Automated classification of coronary atherosclerosis using single lead ECG

Atherosclerosis in the coronary arteries represents a major medical burden throughout the world. While the disease may not manifest with symptoms at early stages, the high risk of heart attack and stroke at first presentation make early detection of this disease process critical. While the electrocardiogram (ECG) exercise stress test is commonly used to diagnose coronary artery atherosclerosis, it is an expensive test that requires cumbersome electrode placement with standardized exercise environment testing. We present an automated method for using a single lead of ECG sensors to classify subjects afflicted by atherosclerosis. This method is not only optimized for streamlined sensor implementation, but also automates classification to address data overload. Using the MIT-BIH database, the framework achieves high accuracy and diagnostic performance, supporting the clinical value of this novel classification method.