A Hybrid Method for Coronary Heart Disease Risk Prediction using Decision Tree and Multi Layer Perceptron

Background/Objectives: The diagnosis of Coronary Heart Disease (CHD) risk prediction is a vital and complicated job in medicine which is closely linked with lifestyle related behaviors. The main intention of this work is to build up a rapid and automatic prediction of CHD risk by integrating Decision Tree (DT) and Multi Layer Perceptron (MLP). Methods/Statistical Analysis: The proposed hybrid method consists of two stages in which risk identification is carried out in the first stage and level prediction is done in the second stage. In the first stage the physical and biochemical factors are classified using C4.5 algorithm in DT. In the second stage the CHD risk identified instances from DT are analyzed using MLP with habitation and medical history attributes. Findings: The obtained classification accuracies of this system are 98.66% for DT and 96.66% for MLP. The performance analysis of the proposed method is evaluated using sensitivity and specificity which helps to reduce the healthcare costs, further invasive CHD risk examination and waiting time of the individuals. Application/Improvements: The proposed work structured an optimal predictive tool for diagnosing CHD risk, which can further serve as an implication sketch for physicians in clinical diagnosis.