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

59 / 146

numerous anatomical districts. The aim of this investigation was to assess
the impact of the MRIdian radiation therapy system through a planning
analysis for rectal cancer treatments.
Material and methods: Ten sets of 3 plans (MRIdian step and shoot
IMRT, RapidArc and 5 beams sliding windows IMRT) were calculated for
10 patients affected by locally advanced rectal cancer. RapidArc (6–
15 MV) and 5 beams (6–15 MV) sliding window IMRT treatment plans
were calculated on Eclipse. The body, the bowel bag and the bladder
were the OaR considered. The prescribed doses for PTV2 were 45 Gy and
55 Gy for PTV1 through simultaneous integrated boost. The PTV V95 and
OaRs QUANTEC dose constraints on the DVHs and Wu’s homogeneity
indexes (HI) were considered for the QA of the plans. The structure sets
were then uploaded on the MRIdian TPS and Co60 step and shoot IMRT
plans (7 groups of 3 fields) were calculated. The DHVs and HIs were then
compared to the RapidArc and IMRT plans in order to evaluate MRIdian’s
performances.
Results: MRIdian showed a better HI when compared to the other tech-
niques for PTV1, while this advantage could not be appreciated for PTV2,
even if a better PTV2 V100 (45 Gy) was observed. Comparable mean doses
for the bladder were registered, while a higher bowel V45 was observed
(even if still in the constraints limits). Low dose body V5 was higher for
the MRIdian plans.
Conclusion: A comparable PTV dose coverage between the 3 plans was
found for rectal cancer, with a HI advantage for the PTV1 for the MRIdian
plan. Differences were described for OaRs, especially for low dose areas.
MRIdian allowed to reach dosimetrical goals comparable to RapidArc and
IMRT gold standards. The evaluation of a possible reduction in PTV margin
and a proper target coverage by MRI based gating will be analyzed when
the system becomes operative at Gemelli ART.
A.182
MONTE CARLO DOSE CALCULATION OF VIEWRAY HYBRID MRI-CO60
RADIOTHERAPY SYSTEM: A REPEATABILITY STUDY
E. Placidi * S. Teodoli N. Dinapoli L. Boldrini G.C. Mattiucci
V. Valentini A. Piermattei L. Azario a UOC di Fisica Sanitaria, Fondazione
Policlinico Universitario A. Gemelli, Roma, Italy; b Divisione di Radioterapia
Oncologica, Gemelli-ART, Fondazione Policlinico Universitario A. Gemelli, Roma,
Italy
Introduction: The ViewRay MRI-Co60 hybrid system (MRIdian®) is
a novel technology that provides soft tissue imaging during radiotherapy
allowing real adaptive radiotherapy possibilities and image guidance.
The combination of Co60 with 0.35 Tesla MRI provides MR-guided
IMRT step and shoot delivery with multiple beams (3 Co60 heads
120° apart). MRIdian dose calculation takes advantage of a full
Monte Carlo-based algorithm. The aim of this work was to evaluate the
repeatability of the dose calculation of MRIdian plans for rectal cancer
treatments.
Material and methods: Ten patients affected by locally advanced rectal
cancer were manually segmented on Eclipse TPS v11. MRIdian IMRT
plans (7 groups of 3 fields each) were calculated 5 times for each patient.
The prescribed doses for PTV2 were 4 Gy and 55 Gy for PTV1 through SIB.
The PTV1 V95, the conformity index CI and the Wu’s homogeneity index
HI were computed for each patient. The coefficient of variation (CV),
defined as the ratio of the standard deviation to the mean, was calculated
for each set to express the precision and repeatability of the Monte Carlo
dose calculation. The estimated beam-on time was also recorded for each
plan.
Results: The time for the optimization and the final dose calculation was
less than 5 minutes for each plan having a total of 21 fields (7 groups of 3
heads). The mean CV for the conformity index was found to be equal to
1.6 ± 0.5% whilst both the mean V95(PTV1) and the HI resulted in a CV
smaller than 1%. The CV for the estimated beam-on time for the 10 pa-
tients was found to be 4.3 ± 2.1% (mean ± std).
