Monte Carlo simulations for imaging in proton therapy

Proton therapy is rapidly gaining importance in the field of radiotherapy, because of its potential to deliver the planned dose over a small depth range and sparing dose to healthy tissue, when compared to conventional radiotherapy. Proton therapy, however, makes the need of new imaging modalities for treatment planning, based on direct measurements of tissue stopping power and eliminating the need to convert tissue density – as measured in conventional X-ray Computed Tomography) – to stopping power, upon which treatment planning is based [1] . The expected benefits of proton CT (pCT) for treatment planning in Proton Radiotherapy are producing great interest worldwide to develop instruments for clinical-quality pCT. The PRaVDA (Proton Radiotherapy Verifications and Dosimetry Applications) consortium has developed a novel purely solid state system and associated image reconstruction for pCT [2,3] . Custom detectors for the PRaVDA instrument include Silicon Strip Detectors (SSD) and Monolithic Active Pixel Sensors (MAPS). To assist design decisions, detector development and development of image reconstruction algorithms, comprehensive Geant4-based Monte Carlo simulations were developed, comprising realistic beam-line sources for two proton sources where the PRaVDA instrument has been tested (MC40 cyclotron at the University of Birmingham and iThemba LABS cyclotron), full device geometry and realistic readout for both SSDs and MAPSs. The full paper will focus on simulations of MAPS for imaging in proton therapy. The full read-out chains of a MAPS – including charge diffusion, collection, sharing and digitalization, has been simulated through the development of ad hoc classes integrated in the standard Geant4 toolkit. An experimental validation of the model (at two different proton sources) will be provided and the contribution of these simulations to assess imaging capabilities of MAPS, in the context of proton therapy, will be highlighted.