Using Deep Learning on Retinal Images to Predict Risk of Atherosclerotic Cardiovascular Disease

What this Trial Is About

Researchers are studying how deep learning algorithms can predict body measurements and cardiovascular disease risk using retinal fundus photographs. This observational study focuses on people with various cardiovascular conditions such as heart attack, heart failure, hypertension with organ damage, high-risk dyslipidemia, and diabetes. It aims to validate the ability of these algorithms to predict cardiovascular events over time. Participants will have fundus photographs taken twice during the first year and again two years later. The study will follow major cardiovascular events for five years to assess the prediction accuracy of the deep learning model. This study does not involve drug or device treatments but focuses on image analysis and risk prediction. During the study, participants will undergo retinal photography and clinical evaluations related to their cardiovascular health. The main outcome measured is the occurrence of major adverse cardiovascular disease events within four years. Researchers will monitor participants health data over several years to verify the algorithms predictive ability and collect follow-up information to support long-term analysis.

Deep Learning of Retinal Photographs and Atherosclerotic Cardiovascular Disease