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

30 / 146

Materials and methods: Ten patients with liver tumour were included in
the study. Six plans were created for each patient. Two plans were gener-
ated with AP of Pinnacle3 TPS using SmartArc technique and two with
traditional planning (MP) by two different expert medical physicists. Other
two experts performed two VMAT plans with Monaco TPS (version 5.0,
Elekta) (VM). Dosimetry comparison was done in terms of the PTV cover-
age, gEUD, OARs sparing, homogeneity index (HI) and conformity index (CI).
Also total monitor units, number of beam segments and beam’s complex-
ity metrics (plan average beam area BA, plan average beam irregularity PI
and plan average beam modulation PM) were evaluated.
Results: Preliminary results indicated that, for same gEUD (p value = 0.99),
relevant differences were found about beam’s complexity metrics (p = 0.23
for BA, 0.01 for PI and 0.05 for PM), HI (p = 0.03), monitor units and OAR
sparing. In particular, median and mean values of monitor units were re-
spectively 3212 and 3646 ± 1529 for AP, 2930 and 2923 ± 447 for MP and
5006 and 4850 ± 570 for VM. Also for OARs, results showed different
behaviours of TPS. In particular, the healthy liver median volume below
15 Gy was 592 cc for AP, 596 cc for MP and 659 cc for VM; the mean values
were 625 ± 150 cc for AP, 632 ± 120 cc for MP and 673 ± 46 cc for VM.
Conclusions: Analysis of first three patients demonstrated that AP and MP
employed much less monitor units with respect to VM and showed a minor
PI. However, in particular complex cases, AP and MP had more difficulty
to spare the organs at risk than VM. Furthermore, there was sensible intra-
patient variability for AP and MP. AP was less human employment time
consuming than both manual planning systems.
A.83
PHENOL COMPOUNDS AS NEW MATERIALS FOR ELECTRON SPIN
RESONANCE DOSIMETRY IN RADIOTHERAPY
S. Gallo M. Marrale G. Iacoviello S. Panzeca S. Altieri
V. Caputo G. Collura F. D’Errico A.M. Gueli A. Longo
M. Bra a Dipartimento di Fisica e Chimica, Università di Palermo, Palermo,
Italy; b Istituto Nazionale di Fisica Nucleare (INFN) Gruppo V SEZ. Catania,
Catania, Italy; c PH3DRA Laboratories, Dipartimento di Fisica e Astronomia,
Università di Catania, Catania, Italy; d Scuola di Specializzazione in Fisica Medica,
Università di Palermo, Palermo, Italy; e Unità Operativa Complessa di Fisica
Medica, ARNAS Ospedale Civico di Palermo, Palermo, Italy; f Dipartimento di
Fisica, Università di Pavia e Istituto Nazionale di Fisica Nucleare (INFN) SEZ.
Pavia, Pavia, Italy; g Dipartimento di Ingegneria Civile e Industriale, Università
di Pisa, Pisa, Italy; h Yale University, Yale, USA; i Scuola di Specializzazione in
Fisica Medica, Università di Catania, Catania, Italy
Introduction: Among the various dosimetric techniques used for charac-
terizing the radiation beams used in radiation therapy, the electron spin
resonance (ESR) arouses increasing interest for applications in various
therapy procedures. Free radicals are known to be produced when a com-
pound is irradiated with ionizing radiations. The concentration of radiation-
induced free radicals is proportional to the absorbed dose and this allows
for dosimetric measurements through ESR technique which enables to quan-
titatively determine the radical concentration.
In this work we report the ESR investigation of phenol pellets and thin films
exposed to various types of radiation beams (clinical photon and electron
beams, mix field neutrons/photons). Phenols are compounds possessing
a benzene ring attached to a OH group. After irradiation the final product
is a stable phenoxy radical.
Materials and methods: Thanks to their size, the phenolic films here pre-
sented are good devices for the dosimetry of beams with high dose gradient
and which require accurate knowledge of the precise dose delivered.
Photon and electron irradiations at various energies were performed with
the Siemens Primus Low linear accelerator present at ARNAS Ospedale Civico
di Palermo, Italy. Thermal neutron irradiations were performed at the
thermal neutron column at the Triga Mark II reactor of LENA (Laboratorio
Energia Nucleare Applicata) of Pavia, Italy. ESR dosimeters were readout
by means of a Bruker ECS106 spectrometer equipped with a TE102 rect-
angular cavity at room temperature.
Results: The dose response of both pellets and films was found to be linear
for all beams used in the dose range analyzed. The signal is very stable in
the first days after irradiation. The possibility of obtaining depth dose profile
was investigated.
Conclusions: In conclusion, the phenols show radiometric features that des-
ignate it as a new material for EPR.
A.84
LONG PERIOD ACCURACY AND REPRODUCIBILITY OF MLC LEAVES
POSITION ACQUIRED IN A DAILY QA PROGRAM USING EPID IMAGES
A. Gambirasio P. Colleoni C. Bianchi M. Fortunato S. Andreoli
a Fisica Sanitaria, A.O. Papa Giovanni XXIII, Bergamo, Italy; b Scuola
Specializzazione in Fisica Medica, Università degli Studi di Milano, Milano, Italy
Introduction: We present the analysis of the EPID images acquired during
the last two years as a part of the daily MLC QA program of two linear ac-
celerators. These images have been used to assess qualitatively MLC leaves
position by a visual evaluation of the displacement between recorded and
expected leaves positions. Specifically developed analysis SW allows evalu-
ating from the images the accuracy and reproducibility of each leaf position.
Materials and methods: The images were acquired on two Varian Medical
Systems linear accelerators both equipped with Millennium 120 MLC and
with Varian AS1000 flat panel detector with an active area of 40 × 30 cm 2
and with 1024 × 768 pixels. Images were acquired with the best quality
image modality. Analysis was performed using an in-house code devel-
oped in MATLAB®.
Results: Results confirm quantitatively the good accuracy and reproduc-
ibility of MLC leaves position already assessed by visual inspection.
Nevertheless, it is possible to highlight some new aspect, which could help
to improve MLC leaves QA. We found that, in each image, differences
between extracted and expected MLC positions are normal distributed. In
some cases, distribution centroid fall definitely far apart zero. This fact can
be related to small systematic errors in initial set-up of the MLC leaves.
Moreover some leaves were found to have a displacement, which lays more
than 3 sigma far from the centroid and this could be related to a non-
optimal functionality of the leaves.
Conclusions: The analysis allows a submillimetric quantitative evalua-
tion of displacement of each leaf form the expected position and confirms
the good accuracy and reproducibility of MLC leaves positioning over time.
However, new aspects, which cannot be found by a visual inspection, can
be identified and can help to improve quality of MLC leaves positioning.
Further investigations are needed to assess if these findings can be related
to a bad initial set-up or prediction of malfunctioning of the leaves.
A.85
AUTOMATIC TOOL FOR MLC LEAVES POSITION AND COLLIMATOR
ROTATION DETECTION USING EPID IMAGES
A. Gambirasio * M. Mori P. Colleoni L. Perna C. Bianchi
M. Fortunato S. Broggi S. Andreol a Fisica Sanitaria, A.O. Papa Giovanni
XXIII, Bergamo, Italy; b Medical Physics Department, IRCCS San Raffaele Scientific
Institute, Milano, Italy; c Scuola Specializzazione in Fisica Medica, Università
degli Studi di Milano, Milano, Italy
Introduction: The aim of this work was to test the reliability and the ef-
ficiency of some methods to analyze EPID static images in order to
automatically detect MLC leaves position and collimator rotation.
Materials and methods: To find local edges, two different detection methods
were considered: the Canny’s method and the isointensity approach. Then,
a homemade edge linking tool was used to identify edge section belong-
ing to each leaf tip. FFT-based algorithm was implemented to evaluate
collimator angle rotation. The code including all these tools has been de-
veloped using MATLAB® as platform. The robustness of these tools was
verified on EPID devices (Varian Medical Systems) with two different res-
olution grids: an AS1000 with 1024 × 768 and AS500 with 512× 384 pixels.
Geometrical and clinic MLC patterns (Varian Millennium 120-leaf MLC) were
tested; position errors were also intentionally introduced in some pat-
terns. The differences between measured and planned MLC position/
collimator rotation were reported.
Results: For EPID with higher resolution, Canny’s and isointensity methods
were able to reach acceptable results, both for geometrical and clinical MLC
patterns; for the EPID with a low resolution the isointensity method turned
e25 Abstracts/Physica Medica 32 (2016) e1–e70