The pandemic caused by the presence of the SARS-CoV-2 virus in our society has been going on for almost a year. It is not an easy fight, among other things due to the potential ease of getting infected. Governments around the world have a problem testing all suspects who may be infected today. The search for new ways of examining patients is ongoing, and it turns out that one of them may be quite unusual. We are talking about a special application that would indicate that a person could potentially be sick with COVID-19.
The Institute of Electrical and Electronics Engineers has published a report indicating that machine learning can be used to analyze the cough of people who suspect they may be sick with COVID-19. Machine learning is supposed to distinguish forced coughing in healthy people as well as in people who are currently sick. The app's team collected 70 000 recordings of coughs from around the world, plus recordings of 200 000 forced coughs. It turned out that the application was 98.5% effective when it comes to indicating that a person has COVID-19. So the results are really impressive, but for now we will have to wait a little longer for the application to be finished. Ultimately, the application is planned to be available for free.