COUGH
The world's leading infectious killer, COVID-19, overtook TB at the worst of the epidemic, killing an estimated 1.5 million people in 2020 and 1.4 million in 2019. (Photo: Annie Spratt/Unsplash(CC BY 2.0))

A group of researchers from MIT believe they unearthed a novel way of detecting a possible existence of COVID-19 infection without employment of an invasive technique. It only involves an iPhone app, which the ressearchers said is capable of analyzing the sound of someone's cough.

There is a new way of detecting possible COVID-19 infection quickly and it has something to do with analyzing cough sounds. MIT researchers made use of an intelligent machine learning technology to develop an iPhone app, which could ascertain if a person has COVID-19 infection just by evaluating how their cough sounded like.

The failure to test at scale has the Achille's heel of mankind in the raging war against the COVID-19 pandemic. Scientists know that coming up with a screening tool or app that is both scalable and accurate would be a game-changer. MIT researchers explored this possibility and recently, they announced they were able to develop an Artificial Intelligence (AI) screening app for iPhone that can analyze cough sounds and determine if a person is infected with the virus. Best of all, it can generate a result within two minutes and can help a person decide whether or not to seek a doctor's appointment and undergo a coronavirus test.

MIT researchers used an already existing iPhone app which lets users check the possibility of contracting COVID-19 after coming into close contact with a positive individual. If the researchers' discovery, which now gives the app the capability to detect possible COVID-19 infection through cough sound analysis by making use of biomarkers, is developed further, your iPhone could become an important tool, not just in curbing the spread of the virus, but also giving you the chance for early treatment, thereby increasing your chances of survival.

  The researchers published their study in the IEEE Journal of Engineering in Medicine and Biology. They claim that they were able to create an AI model that can differentiate health people from asymptomatic persons by analyzing cough sound recordings. According to researchers, their iPhone app, when analyzing cough sounds from COVID-positive people, is accurate at 98.5% of the time. It is also 100% accurate when examining cough sound recordings of asymptomatic people.

Researchers say, however, that people should not only rely on their developed AI and should still visit COVID test facilities to see if they are really infected. Some people questioned why not just go directly to these testing sites, to which the researchers replied that the virus has spread to an extent that when an undiagnosed person coughs, this contributes more to the covert and rapid spread of the virus. Furthermore, 81% of COVID-positive individuals do not develop severe symptoms that necessitate medical help, though they can become active spreaders. With the iPhone app, they developed which can determine whether or not a person has contracted COVID-19 just by analyzing cough sounds, only those who are likely infected can seek an examination at COVID testing sites.