Eye Imaging AI Identifies Elevated Cardiovascular Risk

Many adults at risk for atherosclerotic cardiovascular disease are not identified until they undergo formal primary care assessment. Delayed risk recognition can postpone initiation of statins and lifestyle counseling. A newly introduced technology aims to use routine eye examinations to surface patients who likely meet preventive treatment thresholds. Researchers now report that artificial intelligence applied to retinal photographs can approximate standard cardiovascular risk stratification and prompt referral for guideline-based evaluation.

The system, called CLAiR, was developed by Toku and has received Breakthrough Device designation from the U.S. Food and Drug Administration (FDA). It was evaluated in the United States in a prospective study presented at the American College of Cardiology’s Annual Scientific Session (ACC.26). Investigators positioned the tool as an adjunct that could expand opportunistic screening in eye care settings.

CLAiR analyzes standard fundus photographs to assess the retinal vasculature. The algorithm was trained to recognize patterns in blood vessel appearance that are associated with the development of heart disease. During a clinic visit, the software classifies patients whose estimated 10-year risk of heart attack or stroke is 7.5% or higher, mirroring the threshold commonly used to consider statin therapy with the pooled cohort atherosclerotic cardiovascular disease (ASCVD) risk estimator.

The study enrolled 874 adults aged 40–75 years from 10 eye care and primary care sites across the United States. Participants were not taking lipid‑lowering medications and had no known atherosclerosis; half were female, 19% were Black or African American, and 26% were Hispanic. Compared with the standard ASCVD risk estimator, CLAiR identified at‑risk participants with 91.1% sensitivity and 86.2% specificity. Overall, 26% of participants met the ≥7.5% 10‑year risk category by standard assessment.

Operational metrics favored clinic integration. Retinal imaging required about five minutes, and CLAiR returned results in about 30 seconds. Ninety‑four percent of acquired images were analyzable. The system is not designed for people who are pregnant or those with advanced eye disease, and retinal imaging may not be covered by all vision insurance plans. Investigators noted the need to formalize referral pathways from eye clinics to primary care for confirmatory evaluation and treatment.

“Even just a standard retinal photo provides high resolution imaging of your blood vessels—it's a literal window into vascular tissue,” said Michael V. McConnell, MD, clinical professor of medicine at Stanford University and chief health officer at Toku.

“This approach would not replace the standard cardiovascular risk evaluation, but it's a potential way to bring greater awareness, especially for people who should be on preventive care, but who have not yet had a thorough evaluation,” said McConnell. “For patients to benefit, we need to implement clear pathways to connect your elevated risk from your eye exam to help you see your clinician and ultimately get guideline-based preventive therapy.”