798 / 1096
S782
ESTRO 36
_______________________________________________________________________________________________
scanning techniques. The aim of this work is to investigate
the influence of variations of the nominal beam width on
the dose distribution of cubic dose volumes, which are
often part of a typical QA program.
Material and Methods
For QA purposes, three cubic dose volumes with a spread
out bragg peak (SOBP) of 3x3x3 cm³ are optimized with
the treatment planning system (TPS), syngo PT Planning
(Siemens, Germany). The nominal dose in the SOBP is 0.5
Gy. The depth in water of the centre of the cubes is 50,
125 and 200 mm, respectively.
To perform dose calculations on a water phantom with
variation of initial beam width, the needed algorithms and
base data of our TPS are transfered to MATLAB (The
Mathworks Inc., USA). These are mainly the depth dose
distributions and the double gaussian parametrization of
the beam width. Plans are recalculated with MATLAB with
nominal and varied beam width. Dose distributions are
analysed performing a 3D gamma index analysis with
criterias of 1mm distance to agreement and 5% dose
deviation, normalized to global maximum. The maximum
negative and positive tolerable beam width deviations are
determined where all points still pass the gamma index
acceptance criteria (e.g. gamma index < 1).
Results
The maximum tolerable beam width variations for the
dose cube in a depth of 50 mm amounts to -7% and +10%,
while for the dose cube in 125 mm depth the found values
are -12% and +17%. The most interesting result is found for
the dose cube in 200 mm depth. While the high dose region
shows comparable large possible beam width variations of
-17% to +25%, the maximum tolerable beam width
deviation in the entrance region amounts to -8% and +15%.
Conclusion
The QA plan in small depth shows rather small tolerable
deviations of the initial beam width. It is observed, that
this is mainly affected by the strong variation of the
particle numbers per scan spot in each energy slice,
optimized by the TPS to achieve lateral penumbras as
small as possible. Following this observation, we created
a plan with the same field size parameters consisting
uniform scanned energy slices. Applying the described
beam width variation, resulting tolerable beam width
deviations
are
-13%
and +12% .
For the QA plans in medium and large depth, beam width
tolerances in the SOBP become larger. One reason is that
the additional multiple coulomb scattering in water
dilutes the impact of deviations of the initial beam width.
Depending on the spot scan pattern, the entrance region
can be the region with the highest demands on beam width
accuracy for plans in large depths.
For TPS commissioning, the spot scan pattern should be
inspected with regard to beam spot width variations.
EP-1465 A beam matching procedure for Volumetric
Modulated Arc Therapy
L. Abdullah
1
, C. Constantinescu
1
, M.N. Hussein
1
1
King Faisal Specialist Hospital and Research center,
Biomedical Physics, Jeddah, Saudi Arabia
Purpose or Objective
To describe our experience of commissioning and quality
control tests, along with the adjustments performed for
beam matching in VMAT.
Material and Methods
A Synergy linac upgraded to Agility head was matched to
a Versa HD linac.
Dose calculation for commissioning tests was performed
on Monaco TPS version 5.0, with Monte Carlo algorithm,
inhomogeneity correction, calculation grid size of 1mm,
and statistical uncertainty of 0.5%/control point. The
compliance between dose calculation and measurement
was assessed for a gamma-index (GI) of 3% dose difference
and
3
mm
distance-to-agreement
(DTA).
The manufacturer specifies a routine for acceptance
testing of matched linear accelerators which concerns
only the beam quality and field size.
After appropriate MLC calibration, percentage depth dose
and beam quality index were evaluated for both linacs, in
similar setups. In-plane and cross-plane beam profiles
were acquired for field sizes of 10x10cm², 20x20cm
2
, and
depth of 10 cm. All dosimetric parameters appeared
identical for both linacs, within a tolerance of 1%. Beam
output calibration differed within 0.5%. The quality of
matching was found to be valid for 3D-CRT treatments but
not for VMAT. Specific test beams were further performed
to compare the two linacs for the accuracy of leave and
jaws movement, using the “picket fence” and “sliding-
window” methodology at different gantry and collimator
angles, using electronic portal imaging device (EPID)
dosimetry. The MLC minor leaf and jaw offsets of the
second linac was finely adjusted mechanically in order to
achieve acceptable accuracy of matching.
38 clinical VMAT plans for various sites and different
degrees of modulation were verified on both linacs. The
correlation between the GI passing rate and plan
modulation, assessed by the number of MU and number of
segments, was further investigated using a logistic
regression test.
Results
The beam quality index was 0.682 for linac 1 and 0.686 for
linac 2 after the vendor's matching procedure was
performed. GI analysis of the “picket fence” and “sliding
window” test beams indicated the need for mechanical
adjustment of MLC minor leaf and jaw offsets of linac 2.
After the beam matching accomplished, GI analysis of
VMAT clinical treatment plans indicated good agreement
between the two linacs. The average passing rate between
calculated and delivered dose distribution was 97.7±2.8%
(range: 87.5%-100%) for linac 1 and 93.4±4.7% (range: 83%-
99.%)
for
linac
2.
Weak correlations were found between the GI passing rate
and number of MU (R
2
=0.043, p=0.146) and segments
(R
2
=0.0273, p=0.240), indicating that the degree of
modulation is not to be considered in setting acceptance
criteria for GI analysis.