AI FaceAge Tool Uses Selfies to Predict Biological Age and Cancer Survival

A single selfie might reveal more than just a smile—it could unlock secrets about your biological age and even predict cancer survival. Researchers at Mass General Brigham have unveiled FaceAge, an AI-powered tool that analyzes facial photos to estimate how fast someone is aging biologically, distinct from their chronological years.

Trained on nearly 59,000 images of healthy individuals, FaceAge doesn’t just guess age—it dives deeper. When tested on over 6,000 cancer patients before radiotherapy, the tool pegged their biological age about five years higher than their actual age, hinting at the toll cancer takes. More strikingly, it outperformed seasoned clinicians in predicting life expectancy for 100 palliative care patients, offering a chillingly accurate glimpse into survival odds.

Published in The Lancet Digital Health, the findings suggest FaceAge could reshape clinical decisions. By removing bias tied to a patient’s appearance, it might guide fairer, more precise treatment plans. ‘A selfie could become a clinical tool,’ said Hugo Aerts, PhD, from Mass General Brigham’s AI in Medicine program, noting its potential to inform care strategies.

Yet, the tool isn’t ready for hospitals. More diverse studies across various cancer stages and diseases are needed to ensure it’s robust. Ethical hurdles also loom—issues of data privacy, consent, and the psychological weight of being told you ‘look older’ demand careful handling. As Dr. Harvey Castro, an AI expert, put it, ‘This tech quantifies the eyeball test, but it must augment, not replace, human empathy.’

FaceAge opens a bold new frontier, where a snapshot might one day flag health risks early, from cancer to chronic diseases. For now, it’s a reminder: our faces may tell stories science is only beginning to read.