BMJ | 2021

Sixty seconds on . . . cough in a box

 

