Developing a Medical Linac Monte Carlo Model in the TOPAS Framework

Sammanfattning

This thesis presents the development and validation of a Monte Carlo model of a medical linear accelerator (Linac) using the TOPAS simulation framework, conducted in collaboration with Elekta. The work focuses on three key areas: the implementation of variance reduction techniques, the selection of optimal electromagnetic (EM) physics modules within TOPAS, and a case study on beam symmetry. Results demonstrate that combining secondary biasing with directional splitting significantly improves simulation efficiency without compromising accuracy, with directional splitting toward the primary collimator being the most effective strategy. Among the evaluated EM physicsmodules, all produced comparable dose distributions within a water phantom, but option 2 emerged as the most practical due to its reduced computation time. A mismatch observed between simulated and measured percentage depth dose (PDD) curves is attributed to incomplete tuning of the electron beam configuration rather than deficiencies in the physics models. The beam symmetry study shows that minor misalignments in the beam shaping assembly within 0.4 mm or a small angular offset when measuring dose in the water phantom remain within system requirements, while larger deviations exceed acceptable limits. Overall, this work provides an accurate and computationally efficient framework for Linac modeling, supporting future research and development in radiotherapy beam optimization and system design.