Chinese AI technology shows people age fastest in their 40s


  • AI
  • Monday, 21 Sep 2020

Researchers used 3D images to determine how much younger or older people look compared to their actual age. System based on images and health information about 5,000 Beijing residents. — SCMP

Chinese researchers say they have developed an artificial intelligence technology that can calculate a person’s biological age from a 3D image.

Biological age is a measurement of age based on various biomarkers, and can change due to lifestyle and other health factors.

The team from Peking University and Shanghai Jiao Tong University said they found the signs that someone was either a “fast ager” – with a biological age older than their numerical age – or a “slow ager” was most apparent between the ages of 40 and 50.

“Middle age is the ideal stage for ageing interventions,” team leader Han Jingdong from Peking University said in a paper published in the peer-reviewed journal Nature Metabolism this month.

In a press release posted on the website of the Chinese Academy of Sciences on Tuesday, the team said their study showed that yoghurt, coffee, fruits, chicken, beans and “eating on time” could help to slow down ageing.

In contrast, smoking, drinking alcohol, eating pickled food, taking antibiotics and exposure to ultraviolet could make people look older, which was backed up by the results.

The system was built on a database containing 5,000 people’s facial features in three-dimensions and related health information.

The team tested it on several hundred people in Beijing and found it not only achieved a smaller margin of error between the subjects’ numerical and biological age – 2.7 years – but also explained why people were ageing differently better than other methods used to do so, such as blood tests.

The system allowed Han’s team to build a biological ageing portfolio of people at a relatively low cost. The results showed there are “alternative steady states” and ageing interventions during those periods might be compromised by other factors such as growth hormones or diseases that tended to occur in older populations, they said.

The technology is based on an AI algorithm known as deep convolutional neural networks, could be used as a “superior health estimator”, the study said.

A study by researchers at Tsinghua University in 2017 found that some fast-grow genes in childhood could prolong health and life. A large-scale study on young people in the US, Britain, Israel and New Zealand found that ageing could start as early as 26.

In an experiment conducted by researchers in San Diego in 2018, volunteers were shown the images simulated by computer of their face five years later with or without regular sun tanning. Most volunteers decided to stop going to the beach after seeing what they might look like if they did not.

Han’s study said the new technology could help to manage China’s ageing problems but said the biometric information would have to be well protected.

“It should be carefully guarded against any unethical use,” the paper said. – South China Morning Post

Article type: metered
User Type: anonymous web
User Status:
Campaign ID: 18
Cxense type: free
User access status: 3

SCMP , aging

   

Did you find this article insightful?

Yes
No

100% readers found this article insightful

Across the site