Simulation of the transmitted dose in an EPID using a Monte Carlo method.

Abstract

The BEAMnrc and DOSXYZnrc codes from EGSnrc Monte Carlo (MC) system are considered to be the gold standards for simulating radiotherapy linear accelerators and resulting dose depositions (Rogers, Faddegon et al. 1995). The aim of this project was to setup the EGSnrc system for the simulation of the linear accelerator (linac) head and a Scanning Liquid Ionisation Chamber (SLIC) Electronic Portal Imaging Device (EPID) for calculations of transmitted dose in the EPID. The project was divided into two parts. The head of a 6 MV Varian 600C/D photon linac was first simulated by BEAMnrc. The modelling parameters such as the electron beam energy and the Full Width at Half Maximum (FWHM) of the electron spatial distribution were adjusted until the absorbed dose profiles and the Percentage Depth Dose (PDD) curves, in general agreed better than the measured profiles and PDDs by 2%. The X-ray beam obtained from the modelled linac head was used for the simulation of the transmitted dose in the EPID in the second part of the project. The EPID was simulated by DOSXYZnrc based on the information obtained from Spezi and Lewis 2002 (Spezi and Lewis 2002), who also modelled the Varian SLIC EPID (MK2 Portal Vision system, Varian Inc., Palo Alto, CA, USA). The comparisons between the measured and the simulated transmitted doses were carried out for three different phantom setups consisting of an open field, homogeneous water equivalent phantom and a humanoid phantom (RANDO). These phantom setups were designed so that the accuracy of the MC method for simulating absorbed dose in air, homogeneous and inhomogeneous phantoms could be assessed. In addition, the simulated transmitted dose in an EPID was also compared with values obtained from the Pinnacle treatment planning system (v6.2b, Phillips Medical Systems). In the process of selecting the electron beam energy and FWHM, it was confirmed (Sheikh-Bagheri and Rogers 2002; Keall, Siebers et al. 2003) that the variation of the electron beam FWHM and energy influenced the beam profiles strongly. The PDD was influenced by the electron beam energy less strongly. The increase in the energy led to the increase in the depth of maximum dose. However, the effect could not be observed until the energy change of 0.2 MeV was made. Based on the analysis of the results, it was found that the combination of FWHM and energy of 1.3 mm and 5.7 MeV provided the best match between the measured and MC simulated beam profiles and PDDs. It can be concluded that an accuracy of 1.5% can be achieved in the simulation of the linac head using Monte Carlo method. In the comparison between the Monte Carlo and the measured transmitted dose maps, agreements of 2% were found for both the open field and homogeneous water equivalent phantom setups. The same agreements were also found for the comparison between Monte Carlo and Pinnacle transmitted dose maps for these setups. In the setup where the humanoid phantom RANDO was introduced in between the radiation field and the EPID, a general agreement of about 5% found for the comparison between Monte Carlo and measured transmitted dose maps. Pinnacle and measured transmitted dose map was also compared for this setup and the same agreement was found.