A New Paradigm for Explanation in Medical Causal Reasoning Systems

A significant obstacle confronting medical expert systems that employ causal reasoning is providing explanations without prohibitive computational expense. Another problem is the construction of a suitable user interface for the explanatory system. The interface should allow a physician to explore in detail the explanation of some concepts while avoiding lengthy explanations of lesser concern. A new method has been devised to deal with these problems by flexibly linking explanations with conclusions generated by a causal reasoning system. This approach allows the reasoning module to structure explanations without excessive computational overhead. It also allows the user to obtain additional detail of a causal explanation.