.002 ± 0.0005
10 62.8 1.006 ± 0.0004 1.009 ± 0.0005 1.003 ± 0.0005 1.008 ± 0.0006
20 34.1 1.004 ± 0.0004 1.006 ± 0.0005 1.001 ± 0.0005 1.008 ± 0.0006
35 14.0 1.003 ± 0.0005 1.003 ± 0.0005 1.001 ± 0.0005 1.002 ± 0.0005
10XFFF 2.2 100 1.015 ± 0.0004 1.026 ± 0.0005 1.008 ± 0.0005 1.012 ± 0.0006 1.004 ± 0.0005 1.002 ± 0.0005
10 70.4 1.011 ± 0.0004 1.018 ± 0.0005 1.006 ± 0.0005 1.010 ± 0.0006
20 42.4 1.006 ± 0.0004 1.012 ± 0.0005 1.004 ± 0.0005 1.008 ± 0.0006
35 20.5 1.010 ± 0.0005 1.003 ± 0.0005 1.001 ± 0.0005 1.002 ± 0.0005

Equation (4) can be further reduced and approximated to
) ) −1
( (
L
Prp = OAR √ ≈ (OAR(0.29L))−1,
12
which is the inverse of the off-axis ratio (along
√ the long axis
of the chamber) found at a distance of L/ 12 (where L is
the nominal chamber length) for the given unflattened photon
beam under calibration conditions. The need for this type
of correction factor, Prp, was suggested in the addendum to
TG-51.5
3. RESULTS
3.A. Ion collection efficiency
The values of Pion measured with the two-voltage method
in the unflattened clinical and lower DPP beams are listed in
Table II. Pion varies significantly with depth depending on the
ion chamber model. Lower doses per pulse result in a higher
ion collection efficiency. The CC13 ion chamber exhibited
the largest degree of change in Pion versus depth for both FFF
beams, varying by 1.3% between d max and 35 cm depth for
the 6XFFF beam, and 1.6% for the 10XFFF beam. The effect
was the largest in the 10XFFF beam where the DPP was
the highest. The ion collection efficiency of the two smallest
chambers studied in this work exhibited a minimal (<1%)
dependence on depth.
The reduced ion collection efficiency with depth affects
the measured PDD curve. The 6XFFF and 10XFFF PDDs
measured with the CC13 and CC04 for the clinical and low
(5) DPP beams are shown in Fig. 1 to qualitatively illustrate
the effect of ion recombination on measurements with these
chambers. In line with the Pion results, the PDD for the CC13
(in the clinical DPP beam) shows a marked increase at deep
depths due to a lower measured dose at d max. The ratio of
the clinical and low DPP PDDs measured with the CC13
and smaller CC04 chamber is shown in Tables III and IV
along with the ratio of Pion from Table II. The effect of the
higher DPP of the clinical beams is accounted for by the
variation in Pion with depth. Furthermore, the PDD measured
with the CC04 in the clinical beam agrees with the low DPP
result to within 0.5% to a depth of 35 cm, showing that this
smaller detector can be used for FFF PDD measurement up
to 10 MV without correction for variation in ion collection
efficiency with depth. The CC13 measured PDDs exhibit
the greatest errors due to ion recombination as shown in
Tables III and IV. In the low DPP beam, PDDs values
measured with the CC13 and CC04 agree to within 0.5%.
3.B. Volume-averaging
Profiles for both the 6XFFF and 10XFFF beams were
acquired to quantitatively estimate the effect of volume-
averaging on the ion chambers studied (Fig. 2). The profiles
were fit to a sixth order polynomial (shown as solid lines

T III. Relative percent errors in the measured PDD due to ion recombination inefficiency in the clinical
beam. PDDs corrected for Pion from Table II are shown for reference. Uncertainties are the propagated coefficient
of variation for these measurements from the calculated values of Pion. The last column is the ratio of uncorrected
PDDs between the CC13 and CC04. The largest changes are shown in bold.

Ratio of PDD
Energy Depth (cm) PDD (%) Farmer (%) CC13 (%) CC04 (%) PP3D (%) (CC13/CC04)

6XFFF 10 62.8 0.3 ± 0.06 0.6 ± 0.07 0.3 ± 0.07 0.2 ± 0.08
20 34.1 0.6 ± 0.06 1.0 ± 0.07 0.5 ± 0.07 0.3 ± 0.08
35 14.0 1.3 ± 0.07 0.3 ± 0.07 1.005 ± 0.0007
10XFFF 10 70.4 0.4 ± 0.06 0.9 ± 0.07 0.3 ± 0.07 0.1 ± 0.08
20 42.4 1.0 ± 0.06 1.5 ± 0.07 0.5 ± 0.07 0.4 ± 0.08
35 20.5 1.6 ± 0.07 0.5 ± 0.07 1.011 ± 0.0007

