emirates7 - A pioneering study published in The Lancet Digital Health has introduced an AI-powered tool capable of analyzing facial features from a selfie to estimate how quickly someone’s body is aging — potentially giving doctors a new way to assess cancer risk and tailor treatment plans.
Meet FaceAge: AI that sees more than skin deep
Developed by researchers at Mass General Brigham, the deep-learning model, called FaceAge, was trained using nearly 60,000 images of healthy individuals and later tested on over 6,000 cancer patients about to begin radiotherapy.
Beyond birthdays
Instead of relying on chronological age, FaceAge examines subtle facial characteristics — such as skin condition, eye shape, and muscle tone — to estimate biological age, a more accurate reflection of a person's overall physical resilience.
The study found that, on average, cancer patients appeared nearly five years older biologically than their actual age. Each additional biological year was linked to a notable in survival rates, making this AI estimate a strong new prognostic indicator.
A shift from guesswork to precision
Doctors have long used visual cues informally to gauge a patient's strength or frailty, but these assessments are subjective. FaceAge offers a standardized, evidence-based alternative.
“We can now use AI to estimate biological age from facial images,” explained lead researcher Hugo Aerts. “Our findings this has real clinical value.”
Surpassing human judgment
To assess FaceAge’s impact in real-world scenarios, researchers asked ten clinicians to predict the survival chances of patients receiving palliative radiotherapy. Even when given access to medical records and actual age, their predictions were only marginally better than random guesses.
However, when FaceAge was included, prediction accuracy improved significantly, showing that the algorithm picks up on cues that often elude even trained professionals — potentially transforming how doctors make complex treatment decisions.
Potential far beyond cancer
Because accelerated aging contributes to numerous chronic diseases — including heart conditions and dementia — FaceAge could eventually be used to detect health risks well before symptoms emerge, paving the way for more personalized and preventive care.
“This could mark the start of a new era in biomarker discovery,” said Ray Mak, co-senior author. “Its potential applications stretch well beyond oncology.”
Still early days
Despite its promise, FaceAge isn’t clinic-ready just yet. The model was trained using data from only two hospitals, and factors like lighting, image quality, cultural differences, or cosmetic products could affect its accuracy.
The researchers plan to expand testing to include more diverse populations and settings. They also emphasize the importance of ethical considerations — including informed consent and clear communication about how patient data is used.
FaceAge offers a preview of how everyday selfies might one day help guide medical decisions — transforming ordinary images into tools for early disease detection and personalized healthcare.
Meet FaceAge: AI that sees more than skin deep
Developed by researchers at Mass General Brigham, the deep-learning model, called FaceAge, was trained using nearly 60,000 images of healthy individuals and later tested on over 6,000 cancer patients about to begin radiotherapy.
Beyond birthdays
Instead of relying on chronological age, FaceAge examines subtle facial characteristics — such as skin condition, eye shape, and muscle tone — to estimate biological age, a more accurate reflection of a person's overall physical resilience.
The study found that, on average, cancer patients appeared nearly five years older biologically than their actual age. Each additional biological year was linked to a notable in survival rates, making this AI estimate a strong new prognostic indicator.
A shift from guesswork to precision
Doctors have long used visual cues informally to gauge a patient's strength or frailty, but these assessments are subjective. FaceAge offers a standardized, evidence-based alternative.
“We can now use AI to estimate biological age from facial images,” explained lead researcher Hugo Aerts. “Our findings this has real clinical value.”
Surpassing human judgment
To assess FaceAge’s impact in real-world scenarios, researchers asked ten clinicians to predict the survival chances of patients receiving palliative radiotherapy. Even when given access to medical records and actual age, their predictions were only marginally better than random guesses.
However, when FaceAge was included, prediction accuracy improved significantly, showing that the algorithm picks up on cues that often elude even trained professionals — potentially transforming how doctors make complex treatment decisions.
Potential far beyond cancer
Because accelerated aging contributes to numerous chronic diseases — including heart conditions and dementia — FaceAge could eventually be used to detect health risks well before symptoms emerge, paving the way for more personalized and preventive care.
“This could mark the start of a new era in biomarker discovery,” said Ray Mak, co-senior author. “Its potential applications stretch well beyond oncology.”
Still early days
Despite its promise, FaceAge isn’t clinic-ready just yet. The model was trained using data from only two hospitals, and factors like lighting, image quality, cultural differences, or cosmetic products could affect its accuracy.
The researchers plan to expand testing to include more diverse populations and settings. They also emphasize the importance of ethical considerations — including informed consent and clear communication about how patient data is used.
FaceAge offers a preview of how everyday selfies might one day help guide medical decisions — transforming ordinary images into tools for early disease detection and personalized healthcare.