Fast Monte Carlo log-based framework for robust 4D dose evaluation in proton therapy

Le résumé fourni par la source

Abstract Objective. Cyclotron Centre Bronowice Krakow proton therapy center is establishing a methodology for dosimetric evaluation to support treatments of moving targets, particularly in the mediastinal area. We propose a robust 4D Monte Carlo (MC) dose evaluation approach for proton treatment plans, integrating temporal information on both pencil beam delivery and organ motion in the context of free-breathing treatments. Approach. A specialized tool combining the GPU-accelerated FRED MC code and temporal patient and treatment delivery data was developed. The methodology includes standard 3D dose evaluation for setup and CT calibration uncertainties, plan recalculation across respiratory phases, and robust 4D dose evaluation (4DDE). The 4DDE approach accumulates dose over deformable respiratory phases using log file time information to assess treatment plans at various respiratory cycle starting points. Validation used anonymized 4D computed tomography scans from a Hodgkin lymphoma patient, with deformable image registration extracting motion vector fields. For reference, proton plans were also recalculated on the median tumor position phase image and results were analyzed for dose–volume histograms parameters variability, 3D gamma index (GI) analysis (3%/3 mm) and voxel-wise worst case scenario, relative to the reference treatment planning system calculation. Main results. Exemplary robust 4DDE revealed that considering time provides insight into the worst-case scenarios, resulting in wider dose ranges (GI 55.05% to 88.36% vs 48.26% to 92.92% for 3D and 4D analysis, respectively). The voxel-wise worst-case dose distribution showed reduced target coverage by 47.4% and 20.8% (for 95% and 90% isodose) when accounting for respiratory motion. Significance. The study introduces a fast tool for 4DDE, demonstrating versatility ensures robustness in 3D and 4D evaluations, underscoring the significance of dynamic anatomical considerations in 4D proton radiotherapy.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.