EARLY PREDICTION OF CORONARY ARTERY DISEASE (CAD) USING MACHINE LEARNING

Coronary artery disease (CAD) remains a leading cause of mortality worldwide, with early detection being critical for effective intervention and management. Conventional diagnostic techniques, such as angiography, are invasive, costly, and often inaccessible in low-resource settings, highlighting the need for alternative approaches. This study develops an ensemblebased machine learning model that integrates Random Forest, Support Vector Machine (SVM), and Artificial Neural Network (ANN) classifiers through a soft-voting strategy to improve predictive accuracy for CAD. Using the Statlog and Cleveland heart disease datasets, the proposed ensemble achieved superior performance compared to individual models, with improvements observed in accuracy, sensitivity, specificity, and area under the curve (AUC). The model was subsequently deployed into a web-based decision support system (DSS) to provide an accessible, non-invasive, and cost-effective diagnostic aid, with particular focus on applicability within the Nigerian healthcare context. The system enables clinicians to input patient data and receive real-time risk predictions, thereby supporting early detection and informed clinical decision-making. This work contributes to bridging the gap between advanced machine learning methodologies and practical healthcare implementation in resource-constrained environments, offering a pathway toward scalable, data-driven diagnostic tools for cardiovascular disease management.