Impact of approximate reasoning on diagnosis in the EMERGE system

Computer-assisted medical decision support systems have been shown to be useful in a number of medical applications. One of the important theoretical advances in these systems has been the incorporation of methods for dealing with uncertain and imprecise information, both of which are common occurrences in medical applications. A number of techniques exist including fuzzy logic and fuzzy set theory, belief networks, Bayesian systems, and other techniques from approximate reasoning. In this article, EMERGE, an established medical decision support system, is analyzed in terms of the impact of using approximate reasoning in knowledge base development, user interaction, and decision strategy.