Data Mining Approaches for Early Prediction of Cardiovascular Disease

Abstract: Cardiovascular disease (CVD) is a major global health challenge, contributing significantly to morbidity and mortality. With the continuous rise in incidence rates, there is an urgent need for advanced analytical methods to assist in early detection and diagnosis. This study explores the application of data mining techniques on a Transthoracic Echocardiography Report dataset to predict heart disease. Using the Knowledge Discovery in Databases (KDD) methodology, nine iterative steps were applied to process and analyze 7,339 echocardiography reports collected from a hospital. Three predictive models—J48 Decision Tree, Naïve Bayes, and Neural Network—were developed and evaluated. Experimental results indicate that all models achieved strong predictive performance, with the J48 Decision Tree yielding the highest classification accuracy of 95.56% and superior True Positive Rate. These outcomes demonstrate the potential of data mining-based approaches in enhancing diagnostic reliability and supporting cardiologists in clinical decision-making.

Keywords: Cardiovascular disease, Echocardiography, Data mining, Knowledge Discovery in Databases (KDD), Predictive modeling, Decision Tree, Naïve Bayes, Neural Network.