MLC dosimetric characteristics for small field and IMRT applications

J. E. Bayoutha) and S. M. Morrill

University of Texas Medical Branch, Department of Radiation Oncology, Galveston, Texas 77555
共Received 17 October 2002; accepted for publication 29 June 2003; published 25 August 2003兲
The objective of this work was to measure the performance characteristics of a double-focus
multileaf collimator 共MLC兲 for intensity modulated radiation therapy 共IMRT兲, specifically the
variation in penumbra and leakage for narrow fields as a function of field position over a 20⫻27 cm
space available for segmented MLC IMRT. Measurements were made with 6 MV x rays through a
MLC containing 29 leaf pairs 共27 pairs of 1 cm width兲, and EDR2 film at 10 cm depth in solid
water at 100 cm SAD. Films were digitized with 0.17 mm resolution and converted to dose.
Interleaf and intraleaf transmission were measured along 11 vertical profile locations. Leaf-end
transmission was measured along horizontal profiles for each of 9 different leaf abutments, travel-
ing over a 20 cm range. In-plane penumbra measurements were made through a single leaf re-
tracted, for 7 different leaves. Cross-plane penumbra 共leaf-end兲 measurements were made for all 27
leaf pairs, where the 1 cm field width was placed in 11 different off-axis positions 共20 cm range兲.
Interleaf leakage 共range 1.0%–1.5%兲, intraleaf transmission 共range 0.6%–0.8%兲, and leaf-end
transmission 共range 0.8%–2.7%兲 were consistent for all leaf pairs at a given abutment position. The
penumbra for these 1-cm-wide fields was measured to be 0.36 cm⫾0.03 cm for 99% of the
measurements. In conclusion, the penumbra and leakage of the double-focus MLC were remarkably
consistent for the range of leaf positions studied, producing dosimetric characteristics that are well
suited for IMRT segments where opposing leaf pairs are often separated by 10 mm or less.
© 2003 American Association of Physicists in Medicine. 关DOI: 10.1118/1.1603743兴

Key words: multileaf collimator, double-focused, leakage, transmission, penumbra

I. INTRODUCTION are created using a MLC. Although the general dosimetric

Three-dimensional conformal radiotherapy 共3DCRT兲 has be-
come the standard of care for many types of cancers receiv-
ing radiation therapy. The Radiation Therapy Oncology
Group now relies on this delivery technique in protocols for
the treatment of cancers of the prostate, lung, glioblastoma,
and oropharynx. Optimal dose distributions created using
3DCRT produces isodose surfaces that conform closely to
the planning target volumes, and are often designed to ac-
complish conformal avoidance, i.e., creating steep dose-
gradients between the planning target volumes and nearby
critical structures. The utility of steep dose-gradients has
long been exploited in stereotactic radiosurgery; Podgorsak
et al.1 studied the dose-gradient as a metric when comparing
the dose fall-offs outside the target volume for various radio-
surgical techniques. Further efforts in conformal avoidance
have led to the inclusion of ‘‘island blocks’’ within the treat-
ment portal, as described by Eisbrauch et al.,2 Frass et al.,3
Xiao et al.,4 and Verhey.5 This treatment planning goal is a
primary objective for a subset of 3DCRT, intensity modu-
lated radiation therapy 共IMRT兲.6
One common technique of creating IMRT beams is to
utilize a segmentation algorithm that converts the ideal flu-
ence map into deliverable beam segments using a multileaf
collimator 共MLC兲.7 During the segmentation process the leaf
ends of the MLC, which are used to create the resultant
segment shapes, may be separated by as little as 5–10 mm
for the delivery of IMRT. One of the many dosimetric chal-
lenges for treatment planning systems is to model the dose
distributions created by these segments accurately when they
characteristics of MLCs have been described by others8 –14
the small field dosimetric characteristics of the MLC are not
well known. Furthermore, their behavior for the range of
off-axis positions often utilized in IMRT is not well known.
For example, a given MLC leaf has penumbra off the edge of
the leaf and the end of the leaf. These two sources of pen-
umbra may be different, and each penumbra type potentially
changes both in the in-plane and cross-plane direction. Pre-
vious studies have not provided data on the magnitude of
these changes, and treatment planning systems typically ig-
nore these effects. The objective of this work was to measure
the MLC performance characteristics for IMRT, specifically
the variation in penumbra and leakage/transmission for nar-
row fields as a function of field position over the allowable
position space for IMRT delivery.
II. MATERIALS AND METHODS
Exposures were made with a Siemens 共Siemens Medical
Systems, Concord, CA兲 Primart linear accelerator, containing
29 leaf pairs of which the central 27 leaf pairs are 1 cm wide
centered on the central axis. A complete description of the
design of this double-focused multileaf collimator was pro-
vided by Das et al.14 Each leaf is capable of traveling 10 cm
beyond the central axis, so when considering this limitation
for opposing leaf banks, this design provides 20 cm of al-
lowable position space in the direction of leaf travel and 27
cm perpendicular to leaf travel. For this study the 6.5-cm-
wide leaf pairs 共1 and 29兲 are not considered. The Primart is
a single energy linear accelerator that produces 6 MV x rays,

