.jpg)
A new artificial intelligence model has been developed to measure someone's biological age and predict their chances of surviving cancer by scanning a facial image.
What happened
Scientists at Mass General Brigham Hospital in Boston created the AI model, called Foundation Artificial Intelligence for Health Recognition-Face (FAHR-Face), which was trained on over 40 million facial photographs. The model can determine how a person's cells, tissues, and organ systems are functioning, as well as their likelihood of surviving cancer, by analyzing the facial image.
The researchers fine-tuned the AI model on 34,363 photographs of cancer patients from a research center in the Netherlands to detect signs of mortality risk. The model was found to outperform a previous age estimation model in survival prognostication.
According to the study, the algorithm assesses age by looking at specific areas of the face, including the temples, forehead, and the line connecting the nose to the corner of the mouth. The mortality risk prediction algorithm looks at the bridge of the nose and the area under the eyes.
The study found that among 122 cancer patients predicted to have a high risk of death, over half (52.5%) died within 90 days of the assessment. The researchers noted that the same regions of the face scanned by the AI algorithms are similar to those referenced by Ancient Greek physician Hippocrates in his Book of Prognostics, written in 400 B.C.
