AI predicts cancer patient survival

Researchers at the Massachusetts General Brigham Institute in New York have developed artificial intelligence models that can predict the biological age of cancer patients and their chances of survival using just a single facial photograph. These models rely on analyzing specific areas of the face, such as the temples, forehead, and the area under the eyes, which correspond with signs of wasting and mortality risk—areas that Greek physician Hippocrates described over 2,000 years ago. Study results showed that the model can accurately classify patients according to their risk of death within 90 days; more than half of patients labeled by the model as most at risk of death did, in fact, pass away within that period, while the mortality rate in the lowest-risk group was less than 2%. The study recommends using this technique as a supplementary tool for doctors, not as a replacement for traditional clinical assessments, as it can assist in making more accurate treatment decisions.