Foundation AI Model Predicts Cancer Patient Survival From Facial Photographs, Outperforming Clinical Intuition

Researchers at Mass General Brigham have developed an artificial intelligence system capable of predicting cancer patients' survival outcomes by analyzing photographs of their faces, according to a study published in The Lancet Digital Health. The model, called FAHR-Face (Foundation Artificial Intelligence for Health Recognition), was trained on more than 40 million facial photographs to establish a general representation of facial features, then fine-tuned on 34,363 cancer patient images to predict mortality risk and estimate biological age.