Room: Exhibit Hall
Purpose: Monte Carlo techniques are important tools in radiotherapy due to their ability to accurately calculate dose in heterogeneous mediums. Our clinic has validated TOPAS (version 3.1.p2) for the Mevion S250 double scattering proton therapy system and developed a complete workflow for clinical patient dose calculation. The purpose of this study is to visually and analytically compare the doses calculated with the Varian Eclipse pencil beam convolution (PBC) algorithm and TOPAS Monte Carlo (MC) simulations for four different treatment sites.
Methods: Patients were chosen with treatments in four targets, including abdomen, lung, brain and pelvis. Treatment plans were made for each patient using Varian Eclipse. For each plan, the matching beam line was modeled in TOPAS and the three dimensional dose was simulated. Exported doses from treatment planning software were then visualized with 3D Slicer where the simulated doses were normalized to match the Eclipse dose at the isocenter. 3D gamma analysis was then performed using an in-house software tool with criteria of 5mm in distance, 7% in dose difference and 10% threshold.
Results: Gamma passing rates for the pelvis, brain, abdomen and lung were 95.05%, 88.11%, 79.04% and 70.09%, respectively. The plans in areas with higher tissue heterogeneities were found to have lower gamma passing rates, with the lung having the greatest difference between PBC and MC. The greatest dose differences were often shown at the lateral boundaries of the field and the distal fall off region.
Conclusion: This work highlights the uncertainty in proton dose calculations using pencil beam convolution in heterogeneous materials and shows the potential of using MC simulation in routine clinical practice.