Intelligent Patient Management using Dynamic Models of Clinical Variables

Abstract

The ability to model and predict the progression of disease in a patient can have wide ranging benefits, including the ability to successfully manage bed allocation in hospitals or the increase understanding of the evolution of the disease. This paper describes a new method of modelling the progression of a disease through different stages called a Coxian hidden Markov model. This model can be used to increase understanding of the characteristics of the different stages of the disease and to predict patient survival time given repeated measurements of dynamically changing clinical variables. This knowledge could then be used to provide better patient management