Technical note: MC-GPU breast dosimetry validations with other Monte Carlo codes and phase space file implementation.

Area of Science:

• Medical Physics
• Computational Imaging
• Radiological Sciences

Background:

• Accurate dose calculation is crucial for optimizing breast imaging techniques.
• Monte Carlo (MC) simulations are essential for precise dosimetry but can be computationally intensive.

Purpose of the Study:

• Validate the MC-GPU code for dosimetry in mammography, digital breast tomosynthesis, contrast-enhanced digital mammography, and breast-CT.
• Implement and validate a phase space file generation routine within MC-GPU.

Main Methods:

• Modified MC-GPU (v. 1.5 DBT) to generate phase space files and use PENELOPE v. 2018 cross-sections.
• Performed simulations using homogeneous and anthropomorphic breast phantoms for various breast imaging techniques.
• Compared MC-GPU glandular dose calculations with PENELOPE (v. 2014) + penEasy (v. 2015) and EGSnrc MC codes.

Main Results:

• MC-GPU demonstrated good agreement with other MC codes for mammography, digital breast tomosynthesis, and breast-CT dosimetry.
• Photon spectra entering voxels were similar across MC codes within statistical uncertainties.
• Dose reconstruction from MC-GPU phase space files showed excellent agreement (<0.65% difference) with original MC-GPU scores.
• MC-GPU simulations were 20-40 times faster than PENELOPE simulations.

Conclusions:

• MC-GPU is suitable for breast dosimetric studies across multiple X-ray imaging modalities.
• The validated phase space file implementation enables multiscale MC simulations using combined CPU and GPU codes.
• MC-GPU offers significant performance advantages due to GPU acceleration.