Heart Attack Risk Prediction Using Retinal Eye Images
Ramaraju M,
Keerthana G,
Laxman A,
Sairam M,
Sruthi G
Early detection of heart attack risk is essential for improving patient outcomes in cardiovascular
healthcare, as many conventional diagnostic methods rely on invasive procedures, expensive
laboratory tests, or limited clinical accessibility. This project presents an innovative and completely
non-invasive approach that utilizes retinal eye (fundus) images to predict heart attack risk through
advanced machine learning and image processing techniques. The retina provides a unique window
into systemic vascular health because retinal blood vessels reflect microvascular changes associated
with hypertension, arteriosclerosis, and other cardiovascular disorders. By analyzing visible alterations
in retinal vasculature, the system extracts clinically significant biomarkers such as vessel narrowing,
increased tortuosity, arteriovenous ratio variations, and microvascular abnormalities, which serve as
early indicators of underlying heart disease.