2545 Med. Phys. 30 „9…, September 2003 0094-2405Õ2003Õ30„9…Õ2545Õ8Õ$20.00 © 2003 Am. Assoc. Phys. Med. 2545
2546 J. E. Bayouth and S. M. Morrill: MLC dosimetric characteristics 2546

utilizing a straight waveguide with rotating target that yields

a sharper focal spot. The dosimetric characteristics of this
accelerator are similar to the Siemens Primus accelerator.15
Kodak EDR2 ReadyPak film 共Eastman Kodak Company,
Rochester, NY兲 was used for all radiation dose measure-
ments. This film was chosen over gafchromic film for these
measurements because of its size and cost, and over Kodak
XV film because of its wider range and reduced energy
dependence,16 –19 likely due to a reduction in the amount of
silver halide. The films were placed at 10 cm depth in a solid
water phantom 共30⫻30⫻22 cm兲 and 100 cm SAD. All films
were processed together and a unique H&D curve was cre-
ated for each group of measurements. H&D curves were gen-
erated using a single film calibration method similar to that
presented by Childress et al.,19 where eight nonoverlapping
portals of different dose are exposed on a single film. Al-
though the film depth utilized for the H&D curve was
slightly different 共7.5 cm兲 from the measurement depth for
the remainder of the films 共10 cm兲, EDR2 film is insensitive
to the depth difference.19,20 Each film was digitized with 0.17
mm resolution using the RIT113 film dosimetry system 共Ra-
diological Imaging Technology, Colorado Springs, CO兲 and
converted to dose 共optical density range 0.3–1.6兲. The digital
images were manipulated to reduce noise 共3⫻3 medial filter兲
and remove rotation of the film data with respect to the col-
limator coordinate system.
A. Intra- and Interleaf leakageÕtransmission
Three types of leakage/transmission were measured: inter-
leaf, intraleaf, and leaf-end.7 For all measurements, the
leakage/transmission was reported as the percent of the un-
blocked dose. The percentage of the unblocked dose was
computed as the percentage of dose recorded on the film
divided by the dose that would have been delivered to the
film at the same depth and monitor units from a 10⫻10 cm
field. To measure the interleaf and intraleaf transmission, all
27 1 cm leaf pairs were programmed to fully close, produc-
ing a 0 cm gap between opposing leaves. An example of the
field shape is shown in Fig. 1共a兲. The Y jaws travel in the
in-plane direction, while the X jaws travel in the cross-plane
direction. The Y2 jaw is on the side of the central axis closer
to the electron gun, while the X1 leaf bank is on the left-hand
side in the beams-eye-view. In Fig. 1共a兲 leaf pairs 2 through
28 are abutting each other 10 cm off-axis, while leaf pairs 1
and 29 were opened 3 mm, which allowed the Y1 and Y2
jaw set to 13.6 cm and allow full exposure of the leaves. A
single radiographic film was exposed to 6 MV x rays for
6000 monitor units, which produced the image in Fig. 1共b兲.
Eleven vertical dose profiles were measured across this im- FIG. 1. MLC intra- and interleaf leakage/transmission test pattern. 共a兲 MLC
age, 2 cm apart from one another, perpendicular to the leaf field shape in the beams-eye view. Note the Y-jaws 共not shown兲 are set 1
motion direction. The first profile was 5 mm medial to the mm distal to leaf pairs 2 and 28. 共b兲 Resultant film 共dose兲 from 共a兲.
match of the X1 and X2 leaf bank, for the field shape shown
in Fig. 1共a兲, and the eleventh 10.5 cm from the CAX toward
B. Leaf-end leakageÕtransmission
the X2 leaf bank. A single profile shows the difference in
leakage/transmission for each leaf pair, while multiple verti- Nine sequential field shapes were utilized to assess leaf-
cal profiles across the image in Fig. 1共b兲 from left to right end transmission 共i.e., the transmission of photons through
will show the variation of leakage/transmission in the cross- abutted leaf pairs兲. Ten separate films were exposed: one for
plane direction. each of the nine fields, and a tenth was exposed for all nine

