Newsroom
For someone facing cancer, choosing treatment involves more than deciding what might work against the disease. Doctors also have to consider how much the patient’s body can handle.
Researchers in the United States say an ordinary photograph of a patient’s face could eventually help with that difficult decision.
A team at Mass General Brigham has developed an artificial intelligence system that estimates biological age and survival risk in patients already diagnosed with cancer. The research, published in The Lancet Digital Health, suggests facial features can provide information about health that goes beyond the age on a birth certificate
What does a face reveal?
The researchers trained the underlying AI model on more than 40 million facial photographs spanning different ages and ethnicities. They then developed two tools with different jobs.
FAHR-FaceAge estimates biological age, an indicator of physical condition rather than simply the number of years someone has lived. FAHR-FaceSurvival estimates survival risk directly from a photograph of a cancer patient.
The distinction matters because two people of the same age can have very different levels of strength, frailty and overall health.
In a group of U.S. cancer patients treated with radiotherapy, those the AI assessed as looking at least five years older than their actual age had a 21% higher risk of dying during follow-up, which averaged four years. Those assessed as looking at least five years younger had a 17% lower risk than the rest of the group.
These are differences in relative risk across groups. They do not mean the system can tell an individual patient how long they will live.
Helping with difficult treatment decisions
For families in Cyprus, the potential relevance is a familiar concern: will a loved one be able to cope with surgery, radiotherapy or other demanding cancer treatment?
The researchers hope facial analysis could become an additional, low-cost way to help assess a patient’s physical reserves before treatment decisions are made.
Dr. Raymond Mak, a radiation oncologist involved in the study, said the aim was to make those assessments more consistent and measurable while supporting personalized care.
The two tools appeared to capture different information. Combining their results improved the statistical model’s fit in a separate validation group of cancer patients.
Different tools, different facial clues
The age-estimation tool focused on areas including the forehead, temples and folds running from the nose toward the corners of the mouth. The survival tool drew more heavily on the area below the eyes and around the bridge of the nose.
Researchers also found that smoking, heavy alcohol use and drug use were associated with older facial-age estimates.
Changes in lighting, facial expression, pose and makeup generally had little effect on the age tool’s results, according to the study.
Still, the technology remains under investigation. More research is needed before it can be widely used in medical care.
A photograph cannot diagnose cancer, and a risk estimate is not a verdict on a patient’s future. The potential is to give doctors another useful piece of information when helping patients decide what comes next.
*The Guardian
