Mass General Brigham AI predicts cancer survival from a selfie — SkimNews
Get the Health newsletter
Daily health & science — research, biotech, public health, the studies worth knowing. Free.
- Mass General Brigham researchers developed the FAHR-Face AI system, which predicts cancer survival from a single facial photograph, with results published in the Lancet Digital Health journal.
- The model was trained on more than 40 million facial photographs spanning a range of ethnicities and ages, using a foundation-model approach that can be adapted to multiple health outcomes.
- Two complementary tools emerged from the model: FAHR-FaceAge estimates a patient's biological age, while FAHR-FaceSurvival predicts survival directly from facial features rather than through age.
- US cancer patients treated with radiotherapy who looked five or more years older than their actual age had a 21% higher risk of dying during an average four-year follow-up, while those who looked five or more years younger had a 17% lower risk.
- The two algorithms rely on distinct facial regions: FaceAge analyzes the nasolabial creases, forehead above the eyes, and temples, while FaceSurvival focuses on the area below the eyes and around the nasal bridge.
- Dr. Raymond Mak, a radiation oncologist at Mass General Brigham, said the tool stayed consistent across pose, lighting, expression, and makeup variations, and across most plastic surgery procedures — with only a small shift observed after rhinoplasty.
- Researchers linked heavy alcohol use, drug use, and smoking to looking biologically older in the AI's analysis, though they cautioned that more research is needed before the tool can be used widely in clinical care.
Why it matters: For oncologists weighing whether aging patients can tolerate radiation, surgery, or systemic therapy, the tool offers a low-cost, quantitative frailty screen: a 21% higher four-year mortality risk for patients who look five or more years older than their age gives clinicians a concrete benchmark that goes beyond subjective visual assessment, though the researchers stress physician judgment must remain central.
Ask SkimNews
