Selfie AI Tool Predicts Cancer Survival, Study Finds
In October 2026, researchers at Mass General Brigham published a study in The Lancet Digital Health describing an artificial intelligence system called FAHR-Face that can estimate a cancer patient's biological age and directly predict survival outcomes from a single photograph of the patient's face. The study authors say this is the first demonstration that an AI model can directly predict cancer survival from a simple photograph.
Ray Mak, a radiation oncologist at the Mass General Brigham Cancer Institute and the study's corresponding author, said that combining both the face age and face survival scores into a clinical prognostic model improves survival prediction beyond either score alone. Hugo Aerts, a professor at Harvard University and director of the AI in medicine program at Mass General Brigham, said the work "advances the field of facial health recognition AI a step further by learning survival risk directly from facial photographs of patients with cancer," echoing the authors' claim that this is the first such demonstration.
The underlying model, FAHR-Face (Foundation Artificial Intelligence for Health Recognition), was trained on more than 40 million facial photographs spanning a range of ages and ethnicities, plus photographs of more than 30,000 cancer patients, many of whom had already died. The study found that cancer patients' faces averaged five years older than their chronological age, and that looking older was associated with worse outcomes. In a larger analysis of more than 38,000 cancer patients, those who looked older than their chronological age also had a higher risk of dying. After accounting for age, sex, cancer type and other factors, patients who appeared at least five years older than their actual age had a 21% higher risk of death. Those with the highest risk scores were more likely to die within 90 days of when their photograph was taken.
The researchers developed two complementary tools from FAHR-Face. FAHR-FaceAge estimates a patient's biological age, while FAHR-FaceSurvival predicts mortality risk directly from a facial photograph without going through biological age. The two systems examine different areas of the face: the age-prediction tool focuses on forehead lines, temples and creases around the mouth, while the survival tool concentrates on areas beneath the eyes and around the bridge of the nose. The researchers also found that lifestyle factors such as heavy alcohol use, drug use and smoking were linked to looking biologically older in AI analysis of facial photographs.
The tool is intended to complement doctors' clinical instincts and potentially guide treatment decisions, especially when medical data is scarce. The study reports that FaceAge performed well both in patients treated with curative intent, with life expectancies of several years, and in patients at the end of life, with expected survivals of weeks to months. Mass General Brigham also runs an online, institutional review board-approved FaceAge Research Study in which members of the public can upload a photograph and receive a research estimate of their facial age; the portal has drawn more than 10,000 sign-ups since it quietly launched in March. The team stresses that the technology should be used within a strong regulatory and ethical framework.
The study concludes that facial characteristics visible in a photograph hold information about a person's age that deep learning algorithms can use to enhance the accuracy of survival forecasts for cancer patients. The next steps include integrating the technology into clinical workflows and exploring its potential as an early detection system across a variety of chronic diseases, since many such diseases are increasingly viewed as diseases of aging. Experts also caution that the model is not perfectly accurate and will require further validation before widespread clinical adoption.