An assessment on the use of RadCalc to verify Raystation Electron Monte Carlo plans.

Large differences in monitor units have been observed when RadCalc, a pencil-beam-algorithm based software, is used to verify clinical electron plans from Raystation, a Monte-Carlo-algorithm based planning system. To investigate the problem, a number of clinical plans as well as test plans were created and calculated in both systems, with the resultant monitor units compared. The results revealed that differences between the two systems are significant when the geometry includes inhomogeneities and curved surfaces. The RadCalc pencil-beam-algorithm fails to handle such complexities, particularly in the presence of surface curvature. The error is not negligible and cannot be easily corrected for. It is concluded that RadCalc is not adequate to verify electron Monte Carlo plans from Raystation when complex geometry is involved and alternative methods should be developed.