Article by Vivion O’Kelly
Evidently, we should suspect any product that claims to diagnose or treat too many illnesses, that can be used anywhere easily by any medical non-professional, that uses personal testimonials instead of scientific data, that promises immediate results, that uses the word “miraculous” in its description, that is free and easy to download or that simply appears to be too good to be true. Many individuals and companies are attempting to benefit financially from the current Covid-19 pandemic by promoting or selling fraudulent products which may be dangerous to use. Some of these are test kits sold online.
The U.S. Food and Drug Administration’s official website informs us that, in content current as of 10/22/2020, “the only way to be tested for COVID-19 is to talk to your health care provider”. The World Health Organisation makes no mention of any new testing app in their Covid-19 web page.
On the other hand, Dr Edward Jenner, creator of the world’s first vaccine and saviour of the lives of countless millions of people worldwide since the end of the 18th century, might well have fallen foul of FDA testing guidelines, not to mention the mother of the eight-year-old boy he used as a guinea pig.
And doctors have been telling us to cough for a very long time.
So perhaps we should take a closer look at an apparently miraculous telephone app that promises all the things we should be suspicious of. You download the app, cough into it and receive a diagnosis within minutes. Too good to be true? Maybe not.
The idea has been around for a while. The way we cough, as a result of an illness or forced, and even the way we make certain sounds, can tell a trained professional a lot about the state of our health. Add artificial intelligence and you have a more accurate diagnosis, and this is what the app does. Genius is often simple.
Researchers at MIT collected tens of thousands of coughs that people submitted through web browsers and phones, and when they fed these recordings into the AI model, they were able to identify 98.5 percent of coughs from people who were confirmed to have Covid-19, including 100 percent of coughs from asymptomatics, who reported that they did not have symptoms but tested positive for the virus.
The AI model is based on one that existed for Alzheimer’s research, and was found capable of picking up four biomarkers relating to vocal cord strength, sentiment, lung and respiratory response, and muscular degradation specific to Covid-19.
The next step is to create an app that can be easily used by anyone anywhere, the enormous benefits of which are obvious. Regulatory approval is needed to develop it, but given that coughing into a mobile phone cannot cause physical harm in any way, and that a smartphone app is already available to help doctors diagnose common respiratory diseases remotely in Australia, approval should not be a serious block to its eventual widespread use.
Similar research has been carried out in other countries across the world, including Britain, where researchers from Cambridge University reported a 98.5 percent success rate in identifying positive coronavirus cases with symptoms, based on a combination of breath and cough sounds, and a 100 percent success rate without symptoms, meaning this app can detect both.
Not being a vaccine, it solves only part of the problem, but an important part, giving a new meaning to test and trace. It is also not suitable for diagnosing symptomatic people, as they might have other conditions that produce similar behaviour, although the real strength of the tool lies in its ability to discern asymptomatic coughs from healthy (forced) coughs.
Every effort was made before writing this article to unearth fraudulent reporting, but all information debunking the new app was unsubstantiated or appeared in anonymous websites. The story of its development was recently reported with a positive slant on many major news websites, including the BBC, Apple Insider and CBS.
Will this mean the end of pandemics worldwide? Maybe, in that it would help reduce cases to such an extent that the WHO would no longer be obliged to define Covid-19 as a pandemic. At the very least it would go a long way towards identifying carriers of a virus and dealing with outbreaks before they spread too far, and at best it would indeed be the biggest breakthrough so far in the fight against this awful illness.
It should be added that anyone who wishes to support Cambridge University’s efforts in developing their app can do so by clicking on https://www.covid-19-sounds.org/en/ and following the instructions given.