Medical Physics, Vol. 43, No. 3, March 2016

1511 Sudhyadhom et al.: Absolute and relative measurement errors in FFF beams
T IV. Relative percent errors in the measured PDD due to ion recombination inefficiency in the lower
DPP beam. PDDs corrected for Pion from Table II are shown for reference. The last column is the ratio of
uncorrected PDD between the CC13 and CC04. Uncertainties are the propagated coefficient of variation for these
measurements from the calculated values of Pion.
Energy Depth (cm) PDD (%)
6XFFF 35 14.0
10XFFF 35 20.5
in Fig. 2). R2 values for both fits were greater than 0.998
within the dimensions of the largest ion chamber. These
fits were used to analytically calculate the percent change
in dose expected due to volume-averaging for the 6XFFF
and 10XFFF profiles in two dimensions. These results were
compared with the measured percent change in dose for the
6XFFF and 10XFFF beams as determined through cross-
calibration [as in Eq. (1)]. The air cavity of the Farmer
chamber used in this study (PTW 30013) is 23 mm in
length with a diameter of 6.1 mm, according to manufacturer
specification. The calculated volume-averaging effect of this
chamber (relative to a point measurement) was calculated to
be −0.4% and −0.7% for 6XFFF and 10XFFF, respectively,
for the 10×10 cm2 field at 100 SSD and 10 cm depth. Table V
shows a summary of the volume-averaging effects of the three
other chambers examined in this study as measured (versus the
Farmer chamber volume-averaging effect), as calculated from
our full integration of the analytical fit in Fig. 2 (relative to the
Farmer chamber volume-averaging effect), and as calculated
from the simple expression for Prp = (OAR(0.29L))−1. At
chamber dimensions as small as the CC13 and smaller, the
volume-averaging effect is minimal and potentially within
the overall uncertainty of our measurements. The Prp for the
non-Farmer chambers studied in this work were minimal.
4. DISCUSSION
In this work, two characteristics of unflattened beams were
examined to determine their effect on absolute and relative
dosimetry. Overall, changes in Pion were more pronounced
F. 2. Average profiles (of cross and in-plane) for the 6XFFF and 10XFFF
beams for a 10 × 10 cm2 field. The solid lines are a sixth order polynomial fit
to the respective profiles data. The vertical dotted lines represent the length
of the largest ion chamber studied in this work (Farmer PTW TN30013).
Medical Physics, Vol. 43, No. 3, March 2016
1511
CC13 (%) CC04 (%) Ratio of PDD (CC13/CC04)
0.3 ± 0.07 0.1 ± 0.07 1.002 ± 0.0007
0.3 ± 0.07 0.0 ± 0.07 1.000 ± 0.0007
in the 10XFFF beam due to higher doses per pulse versus
6XFFF (and, more generally, most typical flattened beams).
This effect results in about a 1.6% relative increase in the
PDD at a depth of 35 cm for 10XFFF using the CC13. The
CC13 chamber had the most pronounced change in Pion with
depth of the chambers studied. This effect in FFF beams has
been studied in the past and other groups have found similar
results for other ion chambers.1–3 Chang et al.,1 have reported
Pion at a depth of d max for the CC13 to be 1.015 and 1.023
for 6XFFF and 10XFFF, respectively. These values are in line
with our currently reported Pion values of 1.015 and 1.026 for
6XFFF and 10XFFF, respectively.
The Farmer chamber had a substantial volume-averaging
effect of −0.7% at 10XFFF that was consistent between
calculated and measured values. Our measured volume-
averaging effects were within 0.3% from those that were
calculated for 6XFFF. A possible source of this small
discrepancy is due to the determination of relative detector
response in the flattened and unflattened beams, based on
the kQ values used in our determination of ND, Co-60
w . Xiong
11
and Rogers previously studied uncertainties in kQ for lower
energy FFF beams with Monte Carlo simulations. This effect
is most pronounced in our results for the 6XFFF beam, which
is in line with their findings.
In comparison, the CyberKnife 6XFFF beam volume-
averaging effects were found to be up to 1.5% (Refs. 6, 12,
and 13) for Farmer-type chambers. This volume-averaging
effect for CyberKnife is significantly larger than found in
either TrueBeam FFF beam studied. The difference between
these two beams may be due to the shorter source-to-detector
distance (80 cm compared to 100 cm) and a more rapid
reduction in off-axis ratio, presumably due to different targets.
The magnitude of this effect may be significant even for low
energy (6 MV or less) FFF beams and should be quantified for
each beam, at each institution. Cross-calibration of smaller
chambers (by our method) was shown here as a method
to experimentally observe the volume-averaging effect. For
absolute dosimetry, it is preferable to use a chamber calibrated
at an ADCL.
In order to mitigate the volume-averaging effect, we
propose a new simple analytic correction factor to be used
during TG-51 absolute dose calibration to help account for
this chamber geometry based volume-averaging effect,
( ( ) ) −1
L
Prp = OAR √ ≈ (OAR(0.29L))−1, (6)
12
which is the inverse of the off-axis ratio (along
√ the long axis
of the chamber) found at a distance of L/ 12 (where L is
1512 Sudhyadhom et al.: Absolute and relative measurement errors in FFF beams 1512

