Algorithm can detect 'Covid-cough' with almost 100% accuracy
Researchers at the Massachusetts Institute of Technology (MIT) have developed an algorithm that they claim can identify coronavirus carriers – even if they’re asymptomatic.This appears to be a major breakthrough, following attempts by other teams around the world to attain a similar goal. In July, a team at the University of Cambridge reported an 80% success rate in diagnosing coronavirus using analysis of breath and cough sounds, but the MIT team claims to have achieved a 98.5% success rate in diagnosing people solely based on what the researchers say is the “distinctive sound” of a coronavirus cough, which is undetectable to human ears.MIT scientist Brian Subirana, one of the co-authors of the paper describing the algorithm, described how, “The way you produce sound changes when you have Covid, even if you’re asymptomatic.”The MIT team collected around 70,000 audio samples for the study, each one containing a number of coughs. 2,500 of the samples were from people who had tested positive for coronavirus.“This was a classic piece of artificial intelligence,” expert Calum Chace told the BBC. “It’s the same principle as feeding a machine a lot of X-rays so it learns to detect cancer. It’s an example of AI being helpful, and for once I don’t see a lot of downside in this.”The MIT scientists would need to gain regulatory approval in order to develop an app. They have suggested that if they achieve that, they could use it “for daily screening of students, workers, and public, as schools, jobs, and transport reopen, or for pool testing to quickly alert to outbreaks in groups.”