STUDYING THE OXYGEN EFFECT IN PROTON THERAPY VIA GPU-BASED MICROSCOPIC MONTE CARLO SIMULATION
Abstract
Dissolved oxygen molecules are known to play an important role in radiotherapy. For example, tumor cells can be more radio-resistance under hypoxia than normoxia. There has been multiple hypothesis proposed to explain these phenomena, including the oxygen mediated chemical track evolution. In principle, we can perform microscopic Monte Carlo (MC) simulation to quantitively test this hypothesis. Yet, to avoid the extreme high computational resource consuming, in the current CPU-based MC tools, oxygen is either ignored or taken as a continuous radical-scavenging background. Recently, our research group developed an open-source, GPU-based MC simulation tool, gMicroMC, in which, the oxygen mediated chemical track evolution was modeled in a step-by-step manner. In this project, we applied gMicroMC to study the radical evolution under different oxygen levels for proton induced water radiolysis. Compared to the continuous background method, the oxygen was found consumed slower with gMicroMC. We also documented the errors and issues as a fresh user of gMicroMC. A further automation for some parameter settings can make it more user-friendly.