Medical Physics, Vol. 30, No. 9, September 2003

2547 J. E. Bayouth and S. M. Morrill: MLC dosimetric characteristics 2547

FIG. 2. MLC leaf end leakage/transmission test pattern. This image shows
the resultant dose pattern for all 9 abutment locations for all 27 1 cm leaf
pairs.

fields on a single film. The shape of each field resembled that

shown in Fig. 1共a兲, where the abutment location was trans-
lated by 2.5 cm steps over the course of the sequence. Each
exposure was for 500 MU; Figure 2 shows the single film
exposed for all nine leaf abutment locations. The leaf-end
transmission was not determined from the image in Fig. 2,
but rather from each film receiving a single exposure.
Twenty-seven horizontal dose profiles were measured, one
through the center of each leaf pair, for each of the nine
abutment locations and the maximum transmission through
each abutted leaf pair was recorded. Thus, for an individual
leaf pair we could analyze the leaf-end transmission and its
variation across the IMRT position space.

C. In-plane penumbra
Two basic types of penumbra 共80% to 20%兲 were studied,
in-plane and cross-plane penumbra. Fig. 3共a兲 shows one of
seven field shapes used to measure in-plane penumbra,
where a single leaf on the X2 leaf bank was fully retracted
共in this example, leaf 10兲. In this figure, the in-plane penum-
bra is dependent on the Y2 side of leaf 9 共tongue兲 and the Y1
side of leaf 11 共groove兲. The penumbra is dependent on the
sides of the leaves only when the Y jaws are not abutting the
open port, as is often the case with segmental MLC IMRT.6
A new film was exposed for each of seven leaves retracted
共2, 6, 10, 15, 20, 24, and 28兲. A dose profile perpendicular to
the direction of leaf travel 共vertical兲 passing through the cen-
tral axis was acquired for each film to measure the in-plane
variation of in-plane penumbra. In addition, the cross-plane
variation of in-plane penumbra was analyzed. Eleven total
vertical dose profiles were acquired on each film 共2 cm spac-
ing, from 9.5 cm toward X1 to 10.5 cm toward X2兲.
FIG. 3. MLC penumbra tests. 共a兲 MLC field shape in beams-eye-view 共1 of
7兲 used to assess in-plane penumbra. 共b兲 MLC field shape in beams-eye
view 共1 of 11兲 used to assess cross-plane penumbra.
D. Cross-plane penumbra
Figure 3共b兲 shows the beams-eye-view of one of eleven
field shapes used to measure cross-plane penumbra 共leaf-
end兲, where leaves are separated by 1 cm. In this case, the 1
cm opening is centered at 10 cm off the central axis, toward
the X1 jaw. A new film was exposed for each of eleven field
shapes, where the center of the exposed port moved from 10
cm toward the X1 jaw to 10 cm toward the X2 jaw in 2 cm
increments. A dose profile parallel to the direction of leaf
travel 共horizontal兲 passing through the central axis was ac-

