Suggested keywords: Coronavirus, Covid-19, AI tool, asymptomatic cases, pandemic, health, tech
Researchers at the Massachusetts Institute of Technology (MIT) in the US have developed an AI tool that can help identify asymptomatic Covid-19 cases from healthy individuals.
The app identifies the cases through cough recordings and displays them on smartphones.
The paper was published in the IEEE Journal of Engineering in Medicine and Biology.
The researchers said in their study that the difference in cough cannot be detected by humans as they are not decipherable. However, by employing artificial intelligence (AI) one can distinguish asymptomatic people from healthy ones.
These recordings were submitted by people voluntarily through web browsers and devices such as cellphones and laptops, the authors stated in their paper.
The researchers designed the model in such a way that it can differentiate tens of thousands of samples of coughs, as well as spoken words.
When the researchers introduced new cough recordings to the model, it accurately identified 98.5 per cent of coughs from people who have confirmed Covid-19.
The researchers noted that this included 100 per cent of coughs from asymptomatic cases.
The researchers believe that the model can be made more user friendly so that it can be adopted on a large scale.
The team wrote that users can log in daily, cough into their phones, and instantly get information on whether they might be infected and therefore, should confirm with a formal test.
Brian Subirana, a research scientist in MIT’s Auto-ID Laboratory, said: “The effective implementation of this group diagnostic tool could diminish the spread of the pandemic if everyone uses it before going to a classroom, factory, or restaurant.”
The researchers have gathered over 70,000 recordings, each containing several coughs, amounting to some 200,000 forced-cough audio samples. This, according to the researchers, is “the largest research cough dataset that we know of”.
The researchers successfully identified patterns in the four biomarkers — vocal cord strength, sentiment, lung, respiratory performance, and muscular degradation — that are associated with the coronavirus.