Artificial intelligence (AI) can be used to detect Covid-19 infection in the voice through an app.
This is stated by researchers from the University of Maastricht, in the Netherlands, who presented a study a few days ago at the International Congress of the European Respiratory Society in Barcelona.
Researchers say the AI model used is more accurate than rapid antigen tests, 89% of the time, as well as faster and easier to use.
The team worked on data from the University of Cambridge's Covid-19 Sounds app: it contains 893 audio samples from 4,352 healthy and sick participants, of which 308 tested positive for the virus.
Users were asked to cough three times, breathe deeply through their mouth three to five times, and read a short sentence on the screen three times.
The researchers used a speech analysis technique called "Mel Spectrogram" to identify different characteristics of the voice such as volume, variation and power.
"These results suggest that voice recordings and improved AI algorithms can potentially achieve high accuracy in determining which patients have the infection. They could be used, for example, at entry points for large gatherings, enabling rapid screening." explains Wafaa Aljbawi of the Institute of Data Science at the University of Maastricht.
In Italy, the University of Ferrara developed a system a year ago that is always based on coughs and sound waves, thus identifying the "voice" of the disease.
The recognition method is based on a technique called supervised modal learning.