Medical Physics, Vol. 30, No. 9, September 2003

2548 J. E. Bayouth and S. M. Morrill: MLC dosimetric characteristics
FIG. 4. Inter- and intraleaf leakage across all 27 1 cm leaves, along eleven
different cross-plane locations.
quired for each film to measure the cross-plane variation of
cross-plane penumbra. In addition, the in-plane variation of
cross-plane penumbra was analyzed. A total of 27 horizontal
dose profiles were acquired, one through each leaf pair 共2
through 28兲, for each of the eleven leaf positions.
III. RESULTS
A. Intra- and Interleaf leakageÕtransmission
Interleaf leakage is shown in Fig. 4, where the peak of
each dose profile represents the interleaf leakage. Each line
represents a dose profile perpendicular to the direction of leaf
travel, for one of several off-axis 共cross plane兲 positions.
Although the general trends of the data are shown in Fig. 4,
detailed analysis of the results requires specific data provided
in Tables I–IV. The mean value and the standard deviation of
interleaf leakage along each profile are listed in Table I. The
average interleaf leakage for all 27 peaks was 1.1%⫾0.1%,
ranging from 0.9% to 1.3% 共5–95 percentile兲. These results
show that the in-plane variation of interleaf leakage is small;
with a standard deviation of approximately 10% of the mean
value. These results are also observed graphically, as the
variation from one peak to the next for a given line is small.
A spike in the results is seen at an in-plane distance of 0.5
and 1.5 cm from the central axis. This increase in leakage
corresponds to the space between leaf pairs 13–14 and 14 –
15. Inspection of Fig. 1共b兲 reveals an enhanced interleaf
leakage dose at this location that increases in magnitude fur-
ther away from the central axis. This may be caused by a
slight variation in the mechanical fabrication of leaf X2.14
and/or spatially dependent variation in the separation be-
tween the leaves allowing additional leakage. Not to be con-
fused with the leaf position calibration, this variation could
TABLE I. Interleaf leakage variation in the in-plane direction. Each column shows the average and standard deviation of leakage in the in-plane direction for
a specified cross-plane location.
Cross-plane distance from central axis 共cm兲
9.5 7.5 5.5 3.5 1.5
Average 共%兲 1.11 1.14 1.14 1.12 1.11
Standard dev 共%兲 0.12 0.13 0.13 0.13 0.14
Medical Physics, Vol. 30, No. 9, September 2003
2548
result from a leaf that is too narrow or not angled properly
along its ‘‘double-focus’’ fan line. Table II shows interleaf
leakage between each leaf pair, obtained by averaging the
data in Fig. 4 for all cross-plane measurements 共i.e., for all
lines兲 for each peak. Each column shows the average and
standard deviation of leakage in the cross-plane direction be-
tween a specific leaf-pair. With respect to Fig. 4, these results
show the variation of inter-leaf leakage for a given peak for
all curves. Here the largest variation is seen between leaves
14 –15 and 28 –29, where their standard deviation is nearly
doubled, as seen qualitatively by visual inspection of Fig.
1共b兲. The leakage through the first and last leaf-pair show
less transmission, possibly due to the reduced transmission
allowed by the proximity of the Y jaw. The variation in the
cross-plane direction is less than that seen in the in-plane
direction (p⫽0.004, two-sample t test兲, demonstrating the
double-focus MLC design yields leakage between a leaf and
its neighbor that is slightly more consistent in the direction
parallel to the direction of leaf travel than perpendicular.
The valley of each dose profile is the sum of penumbra
and intraleaf transmission, which was between 0.6% and
0.9% 共5–95 percentile兲 with a mean of 0.8%⫾0.1%. Table
III provides intraleaf leakage variation in the in-plane direc-
tion. The standard deviation of leakage in the in-plane direc-
tion for each cross-plane location varies by less than 0.1%.
Table IV shows the results for intraleaf leakage through each
leaf at several locations in the cross-plane direction. These
results demonstrate the standard deviation of leakage in the
cross-plane direction for leaf varies by approximately 0.05%.
B. Leaf-end leakageÕtransmission
The effects of leaf-end transmission can be seen qualita-
tively in Fig. 2, where the vertical lines are caused by leaf-
end transmission and the horizontal lines are the result of
interleaf transmission. Although the two sets of transmission
lines appear similar in exposure, the interleaf transmission
occurred for all 9 exposures, where the horizontal lines are
the sum of leaf-end transmission occurred for one of the 9
exposures and intraleaf transmission for the remaining 8 ex-
posures. Quantitative results of the leaf-end transmission
measurements are shown in Figs. 5共a兲 and 5共b兲, as taken
from individual films of leaf abutment. The average leaf-end
transmission for all leaf pairs at all measurement points was
1.9%⫾0.3%, with a range of 1.4%–2.5% 共5–95 percentile兲.
Each line in Fig. 5共a兲 represents the transmission values be-
tween leaf pairs 2–28 at a single cross-plane location. Minor
variation is seen along a given line, indicating little in-plane
variation of leaf-end transmission. The grouping of these
⫺0.5 ⫺2.5 ⫺4.5 ⫺6.5 ⫺8.5 ⫺10.5
1.10 1.09 1.06 1.03 0.99 0.96
0.14 0.14 0.14 0.14 0.14 0.13
2549 J. E. Bayouth and S. M. Morrill: MLC dosimetric characteristics 2549

TABLE II. Interleaf leakage variation in the cross-plane direction. Each column shows the average and standard deviation of leakage in the cross-plane
direction between a specific leaf-pair.

