Application of operations research to medical informatics and systems

The application of operations research to the medical field goes back to the late days of World War II and began to flourish in post-war Britain and during the Korean War. This dissertation presents several new applications, beginning with the use of neural networks for medical diagnosis prediction. The second application is a constraint programming approach to residential staffing requirements at Duke University. Due to the massive size of the problem, traditional optimization methods could not be used. The constraint programming/constraint satisfaction method employed succeeded where other methods fail.;The latter half of the dissertation is specifically focused on the use of the simulation in approaching the hybrid system in the medical field. Four specific examples, three from the United States Air Force 375th Medical Group, and one from the Missouri Baptist Medical Center, are treated. The systems are simulated and then optimized using the internal correlations of activities to patient time in clinic. A feedback control algorithm using simulation as the state transfer equation substitute is developed. Finally, a computer interface is designed so that the improvements can be implemented in a real time, human interactive system.