Personalized Medicine

This chapter introduces important concepts in personalized medicine (PM) and discusses the role that mathematical/computational/statistical models can play in the advancement of this field. By drawing examples from published literature, it illustrates the use of models in various problems related to PM. Many chronic disease prevention and management problems involve sequential decisions over a long time horizon. Instead of identifying the optimal solution based on a mathematical model that captures disease progression over time, artificial intelligence (AI)-based approaches learn the underlying connection between a patient's health information and decisions directly from existing data. One-time decision disease models aim to stratify patients into subgroups based on subtype testing and guide intervention decisions accordingly. Many health-related attributes of a patient can be used to define the health state in a model, such as demographic factors, disease stages, risk scores, and critical clinical measures, among others.