Feasibility of an Ant Colony Optimization Algorithm for Multi-leaf Collimator (MLC) Aperture Definition and Beam Weighting in Volumetric Modulated Arc Therapy (VMAT) Radiotherapy Treatment Planning

Abstract

Volumetric Modulated Arc Therapy (VMAT) is a sophisticated radiotherapy treatment delivery modality in which a medical linear accelerator arcs around a patient, with concurrent dynamic variation of multi-leaf collimator aperture, dose rate, and gantry speed, to produce a radiation dose distribution which delivers a highly conformal dose to the target while minimizing the incidental irradiation of normal tissue. Treatment planning for VMAT is an inverse problem, requiring optimization of the linear accelerator parameters to produce the desired radiation dose distribution, which is specified by dose-volume objectives. In this study, the feasibility of an ant colony algorithm for VMAT treatment planning is demonstrated by the ability of the algorithm to produce a treatment plan, which satisfies given dose-volume objectives, for a phantom target/critical structure geometry. Three experiments were conducted: one in which the optimization included only heuristic information, one in which there was exclusively a pheromone trail update, and one where there was both a pheromone trail update and an applied heuristic. The results indicate that the use of both a pheromone trail update and heuristic information during the optimization yields solutions of the highest quality.