Abstract
There is a growing need for physicians to select the appropriate medical activity for diagnosis and treatment. However, many difficulties are encountered during the course of their decision making because of the uncertainty in medical information. The purpose of this study was to develop a decision making system to deal with the inherent fuzziness of clinical data. This system is based on the expected utility maximization, of which the probabilities and utilities are expressed as the membership function of fuzzy set theory. We first calculated fuzzy expected utility on the basis of these probabilities and utilities in each chance node. Then, the truth value of the linguistic preference function was calculated in the decision node. This system easily dealt with the fuzziness of clinical data and expressed the result of the decision making process with a linguistic truth value.
