Narrowed coronary artery detection and classification using angiographic scans

Coronary artery disease is a disease in which a plaque material builds-up in the lumen of blood vessels which feed the heart muscle. This plaque material made of fat, cholesterol, calcium and other cellular components such as blood cells and smooth muscle cells. Egypt is ranked as the world 23^rd death rate due to heart disease [1]. Our research aims to recognize the narrowed Coronary artery from the angiographic scans image. Our approach first enhances Coronary Artery angiographic scans using filters, detects and segments these scans using region growing algorithm and then classifies these scans to normal and abnormal cases using the K-Nearest Neighbor classifier. Our approach successfully enhances the diagnosis of Coronary artery diseases by measuring the lumen diameter and detecting whether the artery is normal or abnormal. In abnormal cases, we successfully classify the blockage as minimal, mild, moderate or critical stages. Our system achieved 94.6% accuracy which proves the efficiency of our approach when compared with other approaches.