Harvard’s ‘FaceAge’ AI Links Cancer Survival to How Old You Look in a Photo

The new artificial intelligence model developed at Harvard University, called Faceaga, estimates biological Age analysis of the facial features in the photos – and can help predict the survival of cancer by indicating how much the patient appears in relation to their real age.

Dressed in pictures 58,851 healthy people, Faceage Later, it was tested on a cancer patient to investigate whether it could look if the appearance older than someone’s chronological age would signal worse health results.

“We found that cancer patients look older on average than their chronological age, and look older with worse general survival,” the report said. “Faceage has shown significant independent prognostic performance in a series of cancer types and stages.”

Chronological age refers to the number of years that a person is alive, while biological age reflects how well – or bad – their body works over that number. According to Harvard researchers, a person’s physical appearance can provide effective biomarkers to determine their biological age.

Faceage is based on earlier work from Eth Zurich, where researchers have created deep expectations (Deks), an open -core deep learning model that evaluates the apparent age from the face pictures. Harvard team also trained Faceage using pictures from Imdb-wiki and UtkfaceTwo of the largest publicly available facial data sets.

Since 2006 Harvard has committed significant funds for understanding and reversal Biological aging. More recently, the University has expanded investment in research aimed at AI focused on diagnosis and treatment Cancerfields that are increasingly converging.

In October 2024, developers at the Harvard Medical School introduced a new AI model, known as the clinical foundation for evaluation of histopathology (MAIN). At that time, researchers noticed that AI surpassed the previously tested models with 96% accuracy in cancer detection.

While faceage research focused on biological age and cancer, researchers said it could lead to wider application.

“These findings can be expanded to diseases outside cancer, motivating deep learning algorithms to translate the patient’s visual appearance into objective, quantitative and clinically useful measures,” said Harvard researchers.

Faceage is the latest tools In the growing movement of medical experts to focus on biological age, using facial analysis to determine the early signs of falling and the care of prevention, not just treatment.

According to experts such as Kiano Katanforoosha, an assistant deep learning professor at Stanford University and the founder of the SKIlls Intelligence Company Worker, a shift towards AI in the biological age research talks about overcoming human restrictions.

“Ai analyzes thousands of features in the face for things that most of us consciously not notice and find patterns that are correlated with biological aging,” said Katanforoosh Decipher. “It is similar to how early deep learning models improved from people when discovering cats in the photos. They did not use intuition. They were trained for millions of examples and learned what was statistically consistent.”

“People are biased and inconsistent,” he added. “Ai is systematically trained on a scale that we can’t match.”

Edited Andrew Hayward