Diagnosis of Coronary Artery Disease via a Novel Fuzzy Expert System Optimized by Cuckoo Search

In this paper, we propose a novel fuzzy expert system for detection of Coronary Artery Disease, using cuckoo search algorithm. This system includes three phases: firstly, at the stage of fuzzy system design, a decision tree is used to extract if-then rules which provide the crisp rules required for Coronary Artery Disease detection. Secondly, the fuzzy system is formed by setting the intervals for fuzzy variables and extracted rules. Finally, Cuckoo Search algorithm is used to optimize fuzzy membership functions. The accuracy of our proposed system is evaluated using Cleveland Cardiac Patient Database. The detection rate is 93.48% employing optimized membership functions. Also, 85.76% accuracy is obtained for predicting the risk of coronary artery disease. The superiority of proposed system is obvious by comparing it to the previously methods; it is more accurate and is also easier to implement.