ESTRO 37 Abstract book

ESTRO 37

S460

Conclusion DIAPIX, a bidimensional pCVD diamond detector with 2- mm pitch pixels covering an active area of 5.0 x 2.5 cm^2, manufactured within the DIAPIX and IRTP projects funded by INFN Italy, has been tested for pretreatment verifications of CK and EBM plans showing on average very good results. Improved pixel contact quality will ensure a higher uniformity of the device response. PO-0874 Proton therapy range verification accuracy considering range mixing for time resolved dosimetry A. Toltz 1 , J. Seuntjens 2 , H. Lu 3 , H. Paganetti 3 1 McGill University, Physics, Montreal, Canada 2 McGill University Health Centre and McGill University, Medical Physics Unit, Montreal, Canada 3 Massachusetts General Hospital, Radiation Oncology, Boston, USA Purpose or Objective With the aim of reducing acute esophageal radiation toxicity in pediatric patients receiving craniospinal irradiation (CSI), we investigated the implementation of an in vivo , adaptive proton therapy range verification methodology. Simulation experiments and in-phantom measurements were conducted to quantify the accuracy of the range verification measurement in the presence of range mixing for our time-resolved diode dosimetry technique. Material and Methods A silicon diode array system has been developed and experimentally tested in phantom and modeled in simulation studies for passively scattered proton beam range verification by correlating the detector signal in time to the water equivalent path length (WEPL). We propose to extend the methodology to verify range distal to the vertebral body in a posterior/anterior beam configuration for pediatric CSI cases by performing a measurement with the small volume dosimeter placed in the esophagus of the anesthetized patient immediately prior to treatment. A set of measurements was performed to establish the relationship between time signal and WEPL for two ‘scout’ beams in a custom phantom comprising slabs of bone-equivalent and water- equivalent plastic in a geometric approximation of the CSI clinical scenario. 'Scout' beams are defined as a 1 cm overshoot of the predicted detector depth with a dose of 4 cGy. Measurements were compared against Monte Carlo simulation in GEANT4 using the Tool for Particle Simulation (TOPAS) and against treatment planning system-calculated radiological path length in XiO on a CT of the phantom. Results The accuracy for determining WEPL with our detector in the presence of range mixing has been quantified to within 2 mm uncertainty in phantom studies through TOPAS simulation, experimental measurements, and TPS for the examined ‘scout’ beams of 10 cm and 15 cm range, corresponding to energies of 168 MeV and 177 MeV

verification of small size high intensity modulated treatments. Material and Methods DIAPIX consists of two diamond matrices (Diamond Detector, UK) of area 2.5×2.5 cm^2 and 300 µm thick. In house Cr/Au electric contacts were realized on the front and rear surfaces. The upper contact is segmented in a 12×12 matrix of pixels of area 1.8×1.8mm^2 and 2 mm pitch. The two pCVD matrices were sandwiched inside slabs of PMMA (see Fig.1). We have previously shown [Journal of Instrumentation,V12,2017] that this device is suitable to be used as a radiation-hard, tissue-equivalent dosimeter and was therefore used in the pretreatment verifications of two clinical lung plans delivered by Cyberknife (CK) and a linear accelerator Elekta Synergy BM (EBM) respectively. The corresponding patient QA plans were generated with Multiplan Cyberknife TPS and Monaco Elekta TPS. To cover the full area of the linac plan the detector was first positioned at the isocenter and then shifted in the latero-lateral and in gun-target direction. The dose distributions measured with DIAPIX were compared with the corresponding ones calculated by the TPS by evaluating the γ-index with an in-house developed MATLAB routine. Before the gamma evaluation, the two dose distributions were registered by using intensity-based algorithms. The passing rate (γ < 1) was evaluated for the passing criteria 3% /3mm.

Results Results: Figure 2 presents the calculated and measured dose distributions for both CK and EBM treatments. For CK the passing rate over the whole matrix was around 72% while for the EBM was 77%. However, due to problems in electrical contacts, several pixels, mostly concentrated in one of the two matrices, were not working properly and considering only an area without DIAPIX bad contacts (black pixels) the passing rate increases considerably. For example, γ <1 corresponding to the highlighted area in the figure reaches 94% for CK and 98% for EBM, respectively.