GPU-accelerated Monte Carlo code for fast dose recalculation in proton beam therapy

Proton therapy has rapidly grown in the past thirty years and it has become a superior alternative to conventional radiotherapy for certain clin- ical indications. Proton therapy offers high dose selectivity due to the protons’ distinct depth dose profile which potentially allows to deliver high dose to the tumor while sparing healthy surrounding tissue. Monte Carlo (MC) simulations, which take explicitly into account all the details in the interaction of particles with human tissues, are considered to be the most reliable tool to reproduce the complexity of mixed-field irradiation in a non- homogeneous environment. The advent of general-purpose programming GPU cards prompted the development of trimmed-down MC-based dose engines, which can significantly reduce the plan recalculation time with re- spect to standard MC codes on CPU hardware. In this contribution, the GPU-accelerated MC treatment planning system (TPS) Fred developed by the University of Rome is presented (Schiavi et al., Phys. Med. Biol. (2017)). The current status of the implementation in Fred of the experi- mentally measured physical beam model data used for treatment planning at the Cyclotron Center Bronowice (CCB) Kraków proton beam therapy centre is reported. Three-dimensional dose distributions of proton pencil beams in a water phantom, i.e. lateral profiles and depth dose distributions, are compared for different beam configurations