How are rule-based models used in expert systems for medicine?

In the context of cognitive science, rule-based models represent a foundational approach to simulating human thinking by applying inference rules to complex symbols. In medical expert systems, these models function through two primary components: a knowledge base and an inference engine.12

The knowledge base is populated with facts acquired from human experts, typically structured as "if-then" production rules (e.g., "If some condition is true, then the following inference can be made"). In medicine, these rules allow the system to map specific symptoms to potential diseases. Because medical conclusions are rarely certain, these systems often attach probability factors to their recommendations, helping physicians evaluate the credibility of the output.23

The inference engine interprets these rules to perform tasks such as diagnosis. A notable historical example is MYCIN, which assisted physicians in diagnosing blood infections by demonstrating that knowing which symptoms correspond to specific diseases is often more critical than understanding the underlying etiology. By tracing the sequence of rules used to reach a conclusion, these systems also serve as valuable learning tools for students.23