Evaluation of the EGSnrc Monte Carlo code for interface dosimetry near high-Z media exposed to kilovolt and Co-60 photons

High atomic number (Z) heterogeneities in tissue exposed to photons with energies of up to about 1 MeV can cause significant dose perturbations in their immediate vicinity. The recently released Monte Carlo (MC) code EGSnrc (Kawrakow 2000a Med. Phys. 27 485-98) was used to investigate the dose perturbation of high-Z heterogeneities in tissue in kilovolt (kV) and Co-60 photon beams. Simulations were performed of measurements with a dedicated thin-window parallel-plate ion chamber near a high-Z interface in a Co-60 photon beam (Nilsson et al 1992 Med. Phys. 19 1413-21). Good agreement was obtained between simulations and measurements for a detailed set of experiments in which the thickness of the ion chamber window, the thickness of the air gap between ion chamber and heterogeneity. the depth of the ion chamber in polystyrene and the material of the interface was varied. The EGSnrc code offers several improvements in the electron and photon production and transport algorithms over the older EGS4/PRESTA code (Nelson et al 1985 Stanford Linear Accelerator Center Report SLAC-265. Bielajew and Rogers 1987 Nucl. Instrum. Methods Phys. Res. B18 165-81). The influence of the new EGSnrc features was investigated for simulations of a planar slab of a high-Z medium embedded in water and exposed to kV or Co-60 photons, It was found that using the new electron transport algorithm in EGSnrc, including relativistic spin effects in elastic scattering, significantly affects the calculation of dose distribution near high-Z interfaces. The simulations were found to be independent of the maximum fractional electron energy loss per step (ESTEPE), which was often a cause for concern in older EGS4 simulations. Concerning the new features of the photon transport algorithm sampling of the photoelectron angular distribution was found to have a significant effect, whereas the effect of binding energies in Compton scatter was found to be negligible. A slight dose artefact very close to high-Z interfaces exposed to kilovolt x-rays was discovered when atomic relaxation processes following excitation were omitted