With the advent of winter in the northern hemisphere and the increase in the number of confirmed cases in Europe and the United States, more people will probably start to worry about whether they have been infected with COVID-19 or become an "asymptomatic infected person" after a little cough. In response, MIT scientists have developed an AI machine learning system that can be used to determine whether a cough is caused by COVID-19.

It is possible to put it into application in the form of a mobile app.

This AI is divided into three neuron networks, one listens to the sound of your cough itself, one checks whether the respiratory system is weakened (whether the cough is getting weaker and weaker), and the third is to determine whether there is emotional instability caused by neurasthenia . Although it sounds very general, it has a high degree of accuracy in actual measurement. Among the samples known to have COVID-19, 98.5% of the patients can be correctly identified, and among the asymptomatic infected people, there are even more. 100% recognition rate.

Scientists are going to make this technology into an app as a "pre-screening" tool. As long as you let it listen when you cough, you can have a little idea of ​​your risk of disease. Of course, this technology still has its limitations, and it can't replace real antibody rapid screening or DNA sampling, but it may relieve some pressure on medical testing systems (and coughers).

The article is reprinted from the Internet. If there is any issue like copyright or others, please contact: lmy01@gdchico.cn to delete it.

资料来源：AI 可以靠著咳嗽聲來判斷你是否得了 COVID-19

https://chinese.engadget.com/ai-covid-19-cough-detection-from-recordings-100045030.html