ESTRO 36 Abstract Book

S29 ESTRO 36 2017 _______________________________________________________________________________________________

dependent quenching can be corrected by adjusting the measured 3D optical density (OD) distribution using a calibration model generated from a Monte Carlo (MC) simulation and a calibration dosimeter. Material and Methods A radiochromic (leucomalachite green) silicone-based 3D dosimeter was irradiated with three spatially separated, unmodulated 60 MeV proton beams (Ø 10 mm collimator), to plateau doses of 5, 10 and 20 Gy. Dosimeter read-out was performed prior to and after irradiation, in a Vista 10 optical CT scanner with 0.25 mm 3 voxel resolution, thus obtaining the change in OD caused by the radiation. MC simulation of 10 8 protons was performed in SHIELDHIT- 12A, giving the dose and dose-averaged LET (dLET) distributions. A calibration model with respect to dose and dLET was generated from the 5 and 20 Gy Bragg peaks: Linear fits to OD as a function of dose was computed for all voxels within limited ranges in dLET. Only voxels within the central part of the beam (Ø 5 mm) with doses above 1% of the maximum dose were used for the model. The slopes and intercepts from the linear fits were smoothed as a function of dLET (Fig. 1). Voxels in the 10 Gy Bragg peak were calibrated using these variables and compared to the dose distribution from the MC simulation (Fig. 2). Results Quenching results in a lower dose response in the Bragg peak of the proton beam (0.013 cm -1 Gy -1 ) compared to the low-LET regions, such as the plateau (0.027 cm -1 Gy -1 at 2.94 keV/µ m), see Fig. 1. The dose response increased to a sharp peak for dLET-values within the plateau region. The calibrated optical measurement was close to identical to the MC simulated dose in the central part of the beam. Less good results were obtained towards the edges of the beam (Fig. 2), which corresponded to areas where the MC beam model was suboptimal. Dose errors larger than 2 % of maximum dose (1.1 Gy) was found in 35 % of all calibrated voxels, while 3.5 % had dose errors larger than 5 % of maximum dose (2.7 Gy).

Conclusion We present the first LET-corrected 3D dose m easurements in a radiochromic dosimeter for proton thera py, showing that verification for single fields has been made possible. Acknowledgements: PL-Grid OC-0063 Energy resolution and range reproducibility of a dedicated phantom for proton PBS daily QA L. Placidi 1 , J. Hrbacek 1 , M. Togno 2 , D.C. Weber 3 , A.J. Lomax 1 1 Paul Scherrer Institute PSI, Medical physics, Villigen PSI, Switzerland 2 IBA, Dosimetry GmbH, Schwarzenbruck, Germany 3 Paul Scherrer Institute PSI, Radiation Oncology, Villigen PSI, Switzerland Purpose or Objective Wedge phantoms coupled with a CCD camera (Fig.1) were suggested as a simple mean to verify range consistency or reproducibility for a specific energy. The method is based on analysing of integral image created by a spot pattern that passed through a wedge. We have investigated the reproducibility and dependence of this method on setup errors (shift and tilt) for a commercially available phantom (Sphinx, IBA Dosimetry) and CCD camera (Lynx,