Physica Medica: European Journal of Medical Physics - Volume 32 - Supplement 1 - February 2016

15 / 146

probe with integrated MOSkin dosimeters was recently developed. The
system allows to achieve both imaging and real time in vivo dosimetry (IVD)
in HDR prostate brachytherapy (BT). The aim of our study was to set up
and calibrate the instrument to perform rectal wall IVD, to quantify dis-
crepancies between calculated and delivered doses, and to minimize the
achieved uncertainties.
Materials and methods: The MOSkin is a specific type of MOSFET dosim-
eter optimized to measure dose in steep dose gradients (sensitive volume:
4.8 × 10−6 mm 3). During treatment, TRUS probe integrated with MOSkins
is left inside the rectum and the delivered doses (Dmeas) are measured. In
this study, 22 treatment sessions were analysed and Dmeas were com-
pared to the doses calculated on the pre-treatment plans (Dpre) and
reconstructed on the images acquired right after the treatment delivery
(Dpost). Moreover, to minimize discrepancies between Dmeas and Dpost, a cor-
rection function (CF) linearly depending on both angles and distances
between MOSkins and each planned source dwell position was empirical-
ly established.
Results: Averages of absolute differences between Dmeas and Dpre, Dmeas and
Dpost, Dpre and Dpost were 6.7 ± 5.1%, 3.6 ± 1.9% and 6.3 ± 4.7%, respectively.
Agreement between Dmeas and Dpost was significantly better than between
Dmeas and Dpre (p = 0.002) and between D pre and Dpost (p = 0.004). Discrep-
ancies between Dpre and Dpost correlated with the time required for treatment
planning. The definition of a linear CF allowed a further relative uncer-
tainty reduction, leading to an agreement between Dmeas and Dpost of
2.2% ± 1.1%.
Conclusions: MOSkin dosimeters integrated to the TRUS probe proved to
be an accurate instrument for rectal wall IVD in HDR prostate BT. Mea-
surement uncertainties were minimized; next step will be to study the
system for real time IVD and to define possible action protocols.
A.32
COMPARISON OF FOUR DIODES FOR BASIC DOSIMETRY: OUTPUT
FACTORS, PERCENTAGE DEPTH DOSE CURVES AND PROFILES FOR FIELD
SIZES FROM 0.8 TO 10.4 CM
M. Casati M. Zani C. Arilli A. Compagnucci L. Marrazzo
S. Scocciant D. Gret S. Pallotta C. Talamonti a Fisica Medica, Azienda
Ospedaliero Universitaria Careggi, Firenze, Italy; b Dip. di Scienze Biomediche
Cliniche e Sperimentali Mario Serio, Università degli studi di Firenze, Firenze,
Italy; c Radioterapia, Azienda Ospedaliero Universitaria Careggi, Firenze, Italy
Introduction: The purpose of this work is to compare four diodes for rel-
ative dose measurements of output factors (OF), percentage depth dose
curves (PDD) and profiles and to assess their suitability for square field size
(fs) from 0.8 cm to 10.4 cm.
Materials and methods: Iba 1D array (diodes 0.6 mm × 1 mm × 0.02 mm)
and three single diodes, Iba Razor (0.6 mm diameter × 0.02 mm), SNC EDGE
(0.8mm diameter × 0.03mm) and Exradin D1V (1.1mm diameter × 0.05mm)
were compared. All data were acquired in RFA water phantom for a
SynergyBM 6MVRX beam.
For fs ≥ 4 cm, Exradin A26 ionization chamber (active vol. 0.015 cc) was
chosen as reference detector, while for fs < 4 cm, Exradin W1 scintillator
(1 mm diameter) was chosen for its radiological water equivalence.
Results: OF were normalized to 3.2 cm fs. For larger fs, Razor and D1V show
an over-response vs A26 of up to 3.9% and 2.4% respectively (10.4 cm fs),
while EDGE and 1D agree with A26 within 0.3% for fs from 4 cm to 10.4 cm.
As fs decreases down to 0.8 cm, with respect to W1, EDGE detector shows
an over-response of up to 1.8%, Iba 1D is in agreement within 1% except
for 0.8 cm fs (positioning error), Razor and D1V exhibit negative devia-
tions within −1.4% and −0.9%, respectively.
PDD CURVES (depths ≤ 20 cm). For Razor and D1V, the over-response to low-
energy scattered radiation with increasing depth is evident: within 1.5%
for 0.8 cm and up to 4% for 10.4 cm. EDGE has deviations within 1.6% for
all fs. For Iba 1D, even for fs 10.4 cm, PDD is in agreement vs A26 within
1%.
PROFILES acquired for all diodes with the active area perpendicular to beam
axis agree (<1 mm distance to agreement) with W1 for 0.8 cm fs.
As fs increases, D1V and Razor confirm their over-response to low-energy
scattered radiation on the tails of the profiles.
Conclusions: Iba 1D has a good response for profiles and OF (fs > 0.8 cm),
but it is an array, difficult to align with beam axis for small fs. Razor and
D1V are suitable only for small fs, while EDGE is suitable for fs > 1.6 cm.
A.33
LINAC COUCH TOP ATTENUATION IN ROTATIONAL RADIOTHERAPY
TREATMENTS: CLINICAL CONSEQUENCES
M. Casati C. Arilli L. Marrazzo C. Talamonti A. Compagnucci
S. Pallotta S. Scoccianti D. Greto a Fisica Medica, Azienda Ospedaliero
Universitaria Careggi, Firenze, Italy; b Dipartimento di Scienze Biomediche,
Sperimentali e Cliniche, Università di Firenze, Firenze, Italy; c Radioterapia,
Azienda Ospedaliero Universitaria Careggi, Firenze, Italy
Introduction: With the introduction of rotational radiotherapy tech-
niques, LINAC couch tops have been modified to minimize their attenuation.
Nonetheless couch tops should be taken into account into planning and
not all TPS have a specific tool for couch modeling. The aim of this work
is to assess the effects of couch attenuation, for VMAT plans on different
sites.
Materials and methods: Our LINAC is equipped with a couch iBEAM evo
with an external carbon structure filled with a foam core. The nominal at-
tenuation at 6MVRX is 2.4% for the body extension (BE) and 1.5% for the
head extension (HE). Our TPS only models the BE, so we acquired a CT scan
of the HE to create a geometric model. We assign relative electron densi-
ties: 1 for carbon and 0.01 for foam core.
For this study, some VMAT plans were optimized and then recalculated (i.e.
without re-optimization), introducing a different couch modeling. Plans were
compared.
A) SNC and body plans, intentionally optimized without considering the
couch, as if the TPS was not capable to keep the couch attenuation into
account, were recalculated after introducing the HE or BE.
B) Some clinical plans (body) optimized with BE were recalculated with a
lateral 4 cm shift of the BE to evaluate the effect of different positions of
the BE during treatment.
C) Some SNC plans were optimized considering the BE, as if the only model
in the TPS was BE, and then recalculated with the HE to evaluate possible
OARs overdosage due to couch attenuation underestimation.
Results: (A) Body plans recalculated introducing BE a posteriori show a re-
duction of PTV coverage of about 2.5%–5%. For SNC plans the reduction is
<0.5% for HE, <1.5% for BE. (B) Plan deterioration seems to be negligible in
most cases. (C) The clinical relevance of the differences is case dependent.
Conclusions: A proper modeling of the LINAC couch is important, espe-
cially in the case of BE. If the TPS does not have a specific tool we suggest
creating a geometric model in the TPS contour atlas.
A.34
CHARACTERIZATION OF THE ENERGY DEPENDENT RESPONSE OF EBT3
RADIOCHROMIC FILM TO PROTON AND CARBON ION BEAMS
R. Castriconi P. Russo G. Mettivie F. Di Lill A. Sarn M. Ciocca
A. Mirandola C. Sini S. Broggi a Dipartimento di Fisica and INFN Sezione
di Napoli, Università di Napoli Federico II, Napoli, Italy; b Centro Nazionale di
Adroterapia Oncologica, Pavia, Italy; c Ospedale San Raffaele di Milano, Milano,
Italy
Introduction:The authors characterized the response of EBT3 radiochromic
films to proton and carbon ion beams. EBT3 films are adopted routinely
for dosimetry and for QA at linear accelerators used for conventional ra-
diotherapy, and for QA in hadrontherapy. In contrast to sparsely ionizing
radiations, irradiation by ions leads to a strongly inhomogeneous dose de-
position. A quenching effect occurs in the film response consisting of an
under-response of the EBT3 to ion beams with respect to MV-photon beams,
likely due to the ion track structure. The authors investigated the re-
sponse of EBT3 films to proton and carbon ion beams, and compared it with
photon and electron beams’ response, for assessment of a dose–response
correction procedure.
Method and materials: The EBT3 films were calibrated in the dose range
of 0.4–20 Gy with proton (63, 150 and 230 MeV) and carbon ion (115 MeV/u
e10 Abstracts/Physica Medica 32 (2016) e1–e70