Applying Belief Rule-Based Inference Methodology Using Evidential Reasoning Approach to Clinical Reporting System

In the healthcare field, a critical issue is how to reason and represent uncertainties that present in clinical knowledge domain, signs, and symptoms to make the correct decisions. Although several researchers have developed various models of clinical modeling, many of them are incapable of handling uncertainties correctly. The paper provides the working details of rule-based inference methodology using evidence reasoning (RIMER) methodology applied to model the inference process and clinical guidelines. In RIMER, belief-degree are embedded in all possible consequences of the rule. It can handle uncertainties and provide a causal relationship between the rules. Traditional IF-THEN rules do not provide a causal relationship between antecedent and consequent attributes of the rule only provide that the rule is either 100% true or 100% false. Also, a case study is used to demonstrate that the results generated by the system using RIMER are more reliable in terms of accuracy and performance compared to results generated manually for a diabetes diagnosis.