804 / 1096
S788
ESTRO 36
_______________________________________________________________________________________________
dose by the accelerator directly and evaluates the dose
delivery error to the plan. We have used True-Beam
(Varian) with 10 MV photon beams and a QA plan which
underwent prostate VMAT. For the QA plan, we prepared
seven modified MLC files in which MLC positions were
manually shifted between 0 mm to 3 mm, respectively.
Then, the dose delivery errors were measured and
analyzed dose deviation, distance-to-agreement (DTA)
and gamma index by the Delta4 Discover system as well as
to the Delta
4
3D dosimetry system. All measurements were
compared against the points that received less than
appropriate percentage dose (<10%) were excluded from
the gamma index calculation.
Results
The results of the Delta4 Discover system and the Delta
4
3D dosimetry system using all available points for the
gamma calculation, 95.7±5.9% and 97.8±3.5% of them
passed the criteria (3%/2mmDTA/Th10%), respectively.
The two systems also had high correlations of dose
deviation and DTA, permitting the routine verification of
VMAT patient-specific QA plan as well as permanent in-
vivo dosimetry during the patient’s treatment course.
Conclusion
In this study, we have found that there was high
correlation between the pass rate and the intentional dose
delivery error, as respect to dose deviation, DTA, and
gamma index of the Delta
4
Discover system and the Delta
4
3D dosimetry system. It was suggested that the dosimetric
verification system under investigation could be useful for
routine patient specific QA.
EP-1475 RBE estimation of different Brachytherapy
sources based on micro- and nanodosimetry
M. Bug
1
, T. Schneider
2
1
Phys. Techn. Bundesanstalt PTB, 6.5 Radiation Effects,
Braunschweig, Germany
2
Phys. Techn. Bundesanstalt PTB, 6.3 Radiation
Protection Dosimetry, Braunschweig, Germany
Purpose or Objective
Depth-dependent RBE values of typical photon-emitting
Brachytherapy (BT)-sources were determined by a
microdosimetric and a nanodosimetric approach. The
microdosimetric approach considers a biological endpoint
while the nanodosimetric approach is entirely based on
the track structure, given by the interactions of the
photons and secondary electrons. The track structure
characterizes the radiation quality on the nanometric
scale.
Material and Methods
Within a cylindrical water phantom, isotropically emitting
BT-sources were positioned 4 cm below the surface.
Studied were Co-60 and Ir-192 representing high-energy
photon-emitting sources, I-125 being a low-energy photon-
emitting source, and Intrabeam
®
- and Axxent
®
-devices as
examples for electronic BT X-ray sources (EBX). Resulting
photon spectra were calculated at several points along the
cylindrical axis within cylindrical voxels of 0.5 mm depth
and 2 mm radius up to a depth of 10 cm.
The
microdosimetric
calculations of RBE are based on
yield coefficients α
dic,
representing the linear component
of the dose-effect relationship for the dicentric
chromosome aberration yield after an irradiation with
monoenergetic photons. The RBE for a given source in a
given point was determined by convoluting the respective
spectrum with the function α
dic
(E), obtained previously by
microdosimetric calculations.
The same depth-dependent photon spectra were used to
determine
nanodosimetric
quantities by Geant4-DNA
calculations. For each initial photon track, target volumes
in size of one DNA convolution which experienced at least
4 ionizations (F4) were identified. Such quantities were
previously shown to describe the DSB yield. Based on the
average minimum distance between these volumes within
each photon track, the RBE was estimated by
normalization to the distance for Co-60 at 0.125 mm
depth.
Results
Relative depth-dependent RBE based on nanodosimetric
quantities are similar to the microdosimetric RBE. For Co-
60 and Ir-192, the RBE increases with depth due to an
increasing contribution of low-energy photons in the
spectra. For the denser ionizing sources, nanodosimetric
RBE values were divided by 1.9. Apart from this factor, the
constant RBE-dependence up to 10 cm for I-125 and the
decrease of RBE for the two EBX sources due to beam
hardening are in good agreement with the
microdosimetric RBE.
Conclusion
RBE based on track structure (nanodoismetric approach)
shows that the average intra-track distance between DNA-
modelling volumes potentially suffering severe damage is
well related to the microdosimetric RBE, based on the
formation of dicentric chromosomes, for several BT-
sources. Apart from a constant normalization factor for
the denser ionizing sources, the depth-dependence is in
excellent agreement. This indicates that the
nanodosimetric photon track characterization performed
in this study is a good descriptor for the radiation quality.
Furthermore, the proposed target volume appears
realistic. Note, that neither the photon fluence nor
biological endpoints were taken into account for this
approach.
EP-1476 Preliminary results of in-vivo dosimetry by
EPID
S. Giancaterino
1
, M. Falco
2
, A. De Nicola
2
, N. Adorante
2
,
M. Di Tommaso
2
, M. Trignani
2
, A. Allajbej
2
, F. Perrotti
2
,
D. Genovesi
2
, F. Greco
3
, M. Grusio
3
, A. Piermattei
3
1
Ospedale Clinicizzato S.S. Annunziata, Radioterapia,
Chieti, Italy
2
University of Chieti SS. Annunziata Hospital,
Department of Radiation Oncology “G. D’Annunzio”-,
Chieti, Italy
3
Università Cattolica del Sacro Cuore, Medical Physics
Institute - Fondazione Policlinico Universitario A.
Gemelli-, Rome, Italy
Purpose or Objective
This study reports in-vivo dose verification (IVD) results
elaborated with SOFTDISO software on 300 cancer patients
treated with 3D-CRT, IMRT and VMAT techniques.
SOFTDISO uses the integral EPID image referred to each
single static or dynamic beam providing a quasi- real-time
test elaboration.
Material and Methods
The selected patients for this study were treated with an
Elekta Synergy Agility LINAC at SS. Annunziata Hospital.
3D-CRT, IMRT and VMAT treatment plans of 300 patients
were randomly selected. IVD tests were processed
with the SOFTDISO software who provides two type of
tests: (i) R ratio between the reconstructed isocenter dose
and the planned one; (ii) transit dosimetry based on γ-
analysis of EPID imaging (P
g
(%) and g
mean
).
Results
We identified class-1 errors, derived from inadequate QCs,
and class-2 errors due to patient morphological changes.
Considering overall (6697) tests, we found out that only
5% of them showed out-of-tolerance mean R values. For
gamma index analysis, in 13% of the overall tests were
found to be out of tolerance. Ignoring class-2 errors, 100%
of patients treated with different radiotherapy techniques
(except 3DCRT breast treatment, for which no class-2
errors were observed) reported mean P
g
(%) values within
tolerance levels. Thus, the percentage of out- of-
tolerance tests decreases from 13% to 7%. However,