design of an optimal controller using optimization techniques

Abstract

newline Cancer in one of the deadliest disease today, and with a fast-growing aging society, it poses a significant current and future concern. Fortunately, developments in technology and science are consistenly contributing to a deduction in cancer mortifity and treatment adverse effects. The Prime objective of this work is to contribute to these advancements by achieving an optimal reduction in the growth of the tumor cells with minimum side effects, thereby improving the prospects of future cancer patients survival and well-being. newline newlineThe thesis focuses on the advancement of radiation therapy, which is used is both curative and palliative cancer treatments. Lonizing is used to target malignant cells in the body during radiation therapy. Radiation damages cancerous cells DNA, preventing them from proliferating further. Radiation, on the other hand, is detrimental to healthy cells, so it is essential that the patients s irradiation be conducted in such a manner that the ceitical oragans in the proximity of the tumor are protected from radiation. newline newlineThe opposing goals of enchacing tumor control and reducing the consequences for normal tissue are at the heart of all cancer therepy procedures. As a result, many components of the cancer therapy planning process may be formulated as optimization problems, leading to the formation of mathematically sound treatment planning algorithms. newline newlineIn the dissertation, we provide three unique optimization approaches to solve the challenges in cancer treatment olanning: 1) Fluence map optimization for inverse treatment planning, which is well suited to dealing with nonconvex dose-volume constraints. 2) Optimal delivery of the fluence map via leaf trajectory optimization, which has the potential to improve treatment plan quality while reducing the time needed to create a clinically acceptable treatment plan; and 3) A multi-therapy treatment pan scheduling using a Markov decision process that balances the treade-off between tumor control and normal-tissue complications, as oncolo