Advances and Applications in Statistics
Volume 9, Issue 2, Pages 205 - 232
(August 2008)
|
|
THE PARTIALLY OBSERVABLE MARKOV DECISION PROCESSES FRAMEWORK IN MEDICAL DECISION MAKING
John E. Goulionis (Greece) and Dimitrios I. Stengos (Greece)
|
Abstract: Acquired Immune Deficiency Syndrome (AIDS) and its cause, the Human Immunodeficiency Virus (HIV) are major healthcare problems throughout the world. For many patients, advances in therapies over the past ten years have changed HIV form from a fatal disease to a chronic. Because of the development of resistance to administered therapies over time, an extension to basic question (when a patient should initiate a therapy) arises: when should a patient switch therapies. We consider the switching and sequencing problem for a patient with (HIV). This problem formulated as a partially observable Markov decision (POMDP) model. The generality of the standard POMDP model, however, limits practical application of the framework due to the computational complexity of associated solution methods. However, for many specialized problems, the full-blown generality of the POMDP approach and its associated solution methods is superfluous. This holds in particular for decision problems, where often the class of admissible solutions is significantly constrained. In this paper, we discuss a specialization of POMDPs that is tailored to a frequently reoccurring type of management problem, and propose a simple solution method that is able to exploit the properties of the specialized form of the patients with (HIV). |
Keywords and phrases: OR in medicine, Markov processes, dynamic programming. |
|
Number of Downloads: 353 | Number of Views: 1200 |
|