FDA Designated Breakthrough AI Analyzes Eye Scans to Predict Heart Disease Risk During Routine Optometry Appointments

New AI system CLAiR analyzes retinal images during routine eye exams to identify cardiovascular risk with 91% sensitivity, matching standard clinical tests.

By: AXL Media

Published: Mar 31, 2026, 4:15 AM EDT

Source: Information for this report was sourced from American College of Cardiology

The Retina as a Biological Window to Vascular Health

The medical community is increasingly viewing the human eye not just as an organ of sight, but as a noninvasive gateway to the body’s entire vascular system. At the American College of Cardiology’s Annual Scientific Session, researchers introduced an artificial intelligence platform capable of assessing a patient’s likelihood of a heart attack or stroke simply by scanning the retina. Because the back of the eye offers a high-resolution view of live blood vessels, it provides a literal window into the state of a patient's arterial health. According to lead author Dr. Michael V. McConnell, this technology addresses a critical gap in preventative care by reaching individuals who may not see a primary care physician regularly but do attend annual eye exams.

Automating Risk Assessment with Deep Learning

While previous medical efforts have used retinal images to monitor conditions like diabetes, those methods historically required specialized human interpretation, limiting their scalability. The new system, known as CLAiR, utilizes deep learning to automate the analysis of blood vessel patterns associated with atherosclerosis. The AI was trained to identify subtle structural markers that indicate a 7.5% or higher risk of a major cardiovascular event over the next decade. This specific threshold is the standard clinical benchmark used to determine if a patient should begin preventative therapies, such as statins. By automating this process, the system allows optometry clinics to provide sophisticated cardiac screening without specialized training.

Validating AI Accuracy Against Standard Clinical Metrics

To test the reliability of the AI, researchers conducted a prospective multi-center trial involving 874 participants across the United States. Each participant underwent traditional cardiovascular screening—including blood pressure and cholesterol tests—which was then compared to the AI’s retinal analysis. The results were highly consistent, with the CLAiR system identifying at-risk individuals with a sensitivity of 91.1% and a specificity of 86.2%. This suggests that the AI is nearly as effective as a comprehensive blood workup in flagging patients who require preventative cardiovascular care. The study included a diverse demographic pool, confirming the system's effectiveness across different ages, sexes, a...