Gun side of leaf

2 3 4 5 6 7 8 9 10

Average 共%兲 0.57 0.96 0.96 1.01 1.02 1.13 1.08 1.05 1.12
Standard dev 共%兲 0.06 0.05 0.05 0.07 0.06 0.07 0.07 0.04 0.07

Gun side of leaf

11 12 13 14 15 16 17 18 19

Average 共%兲 1.09 1.13 1.11 1.24 1.42 1.10 1.10 1.12 1.13
Standard dev 共%兲 0.05 0.06 0.04 0.11 0.06 0.07 0.04 0.07 0.07

Gun side of leaf

20 21 22 23 24 25 26 27 28

Average 共%兲 1.11 1.10 1.05 1.10 1.13 1.04 1.05 1.02 1.12
Standard dev 共%兲 0.05 0.05 0.04 0.06 0.06 0.07 0.08 0.07 0.12

lines around the mean value of 1.9% demonstrates minor
cross-plane variation. There was no difference between in-
plane and cross-plane leaf-end transmission (p⫽0.4). The
leaf-end transmission at 10 cm off the central axis for all leaf
pairs is approximately 0.5% less than the mean, while the
central leaves are 0.5% greater than the mean for the central
leaf pairs. Figure 5共b兲 shows the frequency distribution of
leaf-end transmission, which is nearly normally distributed
about the mean with the 99% confidence interval calculated
to be 1.9%–2.0%. Seventy percent of the measured values
are within 0.2% of the mean, 80% are within 0.4%, and 98%
are within 0.6%. Since the change in transmission for differ-
ent abutment positions is not symmetric about the central
axis, the variation seen is likely most dependent on MLC leaf
position calibration accuracy.
C. Penumbra
Figure 6共a兲 shows a vertical dose profile 共in-plane兲
through each of the retracted leaves 关as shown in Fig. 3共a兲兴,
where the tongue and the groove of each opening generate an
independent penumbra. This graph demonstrates the con-
stancy of projected leaf width and penumbra for each leaf,
regardless of its off-axis distance in the in-plane direction.
That result is further shown in Fig. 6共b兲, where the in-plane
penumbra was tested for cross-plane dependence. Each line
represents a single leaf, and the data points on the line indi-
cate the cross-plane position of the penumbra measurement.
This graph shows very little variation of the in-plane penum-
bra caused by the edge of a leaf, either in the in-plane (p
⫽0.4) or cross-plane direction (p⫽0.3).
Similarly, the cross-plane penumbra caused by the end of
a leaf 关as shown in Fig. 3共b兲兴 is provided in Fig. 6共c兲. Each
line represents a particular leaf, and the data points on the
line indicate the cross-plane position of the penumbra mea-
surement. These results show less than 0.03 cm from their
mean value, regardless of the location of the field edge. Al-
though these figures only show two field edges 共Y2 and X2兲,
the results for all four edges are shown in Fig. 6共d兲. The
80%–20% penumbra for these 1-cm-wide fields was mea-
sured to have a weighted mean of 0.37 cm⫾0.01 cm.
The range in penumbra measured was between 0.31 and 0.43
mm for all cases. The mean cross-plane penumbra was
statistically different (p⬍0.001) for the X1 bank
(0.358⫾0.0007 cm) compared to the X2 bank (0.369
⫾0.0007 cm).
IV. DISCUSSION
The greatest significance of this work may be found in
establishing the dosimetric characteristics of the MLC in or-
der to allow the physicist to review the treatment planning
system’s accuracy in describing these effects. Nelson et al.21
of the RPC showed that all IMRT treatment planning sys-
tems they analyzed struggled to accurately model the dose in
organs at risk adjacent to planning target volumes, regions
presumably shielded by the MLC leaves during the majority

TABLE III. Intraleaf leakage variation in the in-plane direction. Each column shows the average and standard deviation of leakage in the in-plane direction for
a specified cross-plane location.

