Automated coronary artery disease detection using deep learning on ECG datasets

Abstract

Coronary artery disease is a common cardiovascular disease, usually caused by narrowing or occlusion of the coronary artery lumen, which can easily lead to heart failure or sudden death. Electrocardiogram (ECG) is one of the most commonly used diagnostic tools for cardiovascular diseases, and its non-invasive and inexpensive features have made it widely used in major hospitals and clinics around the world. The interpretation of ECG by physicians is time-consuming and requires appropriate expertise and experience. Reliable coronary heart disease detection models can be built using deep learning techniques. In this paper, we propose a feature fusion model that effectively utilizes features at different scales, and the model achieves 88.8% accuracy on the dataset provided by Fu Wai Hospital, and applies the class activation map (CAM) method for interpretable analysis. The proposed model can be used as a diagnostic aid for cardiologists, providing excellent support for coronary heart disease determination and prevention.