Coronary artery disease is one biggest cause of death in the world. The normal way to diagnose it is with invasive coronary angiography. But this method is not perfect. It is expensive, has risks for the patient, an‎d the analysis of the images by doctors can be subjective an‎d take long time. Because of this, new automatic methods are needed. This seminar explores the integration of artificial intelligence (AI) an‎d deep learning (DL) techniques for automated CAD detection from medical imaging, with a focus on X-ray coronary angiography. State-of-the-art deep learning models, including U-Net, ResNet, an‎d hybrid architectures, are examined for tasks such as vessel segmentation an‎d stenosis detection. Multiple datasets (e.g., ARCADE, CADICA, ASOCA, ImageCAS, RCT QCA) are analyzed with respect to their scope, strengths, an‎d limitations. The methodology emphasizes preprocessing techniques (normalization, histogram equalization, data augmentation) an‎d transfer learning strategies to enhance diagnostic performance. Comparative eva‎luation highlights the potential of AI to surpass traditional approaches in accuracy, efficiency, an‎d reproducibility, while addressing challenges such as data scarcity, noise, an‎d class imbalance. The seminar underscores that DL-based CAD diagnostics can provide robust, objective, an‎d scalable support for clinicians, paving the way toward faster, safer, an‎d more equitable cardiovascular care.

5/26/2026 12:00:00 AM

90497

1405/03/10

Coronary Artery Disease Diagnostics from images

بيماري عروق كرونر قلب (CAD) , يادگيري عميق (DL) , شبكه‌هاي عصبي كانولوشن (CNN) , تحليل تصاوير پزشكي , كرونري خودكار

Coronary Artery Disease (CAD) , Deep Learning(DL) , Convolutional Neural Networks (CNN) , Medical Image Analysis , Automated Coronary