20.10.2026 14:20
An artificial intelligence tool developed by US researchers estimates survival by looking at the under-eye area and around the bridge of the nose in facial photographs of cancer patients. The tool is not yet ready for use in hospitals.
According to BBC News, US researchers have developed an artificial intelligence tool that could help doctors assess how seriously ill a patient is from a single facial photograph. Preliminary findings were published in the journal Lancet Digital Health.
The tool, called "FaceSurvival", was trained on photographs of 35,000 cancer patients as well as millions of healthy faces. According to the scientists, the tool can make predictions about patients' survival time by looking especially at features around the eyes.
According to the research, the most determining areas of the face in predicting life expectancy are under the eyes and around the nasal bridge. These criteria held true for different types and different severities of cancer, including those that had spread to the bone or brain. Most of the patients were from the US and the Netherlands, in their 50s and 60s; there were also children and people in their 90s among them.
It was stated that a patent application has been filed for the tool, which was designed not to replace doctors but to assist them. Dr. Raymond Mak, lead researcher from Mass General Brigham in Boston, said that as the population ages, doctors need better ways to measure a patient's physical resilience and frailty before recommending intensive treatments such as radiation, surgery or drug therapy.
The team is also evaluating another tool called "FaceAge", which estimates a person's biological age based on the age they appear rather than their date of birth. According to this tool, lines running from under the nose to the corners of the mouth, forehead wrinkles, crow's feet at the temples and the cheeks are good indicators of the rate of aging. Heavy alcohol and drug use and smoking can accelerate this process.
Mak said that despite differences in pose, lighting, facial expression and makeup, the tool produced "surprisingly consistent" results in real-life photographs, and that the result did not change in most of the cosmetic surgeries tested, with only a small shift seen after rhinoplasty.
The researchers emphasized that the tools are still in the development stage and that much more data is needed, especially from different ethnic groups. Prof. Miroslaw Bober from the University of Surrey in the UK also described the study as a "promising proof of concept" and said it was not ready for clinical use and that predicting an outcome is not the same as explaining it.
Source: BBC