Deep Learning-Based Coronary Artery and Plaque Detection for Heart Disease

Abstract

Coronary artery disease (CAD) is a major cause of mortality worldwide, mainly due to plaque buildup in coronary arteries that restricts blood flow. Early detection is difficult in many clinical settings because it depends on expert analysis of complex CCTA images. To address this, this study proposes a deep learning–based system for automatic coronary artery segmentation and plaque detection. The model uses a U-Net architecture with a 2.5D approach to effectively capture spatial information and accurately identify artery and plaque regions. Hounsfield Unit (HU) analysis is applied to classify plaque types and estimate severity. The system achieves reliable performance and is suitable for real-time implementation. It can be integrated into a web-based application to assist healthcare professionals in faster and more accurate diagnosis.