Cross-plane distance from central axis 共cm兲

9.5 7.5 5.5 3.5 1.5 ⫺0.5 ⫺2.5 ⫺4.5 ⫺6.5 ⫺8.5 ⫺10.5

Average 共%兲 0.84 0.85 0.85 0.84 0.83 0.82 0.80 0.78 0.75 0.72 0.70
Standard dev 共%兲 0.07 0.07 0.08 0.08 0.08 0.08 0.08 0.08 0.09 0.09 0.08

Medical Physics, Vol. 30, No. 9, September 2003

2550 J. E. Bayouth and S. M. Morrill: MLC dosimetric characteristics 2550

TABLE IV. Intraleaf leakage variation in the cross plane direction. Each column shows the average and standard deviation of leakage in the cross-plane
direction through a specific leaf.

X2 bank leaf

2 3 4 5 6 7 8 9 10

Average 共%兲 0.53 0.69 0.72 0.76 0.79 0.81 0.80 0.82 0.85
Standard dev 共%兲 0.07 0.04 0.05 0.06 0.06 0.05 0.05 0.05 0.05

X2 bank leaf

11 12 13 14 15 16 17 18 19

Average 共%兲 0.85 0.85 0.86 0.90 0.84 0.86 0.85 0.86 0.86
Standard dev 共%兲 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05

X2 bank leaf

20 21 22 23 24 25 26 27 28

Average 共%兲 0.85 0.82 0.82 0.82 0.80 0.78 0.74 0.74 0.65
Standard dev 共%兲 0.05 0.05 0.05 0.05 0.06 0.06 0.06 0.06 0.05

of IMRT treatments. Furthermore, these regions are fre-
quently within a high dose-gradient region where accurate
modeling of the dose profile from 5–10 mm from the field
edge may be critical.
FIG. 5. Leaf-end transmission 共a兲 through abutted leaf pairs, for all 27 1 cm
leaf pairs. Each line represents 1 of 9 different abutment locations. 共b兲 fre-
quency distribution of leaf-end transmission.
Interleaf leakage, intraleaf leakage, and leaf-end penum-
bra have been previously described for fields of 10⫻10 cm
and larger for the MLC collimators found on linear accelera-
tors from Elekta 共Huq et al.9兲, Siemens 共Das et al.14兲, and
Varian 共Galvin et al.22兲. These measurements tend to be
through the central axis of the beam and for larger field sizes.
Das et al.14 showed the average intraleaf leakage to be 0.92
⫾0.12% and the average interleaf leakage to be 1.17
⫾0.13% for the same MLC design studied here. Our results
are consistent with these, in both the in-plane and cross-
plane direction, regardless of the measurement position with
respect to the central axis. Also for comparison, the transmis-
sion through leaves reported for MLCs on General Electric,
Philips, and Varian linear accelerators were in the order of
1.2%, 5%, and 4%, respectively for a 6 MV beam.9,14,22
The leaf-end transmission that we have measured has not
been reported previously in the literature for the double-
focused MLC design provided by Siemens. Klein et al.23 re-
ported this value for a single-focused MLC with rounded leaf
edges as large as 28% for 18 MV photons on the central axis,
reducing to 12% for the furthest points measured off-axis.
The values presented here for the double-focused MLC were
substantially more consistent, as would be expected. The
magnitude of the leaf-end transmission will be greatly de-
pendent on the leaf position calibration accuracy of the
MLC. At our institution, we utilize a leaf position calibration
procedure24 that produces 95.5% of the leaf-positions to be
within 1 mm and a mean within 0.4 mm.
The double-focused MLC also demonstrated consistency
in penumbra. Das et al.14 studied the change in penumbra for
a 10⫻10 cm field at several off-axis positions of the field
center. Specifically, they considered the penumbra along the
center leaf, traveling from 5 cm over-travel to 10 cm open,
and reported ⭐0.5% variation in the penumbra. Our results
show the penumbra is consistent both for the leaf-end as well
as across the side-edge of the leaves. The magnitude of the
penumbra measured for fields 1 cm wide 共0.37 cm penum-
bra兲 was significantly less than the values reported by Das

