As COVID-19 is still spreading in many countries, quickly finding people who have the virus is very important. But COVID-19 is not easy to detect. Many people who catch the virus clearly show signs of being ill, like having a fever or coughing. But many others show no signs at all. These people are "asymptomatic" and they can still pass the virus to others.
Now, scientists from Massachusetts Institute of Technology, a famous American college, have developed a method which can quickly detect whether a person has COVID-19 or not just from the way he or she coughs or speaks. The scientists turned to an AI method known as "deep learning".
In deep learning, computer programs collected a lot of information, analyzed（分析）it and built certain models. The information here was tens of thousands of recordings of coughing and speaking from people who had caught the virus. When the scientists tested their AI models, they found that they could detect 98.5% of the people who caught the virus, including 100% of the asymptomatic people.
Brian Subirana, one of the scientists, says, "We think this shows that the way one person produces sound changes when he or she has COVID-19, even if he or she is asymptomatic." The scientists now want to turn their models into an app that can run on mobile phones. If they are able to do that, it could allow people to quickly and easily check whether they have caught the virus.
