A deep-learning model has been developed to estimate chronological age using high-resolution images of eye corners. This non-invasive approach outperforms DNA methylation clocks in accuracy and could help assess lifestyle, medical, and cosmetic interventions for ageing. Wrinkles and skin pigmentation serve as key visual biomarkers.
November 2018 – Aging
Key takeaways
- Eye corners reveal biological ageing: High-resolution images of eye corners can accurately estimate chronological age, making them a powerful non-invasive biomarker. Wrinkles and skin pigmentation in this area provide key insights into ageing, surpassing traditional DNA methylation clocks in precision
- Deep learning improves age prediction: Advanced neural networks trained on thousands of images can predict age with a mean absolute error of just 2.3 years. This technology refines age estimation and offers potential for assessing lifestyle, cosmetic, and medical interventions
- Non-invasive ageing assessment is possible: Unlike invasive blood tests, photographic biomarkers offer a simple and accessible way to track ageing. This approach enables individuals to monitor skin health and evaluate interventions without the need for complex lab testing
- Skin reflects overall health status: Changes in wrinkles and pigmentation correlate with ageing and may indicate broader physiological changes. Monitoring these visual markers can provide insights into longevity and help guide skincare, nutrition, and wellness strategies
Read the article at: Bobrov, Eugene, et al. “PhotoAgeClock: Deep Learning Algorithms for Development of Non-Invasive Visual Biomarkers of Aging.” Aging, vol. 10, no. 11, 2018, https://doi.org/10.18632/aging.101629.