High-speed evaluation of track-structure Monte Carlo electron transport simulations

Area of Science:

• Medical Physics
• Computational Physics
• Radiation Dosimetry

Background:

• Monte Carlo simulation with the track-structure method is crucial for accurate dose estimation.
• This method is computationally intensive due to large electron interaction cross-sections and anisotropic scattering.
• Existing methods like condensed history (CH) may lack accuracy for specific applications.

Purpose of the Study:

• To evaluate the performance of field-programmable gate arrays (FPGAs) for track-structure Monte Carlo electron transport simulations.
• To compare the speed and accuracy of FPGA-based simulations against standard computer implementations.
• To demonstrate the utility of FPGAs in microdosimetry and radiation biology applications.

Main Methods:

• Implementation of track-structure Monte Carlo electron transport simulations on an FPGA.
• Testing the FPGA simulations across five diverse test cases, including complex biological geometries.
• Comparison of computational times and accuracy with standard computer simulations using the condensed history technique.

Main Results:

• FPGA-based track-structure simulations achieved speeds comparable to or exceeding standard computer condensed history simulations.
• For complex scenarios, FPGAs demonstrated over 500x speed improvement compared to standard track-structure simulations.
• The accuracy of FPGA simulations was found to be comparable to traditional methods.

Conclusions:

• FPGAs offer a significant computational advantage for track-structure Monte Carlo electron transport simulations.
• This technology can enable faster and more accurate dose estimations in fields requiring detailed electron transport analysis.
• FPGA acceleration is particularly beneficial for microdosimetry, radiation biology, and simulations involving complex material interfaces or low electron energies.