AI Facial Analysis Predicts Cancer Survival Outcomes

Your face might reveal more about your health than you realise. In a massive study of 24,556 cancer patients aged 60 and older, researchers employed 'FaceAge', a deep learning system designed to estimate biological age from facial photographs. The goal was to see if the difference between a patient's facial appearance and their actual birth date held clues about their prognosis.

The results were striking. The AI found that 65% of the patients had a FaceAge older than their chronological age. Crucially, this gap served as a powerful predictor of survival. Patients appearing 10 or more years older than their actual age faced a significantly higher risk of mortality, including early death within 30 to 60 days. In contrast, those appearing at least five years younger than their calendar age enjoyed better survival outcomes.

This 'discordance' between looked age and real age acts as an independent biomarker. The findings suggest that non-invasive facial assessments could become a vital tool for clinicians when modelling prognosis and deciding on personalised treatment plans.