NEWS

Your facial temperature can help you guess your age…and spot diseases



What if subtle changes in the temperature of your face could be used by artificial intelligence (AI) to guess your age? That’s what a team of researchers from Chinese universities have had fun testing. With success, if we are to believe the British weekly New Scientist, which echoes thestudy published on July 2 in Cell Metabolism.

But that’s not all. The AI-based software that Peking University’s Jing-Dong Han and his colleagues have developed can also detect health problems such as high blood pressure, diabetes and liver disease from thermal images of the face.

To do this, the researchers started by using a commercially available thermal imaging camera. They took the facial temperatures of 2,811 ethnic Han volunteers, aged 20 to 90, during their annual health check-ups at the hospital, and created “heat maps” of their faces.

An expected validation

“By comparing images obtained from volunteers of different ages, Jing-Dong Han and his team discovered that thermal distribution changes with age in very clear patterns, explain New Scientist. Thus, the temperature of the nose drops from around the age of 50 in women and 60 in men, while the temperature of the eye area increases.” Then the researchers trained an AI model to recognize these patterns.

Finally, they tested the model’s performance on volunteers whose health conditions were not previously known, and then on volunteers who also had blood tests and who suffered from certain diseases. Using only facial temperature, the AI ​​guessed the age of these people, to within five years.

Additionally, the researchers found that certain conditions – such as diabetes, high blood pressure in women, and liver disease – led to an acceleration of aging-related changes in facial warmth in patterns that their AI (once trained) could detect.

“The method will have to be validated on populations of other origins, insists the researcher to the magazine, But if it is confirmed that it works on larger cohorts, we could envisage fairly quickly and fairly easily deploying an AI capable of detecting certain metabolic diseases from a mapping of facial temperature.”

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button