Researchers at Mass General Brigham have developed a new artificial intelligence algorithm called FaceAge, which uses facial photographs to estimate biological age and predict cancer survival outcomes. The tool, described in a study published in The Lancet Digital Health, is designed to support personalized medical assessment by offering clinicians a non-invasive, data-driven measure of a patient’s apparent health status.
AI-Based Biomarker Offers Objective Patient Insights
Unlike chronological age, biological age can provide a more accurate reflection of a patient’s physiological condition. FaceAge uses a deep learning algorithm trained on nearly 59,000 facial images from healthy individuals to establish aging baselines. It detects subtle visual cues associated with aging and health decline, effectively digitizing what clinicians refer to as the “eyeball test” used to gauge patient frailty or vitality.
The model was tested on over 6,000 cancer patients at two major medical centers. Results showed that patients with cancer appeared, on average, five years older biologically than their chronological age. More importantly, patients with higher FaceAge scores had significantly worse survival outcomes, even when adjusted for age, sex, and tumor type. These findings suggest the algorithm could serve as a valuable biomarker in clinical oncology.
“We hypothesize that FaceAge could be used as a biomarker in cancer care to quantify a patient’s biological age and help a doctor make these tough decisions,” said Raymond Mak, co-senior author and oncologist at Mass General Brigham.
Clinical Use Cases and Broader Implications
One potential application is in treatment planning. For instance, a 75-year-old patient with a biological age of 65 may be better suited for aggressive therapies than a 60-year-old patient with a biological age of 70. By integrating FaceAge into decision-making, clinicians could gain an added layer of insight that complements traditional diagnostic tools.
The technology could also be extended to other clinical areas, such as cardiovascular surgery, orthopedic interventions, and end-of-life care, where biological resilience plays a critical role in outcomes. As a non-invasive and easily repeatable tool, FaceAge has potential for widespread clinical implementation, although further validation studies are needed before routine adoption.
The news comes amid broader developments in AI-powered medical technologies at Mass General Brigham, including enhanced diagnostic imaging tools. The healthcare industry is increasingly exploring algorithmic assessment methods, with other researchers developing novel approaches like hand-based age verification.
Sources: Eurekalert, Cybernews, ScienceAlert
—
May 9, 2025 – by the ID Tech Editorial Team
