While the coronavirus epidemic has been going on for more than a year, it has not been so far in detecting the disease. For COVID-19While the symptoms caused by is different from person to person, many people do not even show symptoms. For this reason, some health technology companies are also developing simple solutions to predict COVID-19 using data such as lung image, sound and cough.
You may be interested
The number of new cases decreased slightly (June 26)
See Also. “Huge data leak on Facebook: 19 million Turkish Facebook users’ personal data spread to the Internet!”
The last of these is Vocalist It was an artificial intelligence supported application developed by a company named. Using audio, the app can predict if someone is infected with COVID-19.
Could predict coronavirus 80% correctly
The application extracts 512 features from a sound sample and converts it to the image, by identifying vocal biomarkers such as fever, headache and shortness of breath from this image with the help of artificial intelligence. to predict coronavirus works.
Trained with the voices of more than 275,000 people speaking different languages, the algorithm is very successful in predicting coronavirus. So much so that in the research conducted on 2000 coronavirus cases, many of which did not show symptoms, the algorithm was able to detect COVID-19. 81.2% successful happened. Wishing to improve the algorithm’s success in accurately detecting the disease, the company wants to commercialize the application and make it available for more regions.