Personalized Cancer Chemotherapy Schedule: a numerical comparison of performance and robustness in model-based and model-free scheduling methodologies
This work addresses personalized chemotherapy scheduling for cancer patients, representing an incremental improvement by applying existing DRL methods to a known medical challenge.
The paper tackled the problem of designing personalized cancer chemotherapy schedules by comparing model-based optimal control with model-free deep reinforcement learning (DQN and DDPG). The result showed that both DRL methods outperformed optimal control under uncertainty, with DDPG being more efficient at exterminating cancer due to its continuous action space.
Reinforcement learning algorithms are gaining popularity in fields in which optimal scheduling is important, and oncology is not an exception. The complex and uncertain dynamics of cancer limit the performance of traditional model-based scheduling strategies like Optimal Control. Motivated by the recent success of model-free Deep Reinforcement Learning (DRL) in challenging control tasks and in the design of medical treatments, we use Deep Q-Network (DQN) and Deep Deterministic Policy Gradient (DDPG) to design a personalized cancer chemotherapy schedule. We show that both of them succeed in the task and outperform the Optimal Control solution in the presence of uncertainty. Furthermore, we show that DDPG can exterminate cancer more efficiently than DQN presumably due to its continuous action space. Finally, we provide some insight regarding the amount of samples required for the training.