Medical Diagnosis of Coronary Artery Disease Using Fuzzy Rule-Based Classification Approach

Abstract

Coronary Artery Disease (CAD) is one of the leading causes of morbidity and mortality worldwide including India. Although recent advances in modern medical science have led to better diagnosis and treatment of CAD, yet its early detection is still a challenge. Fuzzy classification approaches are used to deal with uncertainty inherent in medical field. These fuzzy rule-based systems are extremely effective tools in disease diagnosis as they are capable to develop potential linguistic models. The aim of this paper is to initially develop a fuzzy rule-based classification system (FRBCS) based on clinical and epidemiological variables of patients and then to determine its accuracy in the diagnosis of CAD. The membership functions for medical attributes were chosen after extensive review of related literature. The rules were formulated as per the opinion of expert physicians. The present work describes the risk factors accountable for CAD, fuzzy modeling of clinical variables, rule evaluation and defuzzification of the fuzzified outputs to crisp values. The accuracy of the proposed fuzzy if–then rule classification system is 89%. Further, the present approach can assist medical practitioners in diagnosing CAD more precisely based on the fuzzy rules.