AI Now Predicts Your Biological Age And If You’ll Survive Cancer — With A Face Scan

A new artificial intelligence model can measure someone’s biological age and predict their chances of surviving cancer just by looking at an image of the person’s face, researchers say.

Scientists at Mass General Brigham Hospital in Boston built a new AI model that can provide what they’re calling a “non-invasive window into health,” according to a published study released on Oct. 19 in the medical journal The Lancet Digital Health. By simply scanning a selfie, the AI can determine how a person’s cells, tissues and organ systems are functioning as well as his or her likelihood of surviving cancer.

“Facial appearance offers a non-invasive window into health,” the study states. “We built Foundation Artificial Intelligence for Health Recognition-Face (FAHR-Face), a foundation model trained on more than 40 million facial photographs and fine-tuned it for two distinct tasks: biological age estimation (FAHR-FaceAge) and survival prediction (FAHR-FaceSurvival).”

The study’s researchers took these more than 40 million unlabeled photographs from WebFace42M — a dataset comprised of 42 million facial recognition images representing 2 million identities taken from across the internet, according to AI research discovery platform Emergent Mind. A research team from Tsinghua University in Beijing, Imperial College London and China-based XForwardAI developed the dataset in collaboration with each other.

After running these datasets through the FAHR models, researchers fine-tuned the AI on 34,363 photographs of patients with cancer from a research center in the Netherlands to detect signs of mortality risk, the study said. The models “outperformed a previous age estimation model in survival prognostication,” the report found.

“To our knowledge, this is the first report of an AI algorithm that can directly predict cancer patient survival from a simple face photograph,” the study in The Lancet Digital Health said.

Among 122 cancer patients found to have a high risk of death, more than half — 52.5% — died within 90 days of the assessment, as predicted, according to the study.

The algorithm assesses age by looking at the area around the temples, the forehead, above the eyes and the line that connects the nose down to the corner of the mouth, Dr. Raymond Mak, one of the study’s authors, explained to CNN.

“In contrast, the mortality risk prediction algorithm is looking at the bridge of the nose and the area under the eyes,” Mak, a radiation oncologist at Mass General, told the outlet. “I was pretty surprised to find out that we could pick up signs of mortality risk from just a facial photograph.”

The same regions of the face that the AI algorithms scan are the same parts Ancient Greek physician and philosopher Hippocrates referenced in the Book of Prognostics, CNN reported. Written thousands of years ago in 400 B.C., the father of medicine described “the nose sharp, the eyes sunken, the temples fallen in” as signs of mortality risk.

“To think that our algorithms might be replicating that wisdom of Hippocrates from millennia ago was mind-blowing to us,” Mak said according to the outlet, adding that the algorithms had no awareness of these ancient texts.

“Our algorithms are trained purely on face photographs and then whether somebody died or not, or their age,” the oncologist said.