Artificial Intelligence Can Reveal Your Body’s ‘Real’ Age – University of Copenhagen

Forward this page to a friend Resize Print Bookmark and Share

Health > News > 2018 > Artificial Intelligence

11 January 2018

Artificial Intelligence Can Reveal Your Body’s ‘Real’ Age

Artificial Intelligence

New artificial intelligence can determine your biological age with great precision. It can reveal whether lifestyle changes and medicinal products can increase your chances of living a long and healthy life. This is the result of new international research conducted in cooperation with the Center for Healthy Aging at the University of Copenhagen.

New artificial intelligence can help us determine whether our bodily age corresponds to our actual age and our risk of developing age-related diseases. If the algorithm finds that you are younger than the age shown on your driver’s licence, you have increased chances of living a long life, new research shows. It can be used to assess the effect of medicinal products and other procedures on our general physical health by measuring a person’s biological age before and after receiving medical treatment, says Doctor and Assistant Professor Morten Scheibye-Knudsen, who is also a part of Department of Cellular and Molecular Medicine. He has contributed to the creation of the artificial intelligence alongside researchers from the US, South Korea and Canada.

‘Historically a lot of money has been spent trying to identify biomarkers of ageing, and by and large all attempts have failed. But using these large complex data sets containing a number of interacting parameters, we are able to create a fairly accurate picture of a person’s biological age’, says Doctor and Assistant Professor Morten Scheibye-Knudsen from the Center for Healthy Aging and the Faculty of Health and Medical Sciences at the University of Copenhagen. He is one of the authors of a new research article describing how the research team has tested the artificial intelligence on a series of blood samples and shown that the artificial intelligence arrives at an age that is fairly close to the person’s actual age. The research article is a so-called ‘proof of concept’ showing that the algorithm is in fact able to determine a person’s age based on a blood sample.

’The artificial intelligence is just as good at predicting your age as if you looked at a picture of the person and had to guess the person’s age. That is, it arrives at an age that is fairly close to the age shown on your driver’s license, but what it really does is measure your biological age, which may be different from your actual age’, says Morten Scheibye-Knudsen and adds that Insilico Medicine, an American company specializing in artificial intelligence, has primarily run the project..

Simple and Cheap Method
The algorithm behind the artificial intelligence requires measurements already provided by a standard blood sample taken by your GP. 

‘What is special and impressive about this method is that it uses measurements gained from standard blood samples, which are cheap to perform. There are other ways of measuring a person’s age, but they are relatively expensive and require special equipment’, Morten Scheibye-Knudsen explains and continues:

‘Right now we actually have no idea how old we will be. Of course we have hypotheses and assumptions, but we also know that age is the greatest risk factor when it comes to widespread diseases like cancer, dementia and cardiovascular diseases. We would like to be better at curing age-related diseases, and artificial intelligence can help us do that’, says Morten Scheibye-Knudsen.

Artificial Intelligence Resembles the World’s Best Chess Computer
The artificial intelligence, Aging.AI, is based on deep, neural networks, which is a type of learning algorithm resembling the one used to develop the revolutionary artificial intelligence AlphaZero, which just beat the world’s best chess computer without having been taught anything but the rules of the game.

Using Aging.AI the researchers have fed the algorithm the age and blood sample answers of 130,000 individuals. Each sample contains 21 parameters typically measured in blood samples like the ones taken by your GB, e.g. cholesterol, inflammation markers (CRP), haemoglobin count and albumin in the blood.

Center for Healthy Aging is supported by the Danish foundation Nordea-fonden.

The new research article is published in the Journal of Gerontology: Biological Sciences.

Contact:
Doctor and Assistant Professor Morten Scheibye-Knudsen
Email: mscheibye@sund.ku.dk
Phone: +45 42 50 70 60