An app that will detect if we have covid-19 just listening to our forced cough

There has come a time when it is "easy" for artificial intelligence to identify a Covid-19 infection simply by listening to a forced cough, and that ability can be used to identify them over the phone.

It is a project by MIT researchers, who have discovered that asymptomatic people in Covid-19 may differ from healthy people in the way they cough. These differences are not decipherable to the human ear but can be detected by artificial intelligence.

Asymptomatics may not be completely free of changes caused by the virus, as they cough differently. In the article recently published in the IEEE Journal of Engineering in Medicine and Biology, the team reports on an artificial intelligence model that distinguishes asymptomatic people from healthy people through forced cough recordings, which people voluntarily send through web browsers and devices such as cell phones and laptops.


The ultimate goal would be for there to be an app on mobile that listened to our forced voice, getting the result of its own COVID-19.

Researchers trained the model in tens of thousands of cough samples as well as spoken words. When they fed the model with new cough recordings, they accurately identified 98.5 per cent of the coughs of people who were confirmed to have Covid-19, including 100 per cent of asymptomatic coughs, which reported no symptoms but had tested positive for the virus.

To date, researchers have collected more than 70,000 recordings, each containing several coughs, equivalent to about 200,000 samples of forced cough audio. They found that the system could detect patterns in the four biomarkers (vocal cord strength, feeling, lung and respiratory performance, and muscle degradation) that are specific to Covid-19.

The strength of the tool lies in its ability to distinguish asymptomatic cough from healthy cough.

They are already working on an easy-to-use application, which if approved by the FDA and adopted on a large scale, could be a free, convenient and non-invasive preselection tool to identify people who are likely to be asymptomatic for Covid-19.

Post a Comment

0 Comments