A selfie may reveal more about a person's health than we imagine. A team at Mass General Brigham, an academic health system in Massachusetts, has built an artificial intelligence tool that predicts a cancer patient's chances of survival from a single photograph of their face.
The system was trained on more than 40 million facial photographs across a range of ethnicities and ages, and the results were published in The Lancet Digital Health. The researchers describe it as one of the most comprehensive demonstrations so far of how AI analysis of a selfie could help assess a patient's health and guide their care.
Two tools from one model
At the core is FAHR-Face, a foundation model first trained broadly on faces and then adapted to two distinct tasks. FAHR-FaceAge estimates biological age, the age the body shows rather than the chronological one. FAHR-FaceSurvival predicts survival directly from the photo, without going through biological age.
The two algorithms look at different parts of the face. FAHR-FaceAge focuses on the creases running from the sides of the nose to the corners of the mouth, the forehead above the eye sockets and the temples. FAHR-FaceSurvival examines the area below the eyes and around the nasal bridge.
"FAHR-FaceSurvival advances facial health recognition AI a step further by learning survival risk directly from facial photographs of patients with cancer," said Hugo Aerts, co-senior author of the study. "To our knowledge, this is the first demonstration that an AI model can directly predict cancer survival from a simple photograph."
The numbers
Among cancer patients treated with radiotherapy in the US, those who looked five or more years older than their actual age had a 21% higher risk of dying during the average four-year follow-up. Those who looked five or more years younger had a 17% lower risk.
The findings were also tested in an independent group of nearly 5,000 cancer patients in the Netherlands, according to experts quoted by the Science Media Centre. The tool is not perfectly accurate, but patients with higher risk scores did tend to die more often.
Possible uses
The clinical goal is to support difficult decisions. The tool could help predict the risk of death or complications after major surgery or cancer treatment, as populations age and doctors need more objective measures of a patient's physical reserve.
"As the population ages, physicians need better ways to estimate physical reserve and frailty before recommending intensive treatments such as radiation, surgery or systemic therapy," said Raymond Mak, a radiation oncologist at Mass General Brigham. "Our goal is to make face photograph-based health assessment more quantitative, consistent and scalable so it can support personalised treatment decisions rather than replace physician judgment."
The study also linked lifestyle to facial ageing: heavy alcohol use, drug use and smoking were associated with looking biologically older. The results held steady across pose, lighting, expression and makeup, and across most cosmetic procedures tested, with only a small shift after rhinoplasty.
Doctors caution that more research is needed before the tool can be widely used. A single photograph offers a simple, low-cost route to support treatment decisions, but the system remains at the study stage.
No comments yet. Be the first!