GPU-Accelerated Monte Carlo Dose Calculation
GPU acceleration is essential to bring Monte Carlo dose calculation into real-time clinical workflows, especially for adaptive treatments guided by in-room MRI. This project focuses on porting the MonteRay dose engine to GPUs to achieve a 10–20× speedup. The work involves profiling, parallelization, and CUDA development, with the goal of enabling second-scale dose computation even in the presence of complex magnetic fields. Special focus will be on handling inhomogeneous magnetic fields, which are unsupported by current clinical engines.
Requirements
- Very good English skills
- Experience with scripting (e.g., Python), Linux, Git, and CMake
- Strong C++ and CUDA skills
- Background in particle physics, medical physics, or computer science is a plus
Posted: 09/07/2025
Contact: peter.lysakovski@uni-heidelberg.de
Mini-Beam Oxygen Ion Therapy in MonteRay
Mini-beams can reduce damage to healthy tissue by spatially fractionating the dose, potentially improving tolerance in sensitive regions. Oxygen ions provide sharper dose distributions and greater biological effectiveness than protons or carbon, making them attractive for resistant tumors. This project extends MonteRay to support oxygen ion therapy and mini-beam treatment planning, including nuclear interaction data for oxygen and relevant filter materials, and implementing mini-beam geometry simulation. May include experimental validation.
Requirements
- Very good English skills
- Experience with scripting (e.g., Python), Linux, Git
- Background in particle physics, medical physics, or computer science is a plus
Posted: 09/07/2025
Contact: peter.lysakovski@uni-heidelberg.de