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

68 / 146

A.212
A NEW METHOD TO CALIBRATE GAFCHROMIC® EBT3 FILMS FOR
ROUTINE QUALITY CONTROL IN RADIOTHERAPY PHOTON BEAMS
S. Spampinato *, L. Raffaele, A. Gueli. Scuola di Specializzazione in Fisica
Medica, Università di Catania, Catania, Italy
Introduction: The properties of Gafchromic® EBT3 films are well known
by the medical physics community. However, their use is often limited by
a tedious calibration procedure, which involves many exposures to create
a curve that links pixel value (PV) with absorbed dose.
This work proposes an alternative method that allows to calibrate in a simple
and fast way these films, using a single exposure and a wedged field of a
radiotherapy LINAC.
Materials and Methods: The used tools are easily found in radiotherapy
units: an RW3 Slab Phantom and a diode array (MapCheck2, SunNuclear).
For the calibration procedure, the array, previously calibrated in absolute
dose, was used to measure a map of a 12 × 12 cm2 6 MV photon field with
a virtual wedge (45°). Subsequently, an EBT3 film has been exposed with
the same experimental setup (SSD 95 cm, SDD 100 cm) and the same field.
The image was acquired with a flatbed scanner (72 dpi, 48bit colour depth)
and for the analysis the red channel was used. To obtain the calibration
curve, the profile along the direction of the wedge passing from CAX (iden-
tified by landmarks) was compared with that measured by the array. To
validate the procedure, standard fields were acquired and compared with
the standard calibration.
Results: The exposure with a wedged field allows having simultaneously
several spatially identifiable points of dose. By comparing the profiles, a
point identified by the array corresponds to a point in the film character-
ized by a PV. In this way, the number of diodes in the profile limits the
number of points available. Comparing PV and dose allows performing the
fit, which, as known from the literature, is well represented by a polyno-
mial of the 3rd degree. The comparison with the standard fields has
confirmed the reliability of this procedure, comparable within 1% with the
standard protocol.
Conclusions: The proposed procedure allows a simple and fast calibra-
tion of EBT3 films and is comparable to the standard technique.
A.213
AN IN-HOUSE SOFTWARE TO CREATE SYNTHETIC-CT BASED ON MRI
S. Spreafico C. Cozzaglio L. Mascaro C. Toraci M. Maddalo
L. Spiazzi * M. Buglione S.M. Magrini R. Moretti a Cattedra di
Radioterapia Università degli studi, Brescia, Italy; b U.O. Fisica Sanitaria A.O.
Spedali Civili di Brescia, Brescia, Italy; c Medipass Srl, Bologna, Italy; d Università
degli Studi di Milano, Milano, Italy
Introduction: MRIs are emerging as images of choice for radiotherapy treat-
ment planning, allowing a target and OARs delineation directly on functional
and multiparametric images.
Geometric uncertainty would reduce thanks to the fact that CT–MR match-
ing would not be necessary anymore.
Treatment planning bases the tissue density definition on the HU/density
curve; therefore, an in-house SW was developed capable to create syn-
thetic Dicom CT (SCT) based on MR images.
Matherial and Methods: Programming language: Php.
Web server: Apache.
Image viewer: ImageJ; Osirix.
Velocity 3.0 (Varian); Pinnacle 10.0 and Isite (Philips).
To test our SW, we used T1 TSE; T2 TSE; Dark Fluid; T1 flow e T1 TSE320
sequences for 3 head patients.
A deformable registration of each MR on the planning CT was performed
and all images were resampled in the CT frame.
Then a deformable registration of each MR on the T1 TSE sequence was
performed and all images were resampled in the T1 TSE frame.
For every patient two images were created:
• CT in the CT frame.
• SCT in the MR frame.
To derived a correct image density distribution, we defined a series of 11
value combinations for the 5 MR that corresponds to different tissue density.
In order to check the results reliability, we performed a dose calculation
for a central target with 12 equidistant fields with 100 UM each/100 frac-
tion for the SCT based on CT frame and the CT itself.
Results: SCT were correctly handled by Pinnacle and dose calculation was
possible.
Calculated dose in the CT is 1150 Gy and the difference between SCT dose
and CT dose was less than 2% in the target region for the reference patient,
while was between 2% and 3% at max for the other 2.
The SW therefore demonstrated the capability to create correct SCT and
can be used to test more analytic and reliable MR density conversion
algorithms.
A.214
VALIDATION OF A NEW COMMERCIAL OPTIMIZATION TOOL FOR VMAT
TREATMENT PLANNING
I. Martinett C. Cozzaglio M. Buglione F. Corrado F. Dus A. Fiume
B. Ghedi F. Saian C. Toraci S.M. Magrin R. Moretti L. Spiazzi
a Cattedra di Radioterapia Università degli Studi, Brescia, Italy; b U.O. Fisica
Sanitaria Spedali Civili di Brescia, Brescia, Italy; c Università degli Studi, Milano,
Italy; d Medipass S.r.l., Bologna, Italy
Introduction: To verify Philips Pinnacle “Auto-planning” (AP) module for
VMAT optimization, its plans will be compared with accepted ones.
AP implementation needs a training phase with a definition of clinical goal,
physics beam settings, and analysis of accepted treatments. Since these ac-
cepted plans are used by the system in order to tune up the optimization
procedure they cannot be used to verify the optimized VMAT quality.
Material and Methods: AP was set up using the results and prescriptions
for 10 prostate patients, 5 pelvic region patients and 5 H&N patients.
The table summarizes the different implemented fractionation scheme.
Prostate P1 P2 P3 P4 P5
Nr fractions 19 20 35 37 39
PTV (Gy) 60.8 64 70 74 78
Pelvic region PR1 PR2 PR3
Nr fractions 25 28 32
PTV T (Gy) 45 50.4 64
PTV low (Gy) / / 54.4
H&N H1 H2 H3
Nr fractions 30 33 35
PTV T (Gy) 60 69.3 70
PTV 1 (Gy) / 66 63
PTV low (Gy) 54 51.1 56
After system implementation, 10 AP treatment planning for each district
already elaborated with our institution and standard procedure were
prepared.
Results in terms of DVHs, MU and mean field area were compared.
Since the plans had different prescriptions to compare original and AP planes
dose differences for each OAR or PTV, relative volume were calculated. Paired
t-student test to check statistical difference was used.
Results: The AP plans had a significantly higher MU number (+23%, p < 0.02)
as a consequence of a smaller mean field area.
The target coverage for standard treatment and AP did not show statisti-
cal difference for relative volume higher than 95%, and less than 5% (the
relevant PTV’s volume range).
OAR had generally a lower mean dose, but higher doses (especially for
rectum and bladder) were significantly higher than standard treatment.
The optimization tool seems to be promising, but operator experience and
decisions seem to be nonetheless important to customize the results to any
specific patient.
e63 Abstracts/Physica Medica 32 (2016) e1–e70