Searching for optimized selection of Monte Carlo dose calculation parameters for scanned beam proton therapy in RayStation

Abstract

To provide a practical guide on the selection of Monte Carlo (MC) dose calculation parameters for scanned beam proton therapy in RayStation, we generated treatment plans with various combinations of two MC dose calculation parameters in the optimization and compared the elapsed time and dosimetric properties of those plans. We selected 23 clinical cases treated with scanned beam proton therapy. For each clinically approved treatment plan, we generated 26 exploring plans (1 plan with clinical pencil beam dose algorithm and 25 plans with MC dose algorithm with varying dose calculation parameters). We recorded the elapsed time for individual plans in the dose calculation using an automated RayStation script. To evaluate the dosimetric quality of the exploring plans, the conformity and homogeneity indices of each plan were calculated and dose-volume histograms for the target volume and organs-at-risk were assessed. We found the combination of 12,000 ions/spot in the optimization and 0.5% uncertainty in the final dose calculation would provide a dosimetrically competent plan with a reasonable amount of dose calculation time. We would suggest using 8000 ions/spot in the optimization and 1.0% uncertainty in the final dose calculation as a starting point in the treatment planning of scanned beam proton therapy. The practical recommendations resulting from this study, related to MC dose calculations, have the potential to enhance treatment planning strategies for scanned beam proton therapy and to make the iterative process of treatment planning more efficient.