Medical Physics, Vol. 30, No. 9, September 2003

2551 J. E. Bayouth and S. M. Morrill: MLC dosimetric characteristics
FIG. 6. In-plane and cross-plane penumbra created by the side and end of the leaf, respectively. 共a兲 An in-plane dose profile through each of 7 different
retracted leaves. 共b兲 In-plane penumbra measured at 11 different cross-plane locations. Each line represents 1 of 7 retracted leaves. 共c兲 Cross-plane penumbra
measured at 11 different cross-plane locations. Each line represents 1 of 27 leaves on the X2 leaf bank, each having a different in-plane position. 共d兲 Combined
penumbra results for all four field edges measured for all locations studied.
et al.14 for square field sizes of 5 to 35 cm per side 共0.75–1.0
a兲
cm penumbra兲, as seen by the Gaussian shape of the dose
profile for these narrow fields. This indicates the change in
penumbra is driven by additional scatter for the larger field
sizes and not due to the position of the leaves or jaw.
V. CONCLUSIONS
Dosimetric characteristics of small and IMRT fields have
been presented for a double-focus MLC. The data can be
used to validate treatment planning algorithms for these
fields. AAPM Task Group 507 reports that MLC designs us-
ing nonfocused leaf ends produce a variation of transmission
through abutted leaf pairs 共as a function of leaf position兲 as
great as a factor of 2, and changes in the penumbra up to 3
mm as a function of leaf position. This work shows the pen-
umbra and leakage of the double-focus MLC to be remark-
ably consistent for the range of leaf positions studied. Al-
though some statistical differences of penumbra and leakage
were identified, these differences are not dosimetrically sig-
nificant for our current applications. These dosimetric char-
acteristics are well designed for small field IMRT segments
like those often seen for segmental MLC6 IMRT field seg-
ment shapes, where opposing leaf pairs are often separated
by 10 mm or less.
Medical Physics, Vol. 30, No. 9, September 2003
2551
Electronic mail: john.bayouth@utmb.edu
1
E. Podgorsak, G. Pike, A. Olivier, M. Pla, and L. Souhami, ‘‘Radiosur-
gery with high energy photon beams: A comparison among techniques,’’
Int. J. Radiat. Oncol., Biol., Phys. 16, 857– 865 共1989兲.
2
A. Eisbrauch, L. H. Marsh, M. K. Martel, J. A. Ship, R. K. Ten Haken, A.
T. Pu, B. A. Frass, and A. S. Lichter, ‘‘Comprehensive irradiation of head
and neck cancer using conformal multisegment fields: Assessment of
target coverage and noninvolved tissue sparing,’’ Int. J. Radiat. Oncol.,
Biol., Phys. 41, 559–568 共1998兲.
3
B. A. Fraass, M. L. Kessler, D. L. McShan, L. H. Marsh, B. A. Watson,
W. J. Dusseau, A. Eisbruch, H. M. Sandler, and A. S. Lichter, ‘‘Optimi-
zation and clinical use of multisegment intensity-modulated radiation
therapy for high-dose conformal therapy,’’ Semin. Radiat. Oncol. 9,
60–77 共1999兲.
4
Y. Xiao, J. Galvin, M. Hossain, and R. Valicenti, ‘‘An optimized forward
planning technique for intensity modulated radiation therapy,’’ Med.
Phys. 27, 2093–2099 共2000兲.
5
L. Verhey, ‘‘Comparison of three-dimensional conformal radiation
therapy and intensity-modulated radiation therapy systems,’’ Semin Ra-
diat. Oncol. 9, 78 –99 共1999兲.
6
Intensity Modulated Radiation Therapy Collaborative Working Group,
‘‘Intensity-modulated radiotherapy: Current status and issues of interest,’’
