Dutch scientists say they have developed technology that is said to highlight those infected with COVID-19 by scanning their voice, according to Imperial College London.
It has a reported accuracy of 89% for positive cases and 83% for negative cases.
Aside from supposedly being more accurate than lateral flow tests, the technology can detect the virus in less than a minute.
The researchers said they decided to develop the software because Covid affects the vocal cords and the upper respiratory tract, meaning it changes a person’s voice.
In a world where some people found to be carrying the virus are forced into quarantine, the technology has major implications.
Researchers used data from University for Cambridge’s crowdsourcing COVID-19 Sounds App, using 893 audio samples from over 4,000 participants.
Using a voice analysis technique called Mel-Spectrogram analysis, the researchers used an artificial intelligence model to distinguish voices of people with Covid and those without. A model called Long-Short Term Memory (LSTM) provided the best results.
“These promising results suggest that simple voice recordings and fine-tuned AI algorithms can potentially achieve high precision in determining which patients have COVID-19 infection”, said Wafaa Aljbawi, one of the researchers....
English (US) ·