Specialists at Mass General Brigham, a non-profit health system in Massachusetts, have built an artificial intelligence tool that can estimate a cancer patient's chance of survival from an ordinary photograph of the face. The findings are published in The Lancet Digital Health.
The tool rests on biological age: not the number of years since birth but how old a person looks. The AI estimates it by analysing age-related features in a face photo. The model was trained on face photographs of about 40 million people of different ethnicities and ages.
Two complementary tools were built from the same model: FAHR-FaceAge and FAHR-FaceSurvival. The first estimates biological age from the picture and then assesses the chance of surviving cancer; the second predicts survival directly from the face, without the age estimate.
The study found the two rely on different parts of the face. FaceAge looks at the folds running from the sides of the nose to the corners of the mouth, the forehead above the eyes and the temples. FaceSurvival assesses the area under the eyes and the upper part of the nose.
Hugo Aerts, a senior co-author, said FaceSurvival is a new step in face-based health analysis and, to his knowledge, the first AI model able to predict cancer survival from an ordinary photo. The researchers call it an important advance in learning about a patient's health from a simple picture.
The tools are still at the research stage; the report does not say how they would be used in treatment decisions. Samakal's international desk based its report on The Guardian.
Photo: Wikimedia Commons (Public domain)
Source: Samakal
Comments
(0)