AI Tool Predicts Cancer Survival from Patient Selfies

Source reports published 19 Oct 2026, 10:00 am – 20 Oct 2026, 5:29 pm AEST

Researchers have developed an AI tool, ‘FaceSurvival’, that analyzes a single patient selfie to estimate survival rates for cancer patients. The tool focuses on facial features like the area under the eyes and the bridge of the nose to predict outcomes, offering a potential aid for personalized treatment decisions.

What happened

• US researchers, led by Dr Raymond Mak at Mass General Brigham, have created an AI tool named ‘FaceSurvival’ that analyzes a patient’s face photograph to predict cancer survival.
• The tool was trained on millions of faces and 35,000 photos of cancer patients, focusing on features around the eyes and nose.
• Preliminary findings, published in the Lancet Digital Health, suggest the tool can estimate survival probabilities based on these facial metrics.
• The system, dubbed FAHR-FaceAge and FAHR-FaceSurvival, estimates biological age and directly predicts survival from a facial photograph.
• Researchers found that patients who appeared older than their chronological age had a higher risk of death during follow-up, while those who appeared younger had a lower risk.
• The tool’s development reflects a growing trend in using AI for health assessment, with researchers highlighting the need for more data and diverse ethnic representation to ensure accuracy.
• They insist the technology is intended to support, not replace, physician judgment in treatment decisions, particularly as the population ages and physicians seek better ways to assess patient reserve and frailty.

Original reporting

• BBC: Can one photo of your face reveal how ill you are?
  2026-10-20T07:29:01.536000+00:00
• Guardian: New AI tool can predict cancer survival by a single patient selfie, study suggests
  2026-10-19T00:00:00+00:00