A Combined Belief Rule based Expert System to Predict Coronary Artery Disease

Coronary Artery Disease (CAD) is the clogging of the coronary arteries caused by cholesterol and fatty acid on the arteries. Millions of people die every year and the rate of dying is increasing alarmingly from cardiovascular disease all over the world including a huge portion in Bangladesh. Although a remarkable improvement has been done for heart related disease, still early detection is a big challenge for the treatment. In addition, Coronary angiography is a standard way for detecting CAD. Unfortunately, it is very expensive as it needs expensive instruments and CAD experts. Current BRB models transform input attributes without combining them. However, for some expert system, combine effort of the attributes is an obvious. In this paper, a BRB model with combining attributes is proposed for the clinical decision support systems. The proposed Combined Belief Rule Based Expert System (CBRBS) can predict the severity of the four classes of CAD. They are normal or no blockage in any artery, blockage in one artery, blockage/ stenosis in two artery, and Blockage or stenosis in 3 arteries. Success Rate, Error rate, Failure rate are measured for these four classes of CAD. The average success was 93.97% of the proposed system.