Related Experiment Video
Updated: May 20, 2026

07:57
Positron Emission Tomography-based Dose Painting Radiation Therapy in a Glioblastoma Rat Model using the Small Animal Radiation Research Platform
Published on: March 24, 2022
Sub-second high dose rate brachytherapy Monte Carlo dose calculations with bGPUMCD
Sami Hissoiny1, Michel D'Amours, Benoıt Ozell
1École polytechnique de Montréal, Département de génie informatique et génie logiciel, Montréal, Québec H3T 1J4, Canada. sami.hissoiny@polymtl.ca
Medical Physics
|July 27, 2012
Summary
GPU-accelerated Monte Carlo code bGPUMCD offers fast and accurate dose calculations for high dose rate brachytherapy. It enables efficient plan optimization and recalculation in clinical settings.
Area of Science:
- Medical Physics
- Computational Science
- Radiation Oncology
Background:
- High dose rate (HDR) brachytherapy requires accurate and efficient dose calculations.
- Monte Carlo (MC) methods provide high accuracy but can be computationally intensive.
- GPU acceleration offers a potential solution for speeding up MC simulations.
Purpose of the Study:
- To evaluate the accuracy and speed of bGPUMCD, a GPU-based MC code for HDR brachytherapy.
- To assess the time for dose calculation using MC-generated dose distribution kernels for plan optimization.
- To determine the accuracy and speed of bGPUMCD for recalculating pre-optimized treatment plans.
Main Methods:
- bGPUMCD was tested using three clinical HDR brachytherapy cases: prostate, breast, and rectum with a shielded applicator.
- Reference dose distributions were generated using GEANT4 for comparison.
- Full dose distributions of pre-optimized plans and single dwell dosimetry were calculated.
- Single source dosimetry was performed based on TG-43 parameters for the microSelectron V2 source.
Main Results:
- Single dwell position dose kernels were computed in 0.25–0.5 seconds.
- Full dose distributions for pre-optimized plans were calculated in approximately 2 seconds.
- bGPUMCD achieved accuracy within 1–2% for key dosimetric parameters (D90, V100) compared to GEANT4.
- Average absolute differences in target voxels were approximately 1–2%, with larger discrepancies in bony regions.
- Single source dosimetry results were within 2% for radial and 1–4% for anisotropic functions.
Conclusions:
- bGPUMCD demonstrates significant potential for rapid MC dose calculations in HDR brachytherapy.
- The code supports all phases of treatment planning, including dose kernel generation and plan recalculation.
- bGPUMCD facilitates faster and more accurate dose assessments in a clinical environment.