Int. J. Radiat. Oncol., Biol., Phys. 51, 880–914 共2001兲.
7
A. Boyer, P. Biggs, J. Galvin, E. Klein, T. LoSasso, D. Low, K. Mah, and
C. Yu, ‘‘Basic applications of multileaf collimators. Report of the AAPM
Radiation Therapy Committee Task Group No. 50,’’ 共Medical Physics,
2001兲.
8
M. R. Sontag and T. H. Steinberg, ‘‘Performance and beam characteris-
tics of the Siemens Primus linear accelerator,’’ Med. Phys. 26, 734 –736
共1999兲.
9
M. S. Huq, Y. Yu, Z. P. Chen, and N. Suntharalingam, ‘‘Dosimetric char-
2552 J. E. Bayouth and S. M. Morrill: MLC dosimetric characteristics
acteristics of a commercial multileaf collimator,’’ Med. Phys. 22, 241–
247 共1995兲.
10
J. R. Palta, D. K. Yeung, and V. Frouhar, ‘‘Dosimetric considerations for
a multileaf collimator system,’’ Med. Phys. 23, 1219–1224 共1996兲.
11
S. L. Meeks, F. J. Bova, S. Kim, W. A. Tome, J. M. Buatti, and W. A.
Friedman, ‘‘Dosimetric characteristics of a double-focused miniature
multileaf collimator,’’ Med. Phys. 26, 729–733 共1999兲.
12
M. Stasi, B. Baiotto, F. Palamara, P. Gabriele, and G. Scielzo, ‘‘Dosim-
etric characterization of a multileaf collimator,’’ Radiol. Med. 共Torino兲
97, 382–388 共1999兲.
13
M. Pasquino, V. Casanova Borca, and S. Tofani, ‘‘Physical-dosimetric
characterization of a multileaf collimator system for clinical implemen-
tation in conformational radiotherapy,’’ Radiol. Med. 共Torino兲 101, 187–
192 共2001兲.
14
I. J. Das, G. E. Desobry, S. W. McNeeley, E. C. Cheng, and T. E. Schulth-
eiss, ‘‘Beam characteristics of a retrofitted double-focused multileaf col-
limator,’’ Med. Phys. 25, 1676 –1684 共1998兲.
15
R. Mitra, J. Bayouth, and I. Das, ‘‘Startup characteristics and dosimetry
of small monitor unit segments in step and shoot IMRT delivery for two
types of digital linear accelerators,’’ Med. Phys. 28, 1202 共2001兲.
16
X. R. Zhu, P. A. Jursinic, D. F. Grimm, F. Lopez, J. J. Rownd, and M. T.
Gillin, ‘‘Evaluation of Kodak EDR2 film for dose verification of intensity
Medical Physics, Vol. 30, No. 9, September 2003
2552
modulated radiation therapy delivered by a static multileaf collimator,’’
Med. Phys. 29, 1687–1692 共2002兲.
17
I. Chetty and P. Charland, ‘‘Dosimetric characterization of Kodak ex-
tended dose range film in megavoltage photon beams,’’ Med. Phys. 29,
1207 共2002兲.
18
M. Bieda, A. Sarkar, and I. Das, ‘‘Accuracy of Kodak’s EDR2 film in
measuring clinical dosimetry data,’’ Med. Phys. 29, 1264 共2002兲.
19
N. Childress, L. Dong, and I. Rosen, ‘‘Rapid radiographic film calibration
for IMRT verification using automated MLC Fields,’’ Med. Phys. 29,
1366 共2002兲.
20
J. Bayouth, S. Zhou, and S. Morrill, ‘‘Analysis of variation in H&D curve
creation for EDR2 film dosimetry,’’ Med. Phys. 30, 1450 共2003兲.
21
A. Nelson, P. Balter, W. Hanson, and G. Ibbott, ‘‘Experience with an
IMRT head and neck QA phantom,’’ Med. Phys. 29, 1366 共2002兲.
22
J. M. Galvin, A. R. Smith, and B. Lally, ‘‘Characterization of a multi-leaf
collimator system,’’ Int. J. Radiat. Oncol., Biol., Phys. 25, 181–192
共1993兲.
23
E. E. Klein, W. B. Harms, D. A. Low, V. Willcut, and J. A. Purdy, ‘‘Clini-
cal implementation of a commercial multileaf collimator: Dosimetry, net-
working, simulation, and quality assurance,’’ Int. J. Radiat. Oncol., Biol.,
Phys. 33, 1195–1208 共1995兲.
24
J. E. Bayouth, D. Wendt, and S. M. Morrill, ‘‘MLC quality assurance
techniques for IMRT applications,’’ Med. Phys. 30, 743–750 共2003兲.