A new GPU-based Monte Carlo code for helium ion therapy.

Area of Science:

• Medical Physics
• Computational Science
• Radiotherapy

Background:

• Helium ion therapy offers potential advantages in radiation oncology.
• Accurate and efficient dose calculation is crucial for treatment planning.
• Existing simulation tools may lack the computational speed required for complex treatments.

Purpose of the Study:

• To extend the capabilities of the GPU-based Monte Carlo code ARCHER for helium ion therapy.
• To develop and validate a new nuclear-event-repeat algorithm for inelastic nuclear reactions.
• To enhance computational efficiency using the CUDA platform.

Main Methods:

• Simulated helium ion transport in voxelized geometry up to 220 MeV/u using a class II condensed-history algorithm.
• Modeled ionization, energy straggling, multiple scattering, and nuclear interactions.
• Developed a novel nuclear-event-repeat algorithm and tracked secondary particles; validated against TOPAS.

Main Results:

• Strong agreement in dose distribution between ARCHER and TOPAS, with local relative errors below 1.31%.
• Achieved high gamma passing rates (e.g., 99.8% for head-neck case) using the 1%/1 mm criterion.
• ARCHER demonstrated significant speedup, simulating 10 million primary helium ions in under 10 seconds on a single GPU, compared to minutes for TOPAS on a multi-CPU system.

Conclusions:

• ARCHER is the first GPU-based dose engine developed for helium ion therapy.
• The code achieves high accuracy, comparable to established methods.
• ARCHER offers substantial computational efficiency, enabling faster and potentially more complex treatment planning.