The tool that could revolutionize care for oncological patients could be called Faceaage. Thanks to artificial intelligence and a photo of the face, it manages to estimate people’s biological age, but above all the probability of survival of cancer patients. To describe it in the magazine The Lancet Digital Health the scientists of the Mass General Brigham based in Massachusetts. The algorithm that allows you to analyze the facial characteristics of people with different types of cancer was developed by the team led by Hugo Aerts. These are the main results of the study.
Research
Faceage was trained using photos of 58,851 presumably healthy individuals from public datasets and on 6,196 oncological patients. The results showed that the cancer patients appeared significantly older than those without cancer. In particular, their faceage was on average about five years older than their biological age. Worst survival outcomes were associated with an older faceage, especially for oncological patients over the age of 85.
In a second moment, the researchers showed the images of 100 patients undergoing palliative radiotherapy to 10 doctors and experts, asking them to predict their short -term life expectancy. Despite the remarkable variability, Faceage emerged significantly improved the ability of the medical team to estimate the longevity of patients.
Scientists are now testing technology on a wider sample of patients, also evaluating their ability to predict diseases, general state of health and life duration.
A selfie to guide the treatment plans
The investigation therefore revealed that oncological patients had a higher count of biological age than healthy people, about five years. According to the researchers, Furthermore, Faceage has passed the doctors in predicting the short -term life expectancy of patients undergoing palliative radiotherapy.
“A simple selfie – says Aerts – contains important information that could contribute to orienting the clinical decision -making process and care plans for patients and doctors. The correlation between apparent and biological age is really important”.
“This work – says Ray Mak, another signature of the article – opens the doors to a completely new world of discovery of biomarkers starting from photographs, and its potential goes far beyond cancer or forecast age. I hope that in the future we can use this technology in a variety of applications, within a solid regulatory and ethical framework, to help save human vines”.