Optimization of Radiation Therapy Techniques for Improved Patient Outcomes

Abstract

The continual recommendation to improve treatment effectiveness and patient outcomes is one of the main aims of medical research and, hence, a culture of innovation is encouraged. The objective of this manuscript is to improve the outcomes of treatments through the implementation of biomimicry ideas during the scheduling of patients using radiation therapy (RT). RT has a crucial role in modern medicine as a resource to kill cancer cells and shrink tumour sizes. However, it takes a considerable amount of time and effort to manually schedule patients for RT. This paper is aiming to optimise patient scheduling for RT using optimisation techniques. Three approaches inspired by biological ideas is utilised to overcome the problem of optimising for online stochastic scheduling. They include GA, FFO and WO – the Genetic Algorithm, Firefly Optimisation and Wolf Optimisation. The implementation of these approaches can handle delicately complex online stochastic scheduling problems. In order to evaluate the approach for solving these kinds of challenging problems, we exhaustively tested these three algorithms to assess their relative efficacy. We used runtime, the objective values and the convergence time to arrive at the conclusion that bio-inspired algorithms were effective for optimising RT patient scheduling. In particular, out of the three approaches we have implemented, WO performed the best among all three algorithms and, in fact, was the best across all metrics that we have tested. The improvement in the outcomes of RT for patients through streamlining procedures, reducing manual intervention and other techniques is a great achievement of this manuscript’s optimisation technique.