Evaluation of a commercial VMC++ Monte Carlo based treatment planning system for electron beams using EGSnrc/BEAMnrc...

In the present work, Monte Carlo (MC) models of electron beams (energies 4, 12 and 18MeV) from an Elekta SL25 medical linear accelerator were simulated using EGSnrc/BEAMnrc user code. The calculated dose distributions were benchmarked by comparison with measurements made in a water phantom for a wide range of open field sizes and insert combinations, at a single source-to-surface distance (SSD) of 100cm. These BEAMnrc models were used to evaluate the accuracy of a commercial MC dose calculation engine for electron beam treatment planning (Oncentra MasterPlan Treament Planning System (OMTPS) version 1.4, Nucletron) for two energies, 4 and 12MeV. Output factors were furthermore measured in the water phantom and compared to BEAMnrc and OMTPS. The overall agreement between predicted and measured output factors was comparable for both BEAMnrc and OMTPS, except for a few asymmetric and/or small insert cutouts, where larger deviations between measurements and the values predicted from BEAMnrc as well as OMTPS computations were recorded. However, in the heterogeneous phantom, differences between BEAMnrc and measurements ranged from 0.5 to 2.0% between two ribs and 0.6-1.0% below the ribs, whereas the range difference between OMTPS and measurements was the same (0.5-4.0%) in both areas. With respect to output factors, the overall agreement between BEAMnrc and measurements was usually within 1.0% whereas differences up to nearly 3.0% were observed for OMTPS. This paper focuses on a comparison for clinical cases, including the effects of electron beam attenuations in a heterogeneous phantom. It, therefore, complements previously reported data (only based on measurements) in one other paper on commissioning of the VMC++ dose calculation engine. These results demonstrate that the VMC++ algorithm is more robust in predicting dose distribution than Pencil beam based algorithms for the electron beams investigated.