A new study suggests that a simple selfie may soon help doctors assess a patient’s “biological age” and predict how well they might respond to treatments such as cancer therapy. While chronological age indicates how many years a person has lived, biological age reflects their physiological condition, influenced by genetics, lifestyle, and overall health. Researchers from Mass General Brigham in Massachusetts have developed a tool called FaceAge, an artificial intelligence algorithm that uses facial analysis to estimate biological age.
This deep learning model was trained on 59,000 images and can analyze a standard facial photograph to generate a biological age score. The tool is inspired by the traditional “eyeball test” doctors often use to judge a patient’s fitness, especially when determining eligibility for intensive treatments. However, FaceAge replaces this subjective visual assessment with a standardized, data-driven alternative.
The research team applied FaceAge to a cohort of 6,200 cancer patients using images taken at the beginning of their treatment. On average, these patients had a biological age five years higher than their chronological age. Notably, patients whose FaceAge score exceeded 85 years tended to have significantly worse survival outcomes. The findings suggest that biological age assessed through facial features may serve as an independent biomarker to help guide clinical decisions.
To demonstrate the accuracy of FaceAge, researchers compared photographs of two well-known actors—Paul Rudd and Wilford Brimley—both aged 50 at the time. Rudd’s biological age was estimated at 42.6, while Brimley’s was calculated to be 69. These assessments aligned with the men’s visible differences in health and vitality, illustrating how the tool can capture subtle indicators of aging.
According to the study’s authors, FaceAge is a low-cost, non-invasive tool that can be used repeatedly to track changes in a person’s health over time. Its practical applications could range from evaluating surgical risk to monitoring long-term wellness. However, researchers emphasized that FaceAge is intended to support—not replace—clinical judgment. It serves as an additional data point, much like blood tests or vital signs.
Further studies are ongoing to validate the tool’s accuracy across other diseases and populations. Researchers are also investigating how interventions like cosmetic surgery or Botox might affect the tool’s readings and whether FaceAge can be generalized beyond oncology settings.
READ MORE: