How quickly are you getting old? Artificial intelligence tells you: just a photo to find out what is the biological (and non -personal) age of your body.
Just a photograph to find out how old she really is our body. There is no mention of the age of the registry, that is, the one marked on the identity card, but of the biological card, which tells how quickly we are aging. This is the promise of Faceagean algorithm of Deep Learning described in the magazine The Lancet Digital Health. Trained on tens of thousands of images, the software managed by artificial intelligence translates the features of the face in a number of years that reflects the overall health status.
First tests. In clinical tests, Faceage has revealed that oncological patients appear biologically older than about five years than healthy peers. But What is the usefulness of this information? According to the authors, such an instrument could help doctors to make crucial decisions, for example whether to opt for aggressive therapies or choose lighter treatments.
The difference between chronological and biological age, in fact, could radically change the approach to complex treatments such as radiotherapy, chemotherapy or delicate surgical interventionspushing towards a more personalized medicine model.
Digital eye. The program was born to refine what doctors call “eyeball test“, That is, the evaluation of a person’s naked eye of the state of health. The new algorithm, however, offers a more objective and precise approach, capable of grasping details that often escape the human eye.
The system was trained on over 58 thousand portraits of adults over 60 in good health and then tested on almost 6 thousand oncological patients treated in the United States and in the Netherlands. The results show that Those who presented a biological age over 85 years, even if chronologically younger, had lower survival prospects. And it is not just a photograph of the current state: the algorithm has proven capable of provide with surprising precision also the risk of mortality within six monthsovercoming the predictive skills of eight expert doctors called to evaluate the same images.
The downside. Like many recently introduced applications in the medical field, Faceage also raises Ethical questions. If on the one hand it can become a precious tool for oncologists, cardiologists or geriatricians, on the other it might interest Insurance companies or employers looking for methods to assess the risks related to health.
The developers ensure that they have checked that the system does not present racial prejudices – one of the most discussed critical issues in the use of the AI - but are already using to process a more advanced version, trained on further 20 thousand patients, to strengthen the accuracy of the forecasts.
At the same time, the variables that could deceive the algorithm are exploring, such as heavy make -up, cosmetic surgery or artificial lighting conditions, factors that could alter the perception of somatic traits by confusing the machine.
Ruthless mirror. Finally, there is the psychological question. Knowing that your body is aging faster than expected could stimulate to change lifestyle, but also generate anxiety or depression. The researchers underline that such a tool must be used with caution and always under the supervision of qualified medical personnel. To deal with these scenarios, in the context of a scientific study designed to collect further data, you are planning a portal open to the public where it will be possible to load a selfie and receive an evaluation. The commercial versions for clinical use, on the other hand, will only arrive after further validations and rigorous controls.
Meanwhile, Faceage has already tried with a famous face: that of the actor Paul Rudd, biologically estimated by 43 years in a photo taken when he had 50; a clue on how This technology could easily become a pop phenomenon destined to make people discuss, climbing over the medical area.
