COMET-PE as an Alternative to Monte Carlo for Photon and Electron Transport

Georgia Institute of Technology, Nuclear and Radiological Engineering and Medical Physics Programs, Atlanta, GA 30332, USA

^* Corresponding Author, E-mail: farzad@gatech.edu

Abstract

Monte Carlo methods are a central component of radiotherapy treatment planning, shielding design, detector modeling, and other applications. Long calculation times, however, can limit the usefulness of these purely stochastic methods. The coarse mesh method for photon and electron transport (COMET-PE) provides an attractive alternative. By combining stochastic pre-computation with a deterministic solver, COMET-PE achieves accuracy comparable to Monte Carlo methods in only a fraction of the time. The method’s implementation has been extended to 3D, and in this work, it is validated by comparison to DOSXYZnrc using a photon radiotherapy benchmark. The comparison demonstrates excellent agreement; of the voxels that received more than 10% of the maximum dose, over 97.3% pass a 2% / 2mm acceptance test and over 99.7% pass a 3% / 3mm test. Furthermore, the method is over an order of magnitude faster than DOSXYZnrc and is able to take advantage of both distributed-memory and shared-memory parallel architectures for increased performance.