In brachytherapy, plans are routinely calculated using the AAPM TG43 formalism which considers the patient as a simple water object. An accurate modeling of the physical processes considering patient heterogeneity using Monte Carlo simulation (MCS) methods is currently too time-consuming and computationally demanding to be routinely used. In this work we implemented and evaluated an accurate and fast MCS on Graphics Processing Units (GPU) for brachytherapy low dose rate (LDR) applications. A previously proposed Geant4 based MCS framework implemented on GPU (GGEMS) was extended to include a hybrid GPU navigator, allowing navigation within voxelized patient specific images and analytically modeled (125)I seeds used in LDR brachytherapy. In addition, dose scoring based on track length estimator including uncertainty calculations was incorporated. The implemented GGEMS-brachy platform was validated using a comparison with Geant4 simulations and reference datasets. Finally, a comparative dosimetry study based on the current clinical standard (TG43) and the proposed platform was performed on twelve prostate cancer patients undergoing LDR brachytherapy. Considering patient 3D CT volumes of 400 × 250 × 65 voxels and an average of 58 implanted seeds, the mean patient dosimetry study run time for a 2% dose uncertainty was 9.35 s (≈500 ms 10(-6) simulated particles) and 2.5 s when using one and four GPUs, respectively. The performance of the proposed GGEMS-brachy platform allows envisaging the use of Monte Carlo simulation based dosimetry studies in brachytherapy compatible with clinical practice. Although the proposed platform was evaluated for prostate cancer, it is equally applicable to other LDR brachytherapy clinical applications. Future extensions will allow its application in high dose rate brachytherapy applications.
Monte Carlo simulations (MCS) applied in particle physics play a key role in medical imaging and particle therapy. In such simulations, particles are transported through voxelized phantoms derived from predominantly patient CT images. However, such voxelized object representation limits the incorporation of fine elements, such as artificial implants from CAD modeling or anatomical and functional details extracted from other imaging modalities. In this work we propose a new hYbrid Voxelized/ANalytical primitive (YVAN) that combines both voxelized and analytical object descriptions within the same MCS, without the need to simultaneously run two parallel simulations, which is the current gold standard methodology. Given that YVAN is simply a new primitive object, it does not require any modifications on the underlying MC navigation code. The new proposed primitive was assessed through a first simple MCS. Results from the YVAN primitive were compared against an MCS using a pure analytical geometry and the layer mass geometry concept. A perfect agreement was found between these simulations, leading to the conclusion that the new hybrid primitive is able to accurately and efficiently handle phantoms defined by a mixture of voxelized and analytical objects. In addition, two application-based evaluation studies in coronary angiography and intra-operative radiotherapy showed that the use of YVAN was 6.5% and 12.2% faster than the layered mass geometry method, respectively, without any associated loss of accuracy. However, the simplification advantages and differences in computational time improvements obtained with YVAN depend on the relative proportion of the analytical and voxelized structures used in the simulation as well as the size and number of triangles used in the description of the analytical object meshes.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.