Detecting Covid-19 by sound

The University of Cambridge and Carnegie Mellon University are seeking to independently develop a project application to detect Covid-19 by listening to the cough and the path of users.

However, the two projects take different approaches to privacy protection.
The University of Cambridge seeks to keep the anonymity of volunteers anonymous, but believes this currently limits its work.

For its part, Carnegie Mellon University says it is essential that users register themselves.

Both initiatives rely on machine learning, in which computers analyze large amounts of data to find models that can be used to solve problems.

The goal is to be able to distinguish Covid-19 from other illnesses, such as the flu or a simple cough.

Participants are asked to breathe and cough into a computer microphone, as well as provide details about their age, sex, location and whether they have recently tested positive for the coronavirus.
They are then asked to read the following sentence three times: “I hope my data can help manage the virus pandemic.”
The goal is to collect enough data to see if the algorithm is able to diagnose people with Covid-19 and possibly the stage of the disease. The analysis shouldn’t take too much time, but it all depends on the quality of the data collected.
On the first day, about 1,200 people provided records, 22 of whom reported having recently tested positive.

For the time being, the project is limited to collecting samples via a website, rather than a smartphone application because Apple and Google limit the number of applications related to coronaviruses, and this application has not yet been validated by both platforms.

The initiative is interesting but the resulting software will not replace the need for further medical testing.

M.D