Clinical diagnosis expert system based on dynamic uncertain causality graph

Abstract

Clinical diagnosis expert system is the focus and hotspots of research from the beginning of the 1960s, many inference techniques have been applied to disease diagnosis. Dynamic Uncertain Causality Graph (DUCG) is the model of graphical probability reasoning. It can represent the quantitative and qualitative causal knowledge by the way of causal graph and can reason in the case of incomplete knowledge. According to DUCG theory, we developed the clinical diagnosis expert system. Using the system, the clinical knowledge can be easily represented as a causal graph. The knowledge base construction can be done by more than one people separately. The consistence check of the so constructed knowledge base is encoded in this system. In the case of incomplete knowledge representation, this system still works well. The examples of hiatal hernia and infectious disease are provided, which demonstrates that our clinical diagnosis expert system is a powerful tool for the clinical diagnosis.