Researchers at MIT have created a tool using neural networks that can detect subtle changes in a person's cough and identify whether a person has been infected with COVID-19, even if they don't have any other symptoms.
In tests, the tool was able to recognize 98.5% of coughs from people with confirmed cases and 100% of coughs from asymptomatic people.
Asymptomatic carriers of the virus are less likely to get tested so they are hard to manage during a pandemic.
While the researchers warn that the method shouldn't be used as a substitute for getting properly tested for the virus, its use could help improve detection rates.
Advertisement
This content is brought to you by bits.ie
