Insights into systemic disease through retinal imaging-based oculomics

AI Generated Summary*

Review of evidence that retinal imaging yields biomarkers of systemic disease, termed oculomics, focusing on cardiovascular disease and dementia. In fundus photography cohorts, a pooled analysis of six studies found wider venules predicted stroke (hazard ratio 1.15 per 20 micron increase). A deep learning model trained on roughly 280,000 patients' fundus photographs estimated sex (AUC 0.97), age (mean absolute error 3.26 years), smoking (AUC 0.71) and cardiac events (AUC 0.70). OCT-measured nerve fiber layer thinning correlates with Alzheimer disease and, in 30,000 UK Biobank participants, predicted poorer cognition three years later. The central argument is that deep learning on large multimodal datasets, exemplified by AlzEye, linking over 250,000 patients' scans to hospital records, could enable scalable risk stratification.

*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.