The macro response Monte Carlo method for electron transport

Abstract

This thesis demonstrates the feasibility of basing dose calculations for electrons in radiotherapy on first-principles single scatter physics, in a calculation time that is comparable to or better than current electron Monte Carlo methods. The macro response Monte Carlo (MRMC) method achieves run times that have potential to be much faster than conventional electron transport methods such as condensed history. The problem is broken down into two separate transport calculations. The first stage is a local, single scatter calculation, which generates probability distribution functions (PDFs) to describe the electron's energy, position, and trajectory after leaving the local geometry, a small sphere or “kugel.” A number of local kugel calculations were run for calcium and carbon, creating a library of kugel data sets over a range of incident energies (0.25–8 MeV) and sizes (0.025 to 0.1 cm in radius). The second transport stage is a global calculation, in which steps that conform to the size of the kugels in the library are taken through the geometry of a phantom. For each step, the appropriate PDFs from the MRMC library are sampled to determine the electron's new energy, position, and trajectory. The electron is immediately advanced to the end of the step and then another kugel is sampled. The process continues in this manner until transport is completed. The MRMC global stepping code was benchmarked as a series of subroutines inside of the Peregrine Monte Carlo code against EGS4 and MCNP for depth dose in simple phantoms having density inhomogeneities. The energy deposition algorithms for spreading dose across 5–10 transport zones per kugel were tested. Most resulting depth dose calculations were within 2%–3% of well-benchmarked Monte Carlo codes.