Predicting Coronary Artery Disease Using Machine Learning

Abstract

Developing a predictive model for detecting Coronary Artery Disease (CAD) is crucial due to its high global fatality rate of approximately 17.9 million people annually. With the advancements in artificial intelligence, the availability of large-scale data, and increased access to computational capability, it is feasible to create robust models that can detect CAD with high precision. This study aims to build a predictive model that can assist health workers in the timely detection of CAD and ultimately reduce mortality. This study performs a comparative analysis of four supervised classification machine learning algorithms- Logistic regression (LR), Support vector machine (SVM), Extreme gradient boosting (XGBoost), and Artificial neural network (ANN),  in predicting the case-control status of the patient. Chi-squared and lasso criteria are employed to select the most relevant ones from the available features. The performance of the employed models is compared using sensitivity, specificity, accuracy, and the area under the receiver operating characteristic (ROC) curve (AUC). The experimental results indicate that the LR model is the most effective and accurate among the models tested, and its implementation can improve the detection of CAD in clinical settings.