Implementation of convolution/superposition model of photon dose calculation

Abstract

The implementation, accuracy and performance of a collapsed cone convolution/superposition dose calculation algorithms is presented. The primary fluence or TERMA (total energy released per unit mass) is calculated using a polyenergetic spectrum which is modified with depth to account for water effective depth hardening and with off axis position to account for softening of the beam. An Inverse square law correction for kernel tilting is applied at the dose deposition site The algorithm uses energy deposited kernels generated by use of EGSnrc Monte Carle Code. In addition, this work also reports on some of the initial test that was conducted during the commissioning of a commercially available 3D treatment planning system. And the algorithm was verified by using AAPM phantom data with open, unwedged, and tilting fields for both symmetric and asymmetric collimator settings with variant energy (4MV, 18MV). Also measured data from variant linac and practical clinical data are used for verifying applicability and feasibility of the algorithm. Results presented in this paper show good agreement between measured data and calculated results in homogeneous and homogeneous media for different fields, and demonstrate the correct behaviour of collapsed cone convolution/superposition dose calculation algorithms. © 2008 Springer-Verlag.