T V. Volume-averaging effects were measured (via cross-calibration) and calculated (through full 2D in-
tegration of radial profiles) for the 6XFFF and 10XFFF beams. The tabulated measured and calculated doses
are shown as percent dose differences relative to the dose in the Farmer chamber. Uncertainties in the measured
Co-60.
volume-averaging effect represent the quadrature propagated coefficient of variation for the flattened beam N D, w
The Prp rows represent the calculated volume-averaging correction factor from (1) “full” 2D integration of the
profile or (2) “simple” estimation using Prp = (OAR(0.29L))−1.

Energy Farmer CC13 CC04 PP3D

6XFFF Measured — 0.2% ± 0.1% 0.1% ± 0.1% 0.3% ± 0.06%

Calculated — 0.3% 0.4% 0.4%
Prp (full) 1.004 1.000 1.000 1.000
Prp (simple) 1.005
10XFFF Measured — 0.7% ± 0.1% 0.7% ± 0.1% 0.7% ± 0.06%
Calculated — 0.7% 0.7% 0.7%
Prp (full) 1.007 1.001 1.000 1.000
Prp (simple) 1.007

the nominal chamber length) for the given unflattened photon
beam under calibration conditions. This measurement when
completed in a small length chamber would result in minimal
volume-averaging effects. In the case of a highly asymmetric
profile, both sides of the profile should be averaged. As a
final note, the chamber or diode utilized to acquire the OAR
value for calculating Prp does not need to be the same as
the one used for dose calibration; however, the shift applied
must be based on the length of the ion chamber used for
calibration. For the 10XFFF beam measured with the Farmer
chamber studied in this work, the measured Prp using this
simplified formalism would be 1.007 (as determined by diode
measured profiles). This value is in line with our measured
and calculated values for the volume-averaging effect. For
smaller chambers, this correction may be a minor effect and
potentially unnecessary. The recently published addendum to
TG-51 suggests that microchambers (smaller than 0.05 cm3)
may not fit with the criteria for reference dosimetry. As
some microchambers [such as the CC04 (Ref. 8)] appear to
have characteristics that are at the boundary for reference
dosimetry, it may be preferable to apply a correction factor
(such as the one proposed in this work) to account for
volume-averaging effects to a dosimeter with well-established
reference dosimetry characteristics.
5. CONCLUSIONS
Corrections may be necessary to reduce measurement
errors in unflattened beams when using detectors commonly
used to make measurements in flattened beams. If these
corrections are not properly applied, worst-case errors in
absolute (and relative) dose calibration approaching 2% (1.6%
for the TrueBeam 10XFFF) may result for FFF beams with
energies ranging from 6 to 10 MV, potentially larger than
2% at higher energies if the current trend of higher dose per
pulse holds with future machines. Alternatives to some of
the larger and most widely used dosimeters (e.g., CC13 and
Farmer) should be carefully considered and tested9 for relative
and absolute dose measurements, respectively. Our proposed
formalism for Prp may be used to mitigate volume-averaging
errors in the absolute calibration of unflattened beams. In
the case of a FFF beam well modeled by a second order
polynomial, usage of the proposed Prp correction factor during
beam calibration with a large volume reference ionization
chamber (such as a Farmer-type) is an accurate alternative to
using a small volume chamber.
The recommendations of the authors, based off this work,
are to minimize the effect of ion recombination in PDD and
profile measurement of FFF beams not through measurement
of Pion but through specific choice of ion chambers for
relative dosimetry. Measurement of Pion across all PDDs and
profiles would nearly double the data acquisition time during
commissioning. In this study, we evaluated a range of ion
chambers to show that mitigation of these uncertainties may
be possible simply by choosing an ion chamber without a
significantly depth-dependent ion recombination effect. In the
case of absolute dosimetry, we recommend choosing an ion
chamber that has a relatively low ion recombination effect
(Pion close to 1). When one must measure Pion, as is the case
of machine calibration/absolute dosimetry, the two-voltage
technique (as recommended in TG-51) for measurement
of Pion is straightforward and should continue to be
followed.
The data reported in this work can be used in conjunction
with guidance provided by TG-51 and the recently published
addendum to determine how to best account for these effects.
Physicists should be aware that certain ion chambers might be
less than ideal for absolute dose calibration and relative dose
measurements in FFF beams. As a result, recommendations
for a preferred ion chamber to use for a particular application
are, at best, complicated. Regardless of volume, ion chambers
should be evaluated on an application specific basis. As
demonstrated in this work, larger ion chambers that fulfill the
stringent reference dosimeter criteria may still not be ideal for
relative dosimetry in FFF beams without proper correction.
On the other hand, smaller volume ion chambers that may
be preferred for scanning purposes may be problematic for
others reasons5,9 or even for simple reasons such as poor
signal to noise. With the growing popularity of FFF beams,
special evaluation and care should be taken to minimize the
errors associated with dosimetry of FFF beams.

Medical Physics, Vol. 43, No. 3, March 2016

1513 Sudhyadhom et al.: Absolute and relative measurement errors in FFF beams
a)Author to whom correspondence should be addressed. Electronic mail:
Atchar.Sudhyadhom@ucsf.edu
1Z. Chang, Q. Wu, J. Adamson, L. Ren, J. Bowsher, H. Yan, A. Thomas, and
F. F. and yin, “Commissioning and dosimetric characteristics of TrueBeam
system: Composite data of three TrueBeam machines,” Med. Phys. 39,
6981–7018 (2012).
2S. F. Kry, R. Popple, A. Molineu, and D. S. Followill, “Ion recombina-
tion correction factors (Pion) for Varian TrueBeamhigh-dose-rate therapy
beams,” J. Appl. Clin. Med. Phys. 13 (2012).
3Y. Wang, S. B. Easterling, and J. Y. Ting, “Ion recombination corrections of
ionization chambers in flattening filter-free photon radiation,” J. Appl. Clin.
Med. Phys. 13 (2012).
4P. R. Almond, P. J. Biggs, B. M. Coursey, W. F. Hanson, M. S. Huq, R.
Nath, and D. W. Rogers, “AAPM’s TG-51 protocol for clinical reference
dosimetry of high-energy photon and electron beams,” Med. Phys. 26,
1847–1870 (1999).
5M. McEwen, L. DeWerd, G. Ibbott, D. Followill, D. W. Rogers, S. Seltzer,
and J. Seuntjens, “Addendum to the AAPM’s TG-51 protocol for clinical
reference dosimetry of high-energy photon beams,” Med. Phys. 41, 041501
(20pp.) (2014).
6T. Kawachi, H. Saitoh, M. Inoue, T. Katayose, A. Myojoyama,
and K. Hatano, “Reference dosimetry condition and beam quality
correction factor for CyberKnife beam,” Med. Phys. 35, 4591–4598
(2008).
Medical Physics, Vol. 43, No. 3, March 2016
1513
7C. Glide-Hurst, M. Bellon, R. Foster, C. Altunbas, M. Speiser, M. Altman, D.
Westerly, N. Wen, B. Zhao, M. Miften, I. J. Chetty, and T. Solberg, “Commis-
sioning of the Varian TrueBeam linear accelerator: A multi-institutional
study,” Med. Phys. 40, 031719 (15pp.) (2013).
8M. Le Roy, L. de Carlan, F. Delaunay, M. Donois, P. Fournier, A. Ostrowsky,
A. Vouillaume, and J. M. Bordy, “Assessment of small volume ionization
chambers as reference dosimeters in high-energy photon beams,” Phys.
Med. Biol. 56, 5637–5650 (2011).
9M. R. McEwen, “Measurement of ionization chamber absorbed dose k
Q
factors in megavoltage photon beams,” Med. Phys. 37, 2179–2193 (2010).
10D. L. Sawkey and B. A. Faddegon, “Determination of electron energy, spec-
tral width, and beam divergence at the exit window for clinical megavoltage
x-ray beams,” Med. Phys. 36, 698–707 (2009).
11G. Xiong and D. W. Rogers, “Relationship between %dd(10) and stopping-
x
power ratios for flattening filter free accelerators: A Monte Carlo study,”
Med. Phys. 35, 2104–2109 (2008).
12P. Francescon, W. Kilby, N. Satariano, and S. Cora, “Monte Carlo simulated
correction factors for machine specific reference field dose calibration and
output factor measurement using fixed and iris collimators on the Cy-
berKnife system,” Phys. Med. Biol. 57, 3741–3758 (2012).
13E. Pantelis, A. Moutsatsos, K. Zourari, W. Kilby, C. Antypas, P. Papagiannis,
P. Karaiskos, E. Georgiou, and L. Sakelliou, “On the implementation of a
recently proposed dosimetric formalism to a robotic radiosurgery system,”
Med. Phys. 37, 2369–2379 (2010).