Deep-learning-based cardiovascular risk stratification using coronary artery calcium scores predicted from retinal photographs

AI Generated Summary*

A deep-learning model learned from retinal photographs taken at one Korean health-screening site to estimate whether CT-measured coronary artery calcium is present, then was tested in Korean, Singaporean and UK Biobank cohorts. In the second Korean screening set, the retinal score reached an AUC of 0.742, versus 0.705 for age. Among 527 high-risk Korean patients, a tertile-based three-tier version had a concordance index of 0.71, comparable to CT calcium categories. Risk-adjusted hazard ratio trends for fatal cardiovascular events were 1.33 in Singapore (n=8551) and 1.21 in the UK Biobank (n=47,679). Adding the score to the Pooled Cohort Equation in borderline and intermediate groups gave a continuous net reclassification index of 0.261.

*This summary was generated by AI and is published unedited. Oku does not alter these summaries. It may contain errors or omissions and is provided for general informational purposes only. Oku does not guarantee its accuracy, completeness, or reliability. For authoritative information, please refer to the original, peer-reviewed article.