Abstract

Coronary artery disease (CAD) remains a leading cause of mortality worldwide, necessitating accurate and timely diagnostic strategies. This study proposes an enhanced one-dimensional convolutional neural network (1D-CNN) model for the automated detection of CAD using 12-lead electrocardiogram (ECG) signals. The model is trained and evaluated on the publicly available PTB-XL dataset, comprising over 21,000 annotated ECG records. To optimize classification performance, the model architecture incorporates 10-second signal segments, adaptive convolutional layers, and strategic dropout regularization. Extensive experiments demonstrate the model’s robust performance, including five-fold cross-validation and ablation studies. It achieves an average accuracy of 94.2%, precision of 93.1%, sensitivity of 92.7%, specificity of 95.4%, and an AUC-ROC of 96.1%. Comparative analysis with existing models confirms the superiority of the proposed approach in balancing diagnostic accuracy with computational efficiency. This work contributes a scalable and interpretable deep learning framework for CAD detection, offering promising implications for intelligent cardiovascular screening and clinical decision support systems.