FaceAge AI Brings New Precision to Biological Age
Doctors often begin medical evaluations with a quick visual check known as the “eyeball test,” where they estimate if a patient looks older or younger than their actual age. This impression can influence important healthcare decisions. Now, this intuitive method is getting a boost from technology. The FaceAge AI algorithm converts a simple headshot into a number that reflects a person’s biological age, offering a more accurate measure than the birthday on their medical chart.
Trained on tens of thousands of images, FaceAge revealed that cancer patients tend to be biologically about five years older than healthy individuals of the same chronological age. This insight could help doctors determine who can handle aggressive treatments and who might benefit from gentler options.
How FaceAge AI Helps Medical Decisions
Imagine two patients: one is a lively 75-year-old with a biological age of 65, and the other is a frail 60-year-old whose biological age reads 70. Aggressive therapy like radiation could be suitable for the first but risky for the second. FaceAge AI could guide decisions not only in cancer care but also in surgeries and end-of-life treatments.
Sharper Lens on Frailty with FaceAge AI
People age at different rates based on genetics, lifestyle, and habits such as smoking or drinking. Although genetic tests can measure aging, they are often expensive. FaceAge AI offers an accessible alternative by analyzing only a selfie. The model was built using nearly 59,000 portraits of healthy adults over 60 from public datasets.
It was then tested on over 6,000 cancer patients in the US and Europe, using photos taken before treatment. On average, cancer patients appeared biologically 4.79 years older than their actual age. Moreover, higher FaceAge AI scores predicted poorer survival outcomes, even after adjusting for age, sex, and tumor type. The risk notably increased for those with a biological age above 85.
Unique Aging Signs Detected by FaceAge AI
Interestingly, FaceAge AI focuses on subtle facial features differently from humans. For instance, it gives less importance to gray hair or balding and pays more attention to changes in facial muscle tone. This approach allowed doctors to improve their predictions about patient survival when combined with their own assessments.
In one test, eight physicians guessed which terminal cancer patients would survive six months based on photos alone. Their accuracy was only slightly better than chance. However, with FaceAge AI data, their predictions improved significantly.
Addressing Bias and Ethical Concerns of FaceAge AI
AI tools can sometimes perform poorly on non-white populations. Preliminary checks showed no major racial bias in FaceAge AI predictions. Nevertheless, developers are training a new model on 20,000 patients to enhance fairness. They are also studying how factors like makeup, cosmetic surgery, or lighting might affect results.
Ethical challenges arise with the ability to estimate biological age from selfies. While this tool can benefit doctors, it might also tempt life insurers or employers to assess risk unfairly. One expert noted, “It is for sure something that needs attention, to assure that these technologies are used only in the benefit for the patient.” Learning that one’s body is older biologically could motivate healthier habits or cause anxiety.
Future Plans for FaceAge AI
The researchers plan to launch a public portal where people can upload images to join a study validating the algorithm further. Clinical versions could follow after additional testing. This technology promises to change how biological age is assessed, offering a powerful tool for personalized healthcare decisions.