A Hybrid Inference Approach for Building Fuzzy DSSs Based on Clinical Guidelines

Clinical practice guidelines are expected to promote more consistent, effective, and efficient medical practices, especially if implemented in clinical Decision Support Systems (DSSs). With the goal of properly representing and efficiently handling clinical guidelines affected by uncertainty and inter-connected between them, this paper proposes a hybrid fuzzy inference approach for building fuzzy DSSs. It provides a set of specifically devised functionalities for best modeling and reasoning on the particular clinical knowledge underpinning guidelines: i) it organizes the whole fuzzy DSS into self-contained sub-systems which are able to independently reason on piece of knowledge according to their peculiar inference scheme; ii) a global inference scheme has been defined for handling and reasoning on such sub-systems, according to the classical crisp expert system approach. As a proof of concept, the proposed approach has been applied to a practical case, showing its capability of supporting multiple levels of inference and, thus, highlighting the possibility of being profitably used to model and reason on complex clinical guidelines in actual medical scenarios.