MIT Researchers Train AI to Identify Covid-19 in Cough of Asymptomatic People, Mobile Apps Incoming


October 30, 2020

Promo Protect all your devices, without slowing them down.
Free 30-day trial
MIT Researchers Train AI to Identify Covid-19 in Cough of Asymptomatic People, Mobile Apps Incoming

Massachusetts Institute of Technology (MIT) researchers demonstrated how they have trained an artificial intelligence model to identify people infected with Covid-19 just by listening to a phone recording of their cough.

It’s exciting to witness the power of IoT. While the spotlight usually belongs to cool gadgets, research like that coming out of MIT shows just how powerful the technology can be. Screening for possible COVID-19 infections is difficult, especially among asymptomatic people. But it’s not impossible when a powerful AI enters the scene.

It turns out that healthy people don’t cough in the same way as infected but asymptomatic people. While people’s ears are not trained to spot differences in coughing sounds, it is possible to program an AI to do so.

According to MIT News, MIT researchers published a paper in the IEEE Journal of Engineering in Medicine and Biology, reporting on an AI model that distinguishes asymptomatic people from healthy individuals through forced-cough recordings.

The AI successfully identified 98.5 percent of coughs from people who were confirmed to have Covid-19, including 100 percent of coughs from asymptomatics. The latter group had no symptoms but already tested positive.

The team is now working on building a simple app for smartphones that would let users simply cough, record it, and let the software analyze it.

“The effective implementation of this group diagnostic tool could diminish the spread of the pandemic if everyone uses it before going to a classroom, a factory, or a restaurant,” said co-author Brian Subirana, a research scientist at MIT’s Auto-ID Laboratory. The other two scientists involved in the project are Jordi Laguarta and FerranHueto.

Interestingly, the software was designed initially to distinguish voice patterns in Alzheimer’s patients. Researchers adapted the neural network for the new purpose. Also, it’s essential to understand that it’s not a diagnostic tool but to identify an unhealthy cough, which in this situation would apply to other health issues like the flu or asthma.




Silviu is a seasoned writer who followed the technology world for almost two decades, covering topics ranging from software to hardware and everything in between.

View all posts

You might also like