AI Model of Retinal Age Can Predict Clinical Aging

TUESDAY, April 28, 2026 (HealthDay News) -- Retinal age, derived from a single fundus photograph, is a biomarker of clinical aging, according to a study published online April 8 in Communications Medicine.Takahiro Ninomiya, M.D., from the Tohoku University Graduate School of Medicine in Sendai, Japan, and colleagues trained an ensemble multitask learning model integrating fundus photographs with glycated hemoglobin using 50,595 quality-controlled fundus photographs from 27,214 disease-free adults. The model was validated on an independent set of 7,288 additional images from disease-free adults.The researchers found that the model achieved mean absolute errors of 2.78 years in internal validation and 3.39 and 8.63 years in two external cohorts, including 135 and 4,992 eyes, respectively. Improved age-prediction accuracy was seen in eyes with ensemble standard deviations below the median (mean absolute error, 2.46 and 2.87 years internally and in the primary external cohort, respectively). After adjustment for age and sex, the retinal age gap (predicted minus chronological age) was significantly higher in participants with diabetes, cardiac disease, or stroke in a systemic disease cohort of 8,467 individuals, indicating older-appearing retinas."Fundus images are noninvasive photos of the eye taken as part of regular health checkups -- so no additional work is needed," Nakazawa said in a statement. "Our model would be a nearly frictionless addition to a clinician's typical workflow."Abstract/Full Text (subscription or payment may be required).Sign up for our weekly HealthDay newsletter

AI Model of Retinal Age Can Predict Clinical Aging

Related Stories

Predictive Model Shows Moderate Performance for Psoriasis Relapse Risk

Swept-Source Optical Coherence Tomography Angiography May Predict Diabetic Nephropathy

Machine Learning Predicts Asthma Risk in Children With Early-Life Atopic Dermatitis

Multimodal AI Model Can Predict Intimate Partner Violence