ESTRO 36 Abstract Book

229 / 1082

S216 ESTRO 36 2017
_______________________________________________________________________________________________
instrumental to allowing introduction of the technique to
our home institutes.
The ability to talk to our hosts and ask questions in a face
to face setting has proved vital in helping us to get a good
grasp of all the techniques required. This also allowed us
to explore potential for collaborations in the future and
for sharing useful reagents such as tumour cell lines
between our groups to facilitate our research.
SP-0414 Experience with the ESTRO mobility grant;
proton irradiation of a 3D dosimeter
E.M. Høye 1, P.S. Skyt1, P. Balling2, J. Swakon3, J.B.B.
Petersen1, M. Rydygier3, G. Mierzwińska3, L.P. Muren1
1 Aarhus University Hospital, Department of Medical
Physics, Aarhus, Denmark
2 Aarhus University, Department of Physics and
Astronomy, Aarhus, Denmark
3 Polish Academy of Sciences, Institute of Nuclear Physics,
Krakow, Poland
In my visit to the Cyclotron Center Bronowice in Kraków I
investigated the performance of a new 3D dosimeter for
proton therapy. The aim of the visit was to study the
known quenching effects in the Bragg peak of proton
beams in our dosimeter. Denmark is currently building its
first proton therapy center, and so a collaboration
agreement was made with the polish center in order to
perform irradiations of the dosimeter.
Dosimeter samples were prepared with different chemical
compositions, and brought to Kraków. A 1D optical laser
scanner was sent to Kraków, in order to allow read out of
the proton depth dose curve in the dosimeters few hours
after irradiation. Based on preliminary results from our
measurements, decisions were made as to which chemical
compositions to investigate further. New dosimeters were
produced for us in Aarhus, and sent to Kraków to be
irradiated in the second week.
The experience was logistically challenging, and many
people were contributing to the success of the project.
The study gave us detailed information as to how the
dosimeter can be further optimised for proton therapy. In
my presentation I will expand on the challenges we met
and on how they were dealt with. I am grateful to ESTRO
for allowing me this opportunity, together with all the
people in Aarhus and Kraków who helped with the
experiments.
Poster Viewing : Session 9: Dosimetry
PV-0415 Verification of pre-treatment DVH
measurements for individual plan QA
J. Stroom1, J. Boita1, M. Rodrigues2, C. Greco 1
1 Fundação Champalimaud, Radiotherapy, Lisboa,
Portugal
2 Mercurius Health, Radiotherapy, Lisbon, Portugal
Purpose or Objective
Radiotherapy plan QA by measurements is almost
mandatory for IMRT and VMAT. Generally, comparison of
planned and measured dose is performed using the
clinically not so relevant gamma analysis. Recently
software has become available to estimate DVHs based on
QA measurements. We have validated two such systems.
Material and Methods
Our new system, 3DVH (v3.3, SunNuclear), converts dose
deviations measured with the cylindrical ArcCheck
phantom to expected dose deviations in the patient,
hence enabling calculation of DVHs for targets and OARs.
Our existing system, PDAPP (NKI-AVL, Amsterdam), uses
back-projection of measured EPID dose images to produce
3D doses in patients or phantoms. These doses and dicom
structure files are subsequently read by in-house software
( pDVH) to calculate DVHs. We performed the following
tests:
1. 3DVH: With the new system, we first measured
30 different clinical plans (VMAT/IMRT) with
various energies (6MV – 10FFF) on different
linacs (Varian/Elekta) and evaluated the
measured 3D dose distributions using 3D gamma
(3%,3mm). We then compared with the clinical
2D ArcCheck analyses of the same
measurements.
2. 3DVH: We subsequently introduced MU errors or
systematic MLC errors (all leafs opened or
closed) in a subgroup of 6 Elekta plans before
measurement and studied the behaviour of
3DVH.
3. 3DVH+pDVH: To compare 3DVH with pDVH, we
made 3 conformal plans (AP, AP-PA, 4-field box)
and one 4-field IMRT plan on a slab-phantom
with PTV and cubic and cylindrical OARs (Fig).
The plans with and without errors were
measured with the slab-phantom for PDAPP , and
with ArcCheck for 3DVH. Mean PTV and OAR
doses were compared.
4. 3DVH+pDVH: For the phantom IMRT plan, we
predict the effect of an X mm MLC error on the
mean PTV dose to be X/<DMLC>, with <DMLC>
the average leaf pair distance in the plan. For
the cube DVHs of the conformal plans leaf
motions should shift the penumbra of the AP/PA
beams into or out of the 60mm cube and the
resulting DVH up and down by X/60, so X=3mm
would yield ΔVD50 ≈ 5% (Fig).
Results
1. Average 3D gamma passing rates of the 30
clinical cases were 97.7±3.5% (1SD), comparable
to the 2D rates of 97.0±2.1%. There was no
correlation between 2D and 3D results.
2. For the patient error tests, PTV DVHs with MU
errors correspond well to expectations. For
OARs and MLC errors, trends are as expected but
quantitative validation is more difficult (Table).
3. Slab phantom results show that generally both
systems accurately measure MU errors.
Differences between 3DVH and pDVH are larger