American scientists have developed a system that analyzes facial features to assess biological age and survival opportunities by offering information that may help doctors in choosing treatment
A common face picture may contain important information on health and the survival of cancer patients. This is suggested by a new American study in which scientists developed an artificial intelligence system ( AI that analyzes facial features to assess biological age and the risk of mortality.
Research was conducted by health – care scientists Mass General Brigham, while results were released on October 19, 2026, in the scientific magazine. The Lancet Digital Health.
Technology can help assess the body’s overall state, especially among patients facing difficult oncological treatments.
Two algorithms that analyze different facial parts
The system, called The FAHR-Face has been trained using more than 40 million face pictures of people of different ages and backgrounds.
Two algorithms with special functions have developed from this base.
FAHR-FaceAge estimates the patient’s biological age, analyzing traits that may indicate whether the organism represents signs of greater aging than calendar age. The algorithm focuses on wrinkles between the nose and the mouth, the face on the eyes and the temples.
FAHR-FaceSurvival, meanwhile, directly estimates the risk of mortality in cancer patients, without necessarily relying on biological age. This algorithm analyzes mainly the area below your eyes and nose roots.
According to researchers, combining both algorithms improves the system’s ability to distinguish patients with higher risk.
You look five years older? Consider the results
One of the most interesting finds relates to the difference between real age and what artificial intelligence values from the face.
During an average period of four years, radio – therapy patients in the United States were observed:
- Patients who, according to algorithms, seemed at least five years older than their actual age had 21 percent higher mortality risks.
- Those who seemed at least five years younger had 17 percent lower risk.
Meanwhile, algorithm FAHR-FaceSurvival indicated that patients classified in the group with the highest risk were above three times the risk of mortality compared to the lower risk group, even after statistical adaptations for other factors.
However, these figures represent statistical links between facial appearance and survival, not a certain prediction about the life span of an individual.
Smoking and alcohol affect aging
Research also identified links between certain lifestyles and apparent facial aging.
Smoking, excessive alcohol consumption, and the use of narcotics were accompanied by a higher – rate biological age.
The most pronounced link was noted in smoking, where people who kept smoking looked biologically older than former smokers and people who had never smoked.
Researchers also found that the results of the algorithm for age remained relatively stable even when lighting changed, the angle of photography, makeup, or facial expression.
How can this technology help doctors?
According to the coauthor of the study, oncologist Raymond Mack, the assessment of the patient’s physical reserves and general state is especially important before intensive treatment.
Great operations, radio therapy, and chemotherapy can be physically difficult, so doctors must carefully evaluate the body’s ability to cope with them.
Such a tool can provide additional information when choosing the most appropriate treatment.
Experts point out, however, that artificial intelligence cannot replace the doctor’s clinical evaluation, laboratory tests, or diagnostic examinations.
It has not yet been confirmed if the use of this system improves therapeutic decisions and the results of patients. Hence, further clinical studies are needed in the most diverse populations before technology is used in daily medical practice. /Telegraphy/