Current progresses and issues in Monte Carlo methods of radiation dose calculations

Monte Carlo simulation of radiation transport is indisputably the most accurate means of predicting dose distributions and other quantities of interest in radiation treatment of patients. Unfortunately, due mainly to inordinately long computation times, it has hither-to-fore been impractical for all but investigational purposes. However, because of continuing advancements in computer technology, algorithms, and variance reduction techniques, it now appears that we are rapidly approaching a time when we can begin exploiting the benefits of Monte Carlo methodology routinely. In particular, the effort at LLNL, leading to the development of the Peregrine Monte Carlo code, has demonstrated that such is indeed the case. The Peregrine and other projects have generated great deal of interest and anticipation and has energized many others to initiate similar but not necessarily overlapping efforts. Examples of these efforts include work by Wang, et al. at MSKCC, Demarco, et al. at UCLA, Nahum, et al. at Royal Marsden, Mah at Stanford, Siebers, et al. at MCV, etc. As a consequence, many questions are being raised primarily about the necessity and potential of Monte Carlo technology, but also about how accurate it really is, what would it take to apply it clinically and make it available widely, when might all this happen, etc. In this paper, it is tried to answer many of these sometimes without explicitly stating them. Many other questions will be answered when Mote Carlo technology is implemented and used for more routine calculations and for in-depth investigations. It should stated that the discussion in this article deals mainly with external beam photon dose calculations although most of it can be generalized to other modalities as well