AI can catch COVID-19 cases based on how people 'produce sound,' MIT researchers find
Researchers at the Massachusetts Institute of Technology have published a paper suggesting artificial intelligence can recognize potential coronavirus cases based on how people sound when they cough, PCMag reports.
The MIT team employed an AI model that was able to accurately identify 98.5 percent of coughs from people who were confirmed to have been infected by the coronavirus. The accuracy actually ticked up to 100 percent among people who were not displaying symptoms. "We think this shows that the way you produce sound changes when you have COVID, even if you're asymptomatic," Brian Subirana, the paper's co-author, said.
The AI was already being used to analyze forced-cough recordings to catch signs of Alzheimer's, and it turned out the researchers didn't have to alter too much to try to pick up patterns specific to the coronavirus, as well.
The Week
Escape your echo chamber. Get the facts behind the news, plus analysis from multiple perspectives.
Sign up for The Week's Free Newsletters
From our morning news briefing to a weekly Good News Newsletter, get the best of The Week delivered directly to your inbox.
From our morning news briefing to a weekly Good News Newsletter, get the best of The Week delivered directly to your inbox.
Now, the MIT group is working to build an app that uses the AI model that could, in theory, serve as a free, non-invasive way to pre-screen for COVID-19. The app would not serve as a diagnostic tool itself, but would instead determine if a person should take a formal test. Read more at PCMag.
