Your odds of surviving cancer could soon be predicted by a selfie

Forget the old saying about cameras adding 10 pounds—this one is interested in how many years they add.

Researchers at Mass General Brigham have developed an AI system that can estimate a person's biological age and predict survival outcomes for cancer patients using a single photograph of their face.

A Snap filter on steroids

After training their AI model, called FAHR-Face, on 40+ million facial images, the researchers used it to develop two separate tools:

1. FAHR-FaceAge estimates biological age from a person’s selfie—which can differ significantly from actual age—and uses that information to assess survival risk.
2. FAHR-FaceSurvival predicts cancer survival outcomes directly from facial features, particularly the areas beneath the eyes and around the nose.

The results: Among US cancer patients undergoing radiotherapy, those whom FAHR-FaceAge identified as having a biological age 5+ years older than their actual age had a 21% higher risk of death. Meanwhile, patients whose faces gave a biological age 5+ years younger had a 17% lower risk of dying.

For FAHR-FaceSurvival, researchers found patients sorted into the highest-risk group had over 3x the mortality hazard of those in the lowest-risk group, after adjusting for clinical factors.

Looking ahead…Researchers hope the technology could eventually help doctors make more personalized cancer treatment decisions, though further testing is needed before the AI can be used in clinical settings.