Although the new coronavirus is not as powerful as SARS in 2002, its terrible feature is that it has been latent for a long time without symptoms, which makes the infected people spread the virus without knowing it. If there is no strong means of prevention, the spread of the virus is almost impossible to control. The most important measure to prevent large-scale spread of the virus is to detect asymptomatic infected people. Proper isolation and observation of such people is an effective way to avoid large-scale infection again.
So far, nucleic acid detection is widely used as a discovery measure. In addition, there is no particularly good low-cost and fast detection method. But recentlyThe MIT News Office has released a new detection method, which only requires patients to collect cough sounds from their mobile phones and then use artificial intelligence models to determine whether they have new coronavirus。
The model distinguishes asymptomatic people from healthy people by submitting cough recordings using browsers, mobile phones, computers and other devices. The researchers trained tens of thousands of cough samples and oral language using neural network models. When they entered new cough records into the model, the model accurately identified 98.5% of the cough diagnosed with 2019 coronavirus disease, including 100% of asymptomatic cough (these cough uploaders were asymptomatic but tested positive for the virus).
Currently, the team is working on integrating the model into a more user-friendly application, which could become aFree, convenient and non-invasive testing tool。 Users can log in every day, cough into their phones, and immediately get information about whether they might be infected.
The research paper is published in the Journal of IEEE medicine and bioengineering. If you are interested in this paper, you can pay attention to my official account: “program ape DD”, send the keyword “cough test” to get the paper PDF.
Welcome to my official account: program ape DD, get exclusive learning resources to help you learn Java. In addition, I give books every week