An AI that estimates biological age from a face photo
A computer looks at a face photo and estimates biological age, a measure of how the body is ageing. A new paper from Mass General Brigham describes the latest version of their FaceAge research. Here's what it does, what it doesn't, and the people research like this is really meant for.
Early, retrospective research. General information, not medical advice.

What the research did
Researchers at Mass General Brigham built an AI model, called FAHR-FaceAge in the paper, that estimates biological age from a face photo. It started from a general model trained on more than 40 million face photos, then was fine-tuned for age using ten public photo collections. The team also tested it on past patient records.
It's retrospective research: it looked back at existing photos and records. It isn't used to guide anyone's care, and it isn't a screening tool. The clinician decides. The corresponding author, Dr. Raymond Mak, says the goal is to "support personalized treatment decisions rather than replace physician judgment."

Where the human stakes are
Many older adults with early lung cancer have needs you can't see, but very few get a full check of how they're managing day to day. The trial registry cites earlier records from Brigham and Women's Hospital and Dana-Farber suggesting fewer than 5% of early-stage lung cancer patients received a full geriatric assessment. That kind of check can lead to practical help, like nutrition advice or physical therapy, before radiation begins.

A separate trial
This trial is separate from the paper. At Brigham and Women's Hospital, researchers are testing whether a short questionnaire, the Practical Geriatric Assessment, gets more older adults referred for support. It's for adults 65 and older with early lung cancer who are expected to have a focused form of radiation. People are randomly assigned, and in one group the care team is told the results and the recommended referrals. The main measure is how many people are referred to services such as nutrition counselling, physical therapy or social work a month later.
Referrals in the trial come from the questionnaire, not the AI. The trial also collects photos so the team can study whether AI like FaceAge could someday help spot who might benefit. The paper didn't test that. ClinicalTrials.gov lists the trial as recruiting, and there are no results yet.

Keep it in perspective
- It's retrospective research, not a screening tool. The clinician decides.
- The paper didn't test whether support helps anyone. The trial looks at referrals to support services.
- An outside expert, Prof Miroslaw Bober of the University of Surrey, called it "a promising proof of concept," but said it is not ready for clinical use and still needs testing against doctors' own judgement and across different hospitals and patient groups.
- The same expert noted that the AI estimates age from how a face looks, rather than directly measuring biological ageing, and that smiling or changing the lighting shifted its estimate by around two years on average.
Who's behind it
Some authors, including senior authors Dr. Mak and Dr. Aerts, report consulting, advisory or research ties to drug and device companies. The study was funded by the European Research Council. Mass General Brigham has filed a provisional patent application on these tools; the first author and both senior authors declared it.

Read more
- The paper: Haugg, Lee, He et al., "Foundation artificial intelligence models for health recognition by use of face photographs (FAHR-Face): a model development and validation study," The Lancet Digital Health, Oct 19, 2026 (open access). Also on PubMed.
- The trial's registry entry (NCT06987890), ClinicalTrials.gov.
- Expert reaction, Science Media Centre, Oct 19, 2026.
- Mass General Brigham's release, as published by Medical Xpress.
Illustrative photos, not from the study. Photos: Anastasia Shuraeva, cottonbro studio, Kampus Production, RDNE Stock project and Tima Miroshnichenko, all via Pexels.
General information, not medical advice. Talk to your own care team about your health.
Get the next one
One free email a week on AI in medicine and dentistry. Minus the hype.
Form not loading? Subscribe here.