AI Facial Analysis Predicts Cancer Survival Rates From Single Photo
Researchers developed an AI system that predicts cancer survival and biological age from a single facial photo, analyzing specific facial regions to assist oncologists in treatment decisions.
Doctors have developed an innovative artificial intelligence system capable of predicting cancer survival rates using a single patient facial photograph. According to a report in The Guardian, the system was developed by experts at the Mass General Brigham academic healthcare network in Massachusetts. The technology is designed to assess an individual's biological age and directly predict survival outcomes for patients diagnosed with cancer.
The details of this medical breakthrough were recently published in the journal Lancet Digital Health. The revolutionary system was trained on data from more than 40 million facial images of people of various ages and ethnicities, representing one of the most comprehensive demonstrations of integrating technology into medical diagnostics. Based on these data, the model estimates the age a person appears to be, contrasting sharply with their actual chronological age.
Two Tools and Facial Region Analysis
Physicians used the technology to develop two complementary tools derived from the same model but gathering different patient information. The first tool assesses biological age and predicts cancer survival by focusing on wrinkles between the nose and the corner of the mouth, the forehead area, and the temples. The second tool predicts survival directly from the image alone by examining the area under the eyes and the bridge of the nose.
Hugo Aerts, one of the study's authors, noted that this is the first demonstration that an AI model can directly predict cancer survival from a simple photograph. Among cancer patients treated with radiation therapy in the United States, those whom the system estimated to look at least five years older than their actual age suffered a 21 percent higher mortality risk. Conversely, patients who looked at least five years younger than their age exhibited a 17 percent lower risk of dying during an average follow-up period of four years.
Lifestyle Impacts and Medical Decisions
Dr. Raymond Mak, a radiation oncology expert, explained in The Guardian that as the population ages, physicians need better ways to assess patients' physical resilience before recommending intensive treatments. The researchers found that lifestyle factors such as heavy alcohol consumption, drug use, and smoking were clearly linked to an older appearance in the system's analysis. To their surprise, the algorithm remained stable despite changes in makeup, lighting, and plastic surgery, except for a slight shift observed following a nose job.
This breakthrough could pave the way for using facial analysis to help physicians make complex decisions regarding the ongoing treatment required for their patients. However, the doctors emphasized that while this is a simple and inexpensive method, further research is required before the system is widely introduced into hospitals. The ultimate goal, the researchers clarified, is to support medical decision-making consistently rather than replace the professional judgment of doctors in the field.
