Nat Commun
AI facial aging tool predicts cancer survival
Clinical takeaway: Monitoring facial aging rate may offer a low-burden way to track changing health status during cancer care, but additional evaluation is needed.
Chronological age helps estimate cancer prognosis, but it can miss wide differences in biological aging. This study tested whether changes in AI-estimated facial age over time could provide a more dynamic marker of cancer survival risk.
Patients whose facial aging accelerated had lower overall survival during follow-up. The signal was stronger when photos were farther apart: adjusted mortality risk was 25% higher when photos were taken within about one year, 37% higher when they were one to two years apart, and 65% higher when they were two to four years apart.
Researchers analyzed routine clinical face photos from 2,276 adults with cancer who received at least two courses of radiation therapy at Brigham and Women’s Hospital from 2012 to 2023. The tool calculated Face Aging Rate, defined as the change in AI-estimated FaceAge divided by the time between photos. Most patients had metastatic disease, and the median time between photos was 286 days.
“Deriving a Face Aging Rate from multiple, routine facial photographs allows for near real-time tracking of an individual’s health,” said co-senior and corresponding author Raymond Mak, MD, a radiation oncologist at Mass General Brigham Cancer Institute and a faculty member in the Artificial Intelligence in Medicine program at Mass General Brigham.
The dynamic measure also added information beyond a single FaceAge reading. Patients with both older-appearing baseline FaceAge and faster subsequent facial aging had the highest mortality risk, but Face Aging Rate became the stronger prognostic signal over longer follow-up. The median patient age at the first radiation therapy course was 63 years.
The tool could eventually help clinicians refine counseling, follow-up intensity, and treatment planning, particularly when balancing symptom relief, quality of life, and treatment toxicity. The study was retrospective, included mostly white patients, and also lacked detailed data on disease progression and treatment toxicity.
“Tracking FaceAge over time from simple photos offers a non-invasive, cost-effective biomarker with potential to inform individuals of their health,” said study co-author Hugo Aerts, PhD, director of the AIM program at Mass General Brigham. “We hope with continued study we can learn how FaceAge may provide prognostic information for patients with other chronic diseases and for healthy individuals.”
Source: Haugg F. Nat Commun. 2026 Apr 28. Face aging rate quantifies change in biological age to predict cancer outcomes