A matheuristic for the selection of beam directions and dose distribution in radiotherapy planning

Abstract

In this paper a matheuristic using a combined Genetic Algorithm (GA) and exact method approach is proposed for selecting the position of the beams and dose distribution in Intensity Modulated Radiotherapy Planning (IMRT). GA selects a set of beams, for which the dose distribution is determined in the process of the GA’s evaluation, using an optimisation model that is solved by an Interior Point method. Two instances are used to evaluate the performance of the matheuristic, comparing to the optimum solution, in terms of solution and computation time, found using the exact methodology of Branch and Bound. The results show that the matheuristic is appropriate to this problem in the case study proposed, as it is extremely faster than the exact method and also have reached the optimum solution in several experiments done.