Po‐Poster ‐ 08: Comparison of two methods for commissioning BEAMnrc Elekta SL25 accelerator model

Abstract

Incident electron beam parameters of Monte Carlo linear accelerator model are typically determined by comparing computed profiles and depth‐dose curves with measurements. Several variations have been suggested in literature: Sheikh 2002 used kerma equivalent off‐axis factors while van der Zee 1999 used cross profiles in large water phantom. In this study, we compared both methods. The BEAMnrc and DOSXYZnrc codes are used to simulate 6MV and 25MV photon beam from medical linear accelerator Elekta SL25. The mean energy and radial width of the incident electron beam intensity distributions (assumed Gaussian and cylindrically symmetric) are derived by both in air and in water technique. Excluding region of electronic disequilibrium (build‐up and penumbra) local dose difference between simulated and measured curves must be under 2% to found agreement. The statistical uncertainties on calculated dose at each point is about 0.7% (or better) at 1 sigma level. We show that off‐axis curves are more sensitive to mean energy of electron beam than depth‐dose curves. The matching mean energy and radial parameters of the incident electrons beam (6MV & 25MV) for both in air and in water technique converge within 6%. Particular attention is also devoted to measurement in air and their consequences on matching parameters.