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

Abstract

Purpose: This work presents an effort to extend the capabilities of the previously introduced GPU-based Monte Carlo code ARCHER for helium ion therapy.

Methods: ARCHER performs helium ion transport simulations in voxelized geometry, covering kinetic energy levels up to 220 MeV/u. The physical processes are modeled using a class II condensed-history algorithm, considering ionization, energy straggling, multiple scattering, and elastic and inelastic nuclear interactions. A new nuclear-event-repeat algorithm is proposed to generate inelastic nuclear reaction products. Secondary protons, deuterons, tritons, and ³He particles are tracked, while other particles either deposit their energy locally or are ignored. The code is developed under the compute unified device architecture (CUDA) platform to improve computational efficiency. Validations are conducted by benchmarking our code against TOPAS in different phantoms.

Results: Dose distribution comparisons demonstrate strong agreement between our code and TOPAS. The mean point-by-point local relative errors in the region where the dose exceeds 10% of the maximum dose range from 0.25% to 1.31% for all phantoms. In the strict 1%/1 mm criterion, gamma passing rates for a head-neck case, chest case, and prostate case are 99.8%, 96.9%, and 99.6%, respectively. Except for the lung phantom, ARCHER takes less than 10 s to simulate 10 million primary helium ions using a single NVIDIA GeForce RTX 3080 card (NVIDIA Corporation, Santa Clara, USA), while TOPAS requires several minutes on a computational platform with two Intel Xeon Gold 6348 CPUs (Intel Corporation, Santa Clara, USA) with 56 cores.

Conclusion: This work presents the development and benchmarking of the first GPU-based dose engine for helium ion therapy. The code has been proven to achieve high levels of accuracy and efficiency.