An algorithm has been created that detects coronavirus by coughing


Scientists working at the Massachusetts Institute of Technology have managed to develop a special algorithm that can distinguish between the cough of a healthy person and the cough of a patient with a coronavirus. This will greatly help in recognizing infected people who are asymptomatic and unknowingly spread the virus in the community. The algorithm will make it possible to recognize such people in advance, which will make it possible to reduce the risks of the spread of the virus.

Even before the start of the pandemic, these same scientists were working on creating an artificial intelligence model capable of recognizing various diseases by the human voice. For this, 200,000 recordings of voices and coughs were used, which was needed to train the algorithms. This made it possible to obtain an excellent result, since artificial intelligence eventually learned to distinguish between the speech and cough of sick people from healthy people.

Since the coronavirus has similar symptoms to colds known to the world, the algorithm began to cope with the recognition of this virus as well. At the level of neuromuscular disorders, taking into account many other factors, the algorithm detects the presence of the virus even in asymptomatic patients.

Now this testing model is being calibrated and tested on a large number of people in order to achieve the most ideal result. So far, the algorithm with an updated database, where records of new voices and coughs have been added, shows the result with an accuracy of 98.5%. A mobile application is also being developed that will help everyone determine the fact of infection by coughing. This will allow self-diagnostics for everyone, just by coughing into the microphone of their smartphone.

Rate the article
Share with friends
Coronavirus Online
Add a comment

By clicking on the "Send comment" button, I consent to the processing of personal data and accept the privacy policy .

For any site suggestions: [email protected]