A hybrid method to determine of an optimal control strategy for drug administration in tumor treatment

In this study, we present an optimal control strategy for drug administration to patients with cancer, through the minimization of both the cancerous cells concentration and the prescribed drug concentration. There are quite a number of modern optimization algorithms proposed in the last two decades to solve optimization problems. We suggest and investigate a hybrid of gradient-based information, as a means to improve the search process performed by a multi-objective evolutionary algorithm (MOEA). To show our proposed coupling, the widely used None-Dominated Genetic Algorithm II (NSGA-II) as our global search engine hybridized with Local Search (LS) algorithm represent by gradient information. This is an attempt to propose an adaptive mechanism that allows the local and global search strategies to dynamically interleave. The proposed algorithm is tested to solve multi-objective optimization problem. Satisfactory results obtained in the tests show that hybrid approach (HGA) can effectively balance searching ability of global exploitation and local exploration. The Pareto’s Curve obtained supplies a set of optimal protocols from which an optimal strategy for drug administration can be chosen, according to a given criterion