A Treatment Decision Modeling and Optimizing Technology

描述（由申请人提供）：

DESCRIPTION (provided by applicant): A novel disease-independent decision-making and optimization technology is proposed by combining fuzzy systems technology with discrete event systems technology. The former technology is essential because, first, many symptoms, diagnoses, and treatment outcomes are not only patient-dependent, but also imprecise, immeasurable, and subjective with uncertainty, and, second, a clinician's subjective knowledge and experience must be utilized. Discrete event systems technology is most suitable for handling complex systems with numerous components and states. The proposed new technology is intrinsically flexible, dynamic, and efficient due to its unique ability to compactly represent, for any disease, a large amount of complex information on patient and treatment in component models and then use them to automatically synthesize an overall treatment model for optimization. This modular, bottom-up paradigm ensures the structural and design similarity of decision systems for different diseases. Furthermore, it permits easy knowledge upgrade and fast treatment strategy evolution. Our preliminary study has established part of the theoretical foundation for the proposed technology. In this three-year pilot phase project, the feasibility, principle and structure for making treatment decision and optimization for diseases in general will be established. To validate and refine the theoretical and structural development of the proposed technology and determine its clinical utility, the established structure will be used as a system template for implementing an HIV/AIDS treatment decision system. The system will be tested retrospectively using our HIV/AIDS database containing 2,200 patients. Statistical methods will be used to assess the preliminary accuracy and utility of the system. At the end of the project, two software packages running in MS Windows will be available to the scientific community through the Internet for research purposes. One will be a generic treatment decision system that is customizable for a specific disease. The other will be for HIV/AIDS treatment decision-making. The technology and software developed in this project will lay the foundation for more rigorous and larger scale development of treatment decision systems for many diseases in the future. Once fully developed, the technology can produce optimized treatment decisions for any given patient. It can be used to objectively and quantitatively analyze and compare different treatment options and selections, either prospectively or retrospectively.