A GPU-accelerated Monte Carlo dose calculation platform and its application toward validating an MRI-guided radiation
Yuhe Wang1, Thomas R Mazur1, Olga Green1
1Department of Radiation Oncology, Washington University School of Medicine, 4921 Parkview Place, Campus Box 8224, St. Louis, Missouri 63110.
Medical Physics
|July 3, 2016
Summary
A new GPU-accelerated Monte Carlo platform (gPenelope) was developed for magnetic field IMRT. It validated the MRIdian head model, ensuring accurate dose calculations for clinical treatment plans.
Area of Science:
- Medical Physics
- Computational Biology
- Radiotherapy
Background:
- Clinical commissioning of Intensity-Modulated Radiation Therapy (IMRT) in magnetic fields presents significant challenges.
- Accurate dose calculation is crucial for quality assurance in advanced radiotherapy techniques.
Purpose of the Study:
- To develop a GPU-accelerated Monte Carlo dose calculation platform (gPenelope) based on the Penelo_pe code.
- To validate the vendor-provided MRIdian head model for clinical IMRT treatment plans within a 0.35 T magnetic field.
Main Methods:
- Translated Penelo_pe from Fortran to C++, then adapted to CUDA for GPU optimization.
- Incorporated voxelized transport, Woodcock tracking, and enhanced electron/positron propagation in magnetic fields.
- Integrated the MRIdian head model and experimentally validated gPenelope against the MRIdian system.
Main Results:
- Achieved a 152x acceleration factor over the original code while preserving accuracy.
- MRIdian dose calculations agreed with gPenelope for 16 diverse treatment plans, with a mean gamma passing rate of 99.1% ± 0.6% (2%/2 mm).
Conclusions:
- Developed an accurate and rapid GPU-accelerated Monte Carlo simulation platform (gPenelope).
- Validated the MRIdian head model and its fast Monte Carlo engine for accurate radiation transport modeling.
- gPenelope platform is suitable for future applications in dose validation, IMRT optimization, and dosimetry for MR-IGRT systems.
