A Control System Approach to the Management of Thyroid Disease

Abstract A control system approach to health care is developed, illustrated in the context of thyroid disease with the emphasis being placed upon patient management. This approach requires a compatible information handling system and the implementation of an improved information system in the clinical environment is described. Diagnostic test selection is then considered. Applying multi-variate analysis including feature space techniques of pattern recognition, optimal sub-sets of clinical signs and laboratory tests are identified so as to establish an appropriate optimal diagnostic strategy in terms of best prediction of thyroid disease states. With time series data for a particular patient of diagnosis and patient state estimate, models are being developed relating treatment to change of patient state. Such models provide a basis for suggesting an adaptive control strategy for improved patient management.