BOSTON (WHDH) - Doctors at Mass General Brigham (MGB) created artificial intelligence (AI) models that analyze pictures of faces for key indicators of a person’s health, and say the technology could help predict one’s chances of beating cancer.
Foundation Artificial Intelligence for Health Recognition (FAHR-Face) is an AI foundation model that can estimate someone’s biological age and survival outcomes in cancer patients. Mass General Brigham said the AI technology was trained using more than 40 million facial images.
“For a separate AI algorithm, we trained it just on the face photograph plus whether the patient lived or died, and for how long. And that AI learned the facial features that could help or predict whether someone is going to live longer or shorter,” said Dr. Raymond Mak, MD, a Radiation Oncologist at Mass General Brigham Cancer Institute and a faculty member in the Artificial Intelligence in Medicine (AIM) program at Mass General Brigham.
The models specifically assess different aspects of the face to guage a cancer patient’s overall health and biological age. Biological age reflects how well the body has aged at the cellular and physical level.
“For the age estimation algorithm, it looked at the areas above the eyes, the forehead, the areas of the temples, and the lines that connect your nose to the corner of your mouth. In contrast, the algorithm that predicts mortality risk looks at totally different parts of the face. It looks at the bridge of your nose, and looks at the areas under the eyes,” Mak explained.
MGB said cosmetic changes such as facial surgery, botox, and makeup did not affect the AI models’ performances. Mak said as people age, the models will better estimate a person’s physical reserve or fragility before recommending treatments like radiation or surgery.
“As the population ages, physicians need better ways to estimate physical reserve and frailty before recommending intensive treatments such as radiation, surgery, or systemic therapy,” Mak said. “Our goal is to make face photograph-based health assessment more quantitative, consistent, and scalable so it can support personalized treatment decisions rather than replace physician judgment.”
Mak said researchers are amazed to see the capabilities of their once bold idea.
“When we first started it was one of those high-risk, probably isn’t going to work but why not, let’s give it a try,” Mak said. “It has been amazing to see that AI can actually pick up on these subtle features that the human eye can’t really see, and help us better prognosticate cancer patient care.”
(Copyright (c) 2026 Sunbeam Television. All Rights Reserved. This material may not be published, broadcast, rewritten, or redistributed.)