Conclusion: MRI-guided radiotherapy is a novel approach that may be ad-
vantageous over current treatment techniques by allowing PTV reduction.
Dose optimization and calculation time are done with full Monte Carlo with
a very short calculation time.
The plan parameters under consideration proved that the Monte Carlo dose
calculation is stable with a difference of the order of 4% in the estimated
beam-on time.
A.183
GAFCHROMIC EBT3 CALIBRATION: A NEW CURVE TYPE REDUCING THE
DEPENDENCE ON FILMS LOT
A. Poggiu*, R. Bona, C. Marchioni, F. Pinna, M. Tamponi, P. Marini. SSD. Fisica
Sanitaria – ASL Sassari, Sassari, Italy
Introduction: The aim of this work is to describe a new form of calibra-
tion curve for EBT3 Gafchromic film, using a net optical density ratio, and
to compare it with other two types of calibration curves, commonly used
so far, using the optical density or the net optical density.
Materials and methods: Gafchromic EBT3 films were irradiated by photon
beams and scanned through a 10000XL Epson scanner.
A lot of films were used for calibration (exp. A1) and the three different
curves were built. Six other experiments were done using the same lot (A2,
A3, A4) or different lots (B, C, D) in order to evaluate changes in the accu-
racy and uncertainty of the radiochromic dose, due to the different forms
of the calibration curve, or to the different lots used, to the different time-
after-exposure, or finally to the different scanning orientation.
A general logarithmic dependence of the optical density on the dose was
used to build the three curves, as already validated in the literature. A two
layer model was used to explain, in the monochromatic approximation, the
particular dependence of the new curve type on the sole linear attenua-
tion coefficient of the film layer sensitive to the dose.
Results: Using the same lot, the same scanning orientation and the same
time-after-exposure, the new form of calibration curve has shown equiv-
alent performances as the previous ones, for the accuracy (1.3%) and the
overall uncertainty (2.1%).
Interestingly, a change in lot and/or scanning orientation led to unaccept-
able results for the first two curves, but to quite good results for the new
form of curve: for the 4 lots used accuracy was less than 2.2% (same ori-
entation) and less than 2.8% (changing orientation).
Conclusions: A calibration curve based on the net optical density ratio allows
to limit the lot dependence (and the scanning direction and time-after-
exposure dependence) until values that can be acceptable in the practical
use of Gafchromic EBT3 films, limiting the need of recalibration.
A.184
DIGITAL CAMERA SYSTEM AS A CHEAP AND EASY PREDICTION TOOL OF
SKIN TOXICITY IN BREAST RADIOTHERAPY: VALIDATION AND FUTURE
PERSPECTIVES
M. Poli * S. Bresciani A. Miranti A. Di Dia A. Maggio P. Gabriele
M. Stasi a Medical Physics, Candiolo Cancer Institute – FPO- IRCCS, Candiolo,
Italy; b Radiotherapy, Candiolo Cancer Institute – FPO- IRCCS, Candiolo, Italy
Introduction: In this work the validation of the use of a digital camera,
as a cheap and easy tool for early prediction of radiotherapy skin side effects,
is presented.
Materials and methods: 13 patients undergoing breast radiotherapy were
photographed once a week with a reflex digital camera (Canon 30D, 50 mm
Tamron lens). 9 patients were treated with 3DCRT (Varian TrueBeam STx
linac) and 4 patients with tomotherapy. All photographic shots were ac-
quired in rawmode with the same setup, converted in TIFF format and post-
processed in Lab color space (Color Space Converter plugin, ImageJ) to
amplify color differences. From the red channel (a*), a skin redness map
was obtained for each photographed fraction. Two regions of interest (ROIs)
were identified: one inside the treatment field (IF) and one out-of-field (OF).
OF ROI was used as baseline. Redness value histograms were acquired,
plotted and used to evaluate the degree of skin redness level. ROI-
redness (RR) was defined as the maximum redness value of the related
histogram. IF RR values were plotted as a function of fraction dose and fitted
with a line: its slope was defined as RR gradient. Skin toxicity, evaluated
with RTOG criteria, was compared to RR gradient.
e54 Abstracts/Physica Medica 32 (2016) e1–e70