A voice-based AI diagnostic application for Covid-19 infection is under development and is expected to effectively support a large number of users.
Scientists at the University of Cambridge are calling on volunteers to send audio data when breathing, coughing and talking to support their new Covid-19 diagnostic research.
This application is called Covid-19 Sounds. Although it has not been proven specifically, scientists believe that using sound analysis algorithms in the respiratory process can detect early symptoms of infection.
“During the examination, the doctor mostly listens to how the patient breathes, the rate at which he breathes and has a dry cough,” said Professor Cecilia Mascolo, head of the development team.
AI technology can detect even the smallest changes in the voice of an infected person. Currently, the data warehouse of this study still needs to supplement various sources from volunteers.
The development team will build an application similar to the Covid Symptom Tracker (Covid-19 symptom tracking software) that has been released for British people to take daily health checks.
Users will fill out a questionnaire about medical history and demographics. They then upload voice samples, including coughing, breathing, and saying a sentence through the phone microphone. Finally, users just need to add the location and press the submit button. The whole process takes about a minute.
Currently, Covid-19 Sounds application is available on Chrome, Firefox with versions for Android, iOS. Collected data will be stored on Cambridge University’s servers and only used for research, not tracking or retrieving personal data for other purposes.
Source: Zing News