ESTRO 38 Abstract book

S913 ESTRO 38

[5] F. Gómez, C. Fleta, S. Esteban, et al., Physics in Medicine and Biology 2016, 61 4036 EP-1697 Real time small animal irradiator dosimetry using Radioluminescence imaging A. Spinelli 1 , E. D'Agostino 2 , S. Broggi 3 , C. Fiorino 3 , F. Boschi 4 1 San Raffaele Scientific Institute, Experimental Imaging Centre, Milano, Italy ; 2 DoseVue NV, Research and Development, Turnhout, Belgium ; 3 San Raffaele Scientific Institute, Medical Physics, Milano, Italy ; 4 University of Verona, Department of Computer Science, Verona, Italy Purpose or Objective The main objective of this work was the development of a novel real-time 2D dosimetry approach for small animal external radiotherapy using radioluminescence imaging (RLI) with a commercial CMOS detector. Radioluminescence light is generated by a flexible and thin scintillator composed of green-emitting rare earth phosphors. Material and Methods A CMOS detector was mounted on a photographic tripod and coupled with a F=1.4, 8 mm C-mount lens (Edmund Optics) facing the scintillator screen taped on the back side of the animal stage. Measurements were performed on the small animal image-guided platform SmART (Precision X-Ray Inc.). As shown in figure 1 the CMOS was inserted in the XRAD225Cx SmART cabinet, 40 cm away from the animal stage. The SmART cabinet was light-tight and, thus, allowed to investigate RLI with different lighting conditions by switching on/off the internal led during image acquisition.

using novel silicon 3D-microdetectors [3]. Results were verified with both Geant4 and FLUKA simulations. Material and Methods The new 3D-microdetectors used are a type of diode with a 3D-cylindrical electrode etching (15 μm diameter, 5 μm depth) with an inner volume that matches a sensitive volume that simulates a subcellular structure. Details are described elsewhere [3,4]. The 18 MeV proton beams were generated in a Cyclone 18/9 cyclotron at flux rates of 10 8 p/cm 2 s. The proton beam average energy was modulated with an in-house wedge system formed by two equal 10º angle wedges made of 1.19 g/cm 3 Lucite that provides a continuously variable thickness from 3 mm up to 30 mm with an uncertainty of 30 µm [5]. The cyclotron proton beamline was simulated with Geant4.10.4. The so simulated outgoing energy spectrum was used as input-file within Geant4 and FLUKA simulations of the 3D- microdetector performance. Results Fig. 1 shows the most probable lineal energy through the Bragg curve and measured microdosimetry spectra. The experimental silicon microdosimetric spectra showed clearly the variations of the lineal energy frequency distribution along the Bragg curve.

Figure 1: Left: Most probable lineal energy versus PMMA depth curve for the 18 MeV protons measured with the novel 3D-microdetector. Right: Measured microdosimetry spectra as a function of the PMMA thickness. Conclusion Measurements at various depths along the Bragg curve of a 18 MeV proton beam were performed at high fluence rate with a novel 3D-microdetector and both microdosimetric spectra and Bragg curves were obtained. The results show that this new device is useful to microdosimetry characterization of clinical proton beams. [1] Schardt D. and Elsässer T., Reviews of Modern Physics, Vo. 82, 2010. [2] Bassler N, Jäkel O., Sondergaard CS, et al. Acta Oncol. 2010 Oct;49(7):1170-6. [3] C. Guardiola, F. Gómez, G. Pellegrini, et al., Applied Physics Letters 107, 023505 (2015). [4] C. Fleta, S. Esteban, M. Baselga et al., 2015 JINST 10 P10001

Figure 1 RLI acquisition setup. The CMOS setting were as follow: gain=22, brightness=50 and exposure time=1/500 sec. A sequence of 60 images were acquired every 1 sec and summed. RLI data were corrected for perspective distortion with an home-made program written in Matlab. The SmART setting were as follow: tube voltage=225kV, current= 1, 3, 6, 9 and 13 mA, the corresponding dose rate at 13 mA was 4.2 Gy/min. Four fields were tested, 2 circular beams with diameter equal to 1 mm and 3 mm and 2 square fields (10x10 mm and 20x20 mm). Measured planar 2D dose distributions and dose profiles were compared against dose calculation performed with Monte Carlo (MC) simulations, including