The steps in the app named as myCOPD, involves it asking the user to cough, breathe deeply and then reading a short sentence three times. 

According to reports, the app demonstrated an 83 per cent of ratio while identifying positive cases. On the other hand it identified negative cases 83 per cent of time. Both these reports were identified in less than a minute. 

Wafaa Alijbawi, researcher from the University of Maastricht, in the Netherlands, asserted that a user can receive the tests at no cost and are simple to interpret. He expressed its possible effective use at entry points for large gatherings, enabling rapid screening. 

She stated that while lateral flow tests missed 44 out of 100 positive cases, the myCOPD app would just miss 11. However, the app might also falsely detect 17 in 100 non-infected people. 

Cambridge University collected 893 audio samples from 4352 participants. This even included 308 covid patients. The app used this data. Covid changes a person’s voice by affecting the upper respiratory tract and vocal cords. 

The research process for this study in developing the app, used a voiced analysis technique called Mel Spectrogram. 

The developers developed an artificial intelligence system based on neural networks, which mimic how the human brain operates, to facilitate distinguishing Covid patient voices from non-sufferers. 

The post myCOPD: researchers develop AI-powered Covid-19 detection app appeared first on The Tech Outlook.