985 / 1096
S969
ESTRO 36
_______________________________________________________________________________________________
Material and Methods
The leaf position is defined as the position of the 50% of
the dose profile. This measurement depends on relative
position of the beam source and of the leaves. In order to
validate the EPID measurements of the absolute leaf
positions, 10 dose profiles, at the center of different leaf
pairs of the same MLC field, were acquired with an Elekta
iViewGT EPID and with a diode positioned in a water
phantom. The comparison between the two detectors was
performed by Matlab. Garden Fence (GF) was chosen as
test of the leaf position accuracy and a preliminary study
on the gap width was conducted. Leaf position accuracy
was checked automatically with DC by acquiring GF at the
4 cardinal gantry angles and with all the beam energies (6,
10 and 15MV), while the reproducibility was tested with 5
GF repeated in one day and 6 repeated in a time interval
of 70 days.
Results
The difference between EPID and diode absolute
measurement of the leaf positions was less than 0.8mm
for all the analyzed leaves, resulting from the summation
of an error due to the isocenter identification (0,5mm)
plus the leaf positioning error (0.2mm). The gap width
study revealed that, because of the penumbra widening
observed in small fields, the leaf position could be
accurately measured as the 50% of the edge profile, only
if the gap width is equal or larger than 16mm with 6MV
beam. Therefore, GF with 20mm gap was chosen as leaf
position accuracy test for all the energies in order to
distinguish the effect of beam source from that of leaf
positioning. For the GF at different gantry angles the
difference between the measured and the prescribed
position was well within 1.0mm for all the leaves.
Moreover, reproducibility of each leaf position resulted to
differ from its average value less than 0.4mm.
Conclusion
This work permitted to assess the accuracy and the
repeatability of the Elekta Agility MLC leaf positioning by
the combined use of the Elekta IviewGT EPID and the
Dosimetry Check software through the acquisition and the
analysis of Garden Fence test. This system was validated
comparing the EPID with a diode in a water phantom and
assessing the minimum gap width necessary for an
accurate leaf position measurement at all energies which
is useful to distinguish issues related to beam symmetry
from those related to leaf positioning.
EP-1760 A simple method for estimating the
longitudinal isocentre shift due to gantry motion
R. Hudej
1
, D. Brojan
1
, S. Pulko
2
, P. Peterlin
1
1
Institute of Oncology Ljubljana, Department of
Radiophysics, Ljubljana, Slovenia
2
University Clinical Centre Maribor, Department of
Oncology- Radiotherapy Unit, Maribor, Slovenia
Purpose or Objective
The isocentre as a point of intersection of the three
rotational axes (gantry, collimator and treatment couch)
ideally remains fixed in space during the rotation of
gantry, collimator, or the treatment couch. Due to the
mechanical limitations, gantry sags slightly, and
consequently the radiation isocentre shifts slightly
towards the treatment couch when the gantry rotates
from the uppermost to the lowermost position. The
purpose of this study is to assess this shift.
Material and Methods
A strip of radiochromic film embedded in a suitable water-
equivalent phantom is irradiated with a cross-line half-slit
field from the top (0°). Then the gantry is rotated to the
lowermost position (180°) without moving the jaws and
the phantom is irradiated again. The film is scanned and
analysed with an image analysis script. The central lines
of both half-slit images are determined, then the
intersection angle between them is calculated, and finally
the distance between the intersections of extrapolated
lines with the 'sagittal” plane is calculated.
Results
This method was tested on 7 linacs of different makes and
models (Elekta Synergy Platform, Elekta Versa HD, Varian
Unique, Novalis Tx) in the authors' radiotherapy centres.
The average distance by which the isocentre moves
between both gantry positions was found to be 1.04 mm
(SD 0.30 mm), with the whole range covered by the [0.53,
1.48] interval. The two lowest values were achieved on
the two single energy Varian Unique linacs. It was found
out that the longitudinal isocentre shift is largely
independent of the gantry isocentre wobble determined
by the star-shot test. We also tested the alignment of the
collimator 0° setting with the gantry rotation plane. The
average deviation was found to be 0.16° (SD 0.10°), range
[0.04°, 0.31°].
Conclusion
The results appear consistent, but it would be helpful to
test the method on a wider pool of treatment machines
over a longer period of time. The longitudinal isocentre
shift during gantry rotation is a non-negligible parameter
which needs to be incorporated into the uncertainty
budget which is the basis for the CTV-PTV margin.
EP-1761 Workflow development for the clinical
implementation of an MR-guided linear accelerator
T. Stanescu
1
, A. Berlin
2
, L. Dawson
2
, J. Abed
2
, A.
Simeonov
2
, T. Craig
2
, D. Letourneau
2
, D. Jaffray
2
1
Stanescu Teodor, Radiation Physics- PMH, Toronto,
Canada
2
Princess Margaret Cancer Centre, RMP, Toronto, Canada
Purpose or Objective
Development of clinical workflows for the implementation
of a new external beam radiation therapy environment
which relies on hybrid MR-CBCT in-room imaging guidance.
Material and Methods
A standard radiation therapy 6X linac (TrueBeam, Varian
Medical System, Palo Alto, CA) was integrated with a 1.5
T diagnostic MR scanner (IMRIS, Minnetonka, MN). The MR
can move on rails and was tuned up to perform optimal
imaging inside the treatment room in the proximity of the
linac. The patient load is transferred directly between the
MR diagnostic table and the linac IGRT couch via a
hovercraft system (Zephyr XL, Diacor, Salt Lake City, UT).
No special MR safety requirements were employed
regarding the curation of the linac room – the linac/couch
can be freely operated mechanically when the MR magnet
is present - only typical MR room screening for
ferromagnetic content was implemented. Comprehensive
testing was completed to confirm negligible magnetic
field coupling between the MR and the TrueBeam system
(linac and patient table). Since the linac retains its default
features and an MR imager is available in the linac vault a
combined MR-kV approach can be employed for the
patient setup verification and treatment delivery. A new
software tool was developed in collaboration with Varian
to provide the computation and implementation of
treatment couch shifts based on soft-tissue information,
i.e. image matching between plan MR and guidance MR. In
this study, clinical workflows for liver and prostate were
developed and tested. Each site posed challenges from
patient image data planning and acquisition to RT planning
and in-room guidance. The approach was to integrate the
capabilities offered by the new technology in existing
processes.
Results
MR imaging protocols for planning and guidance were
established. The guidance scans were optimized to
minimize session time with negligible penalty on the
accuracy of the image matching process (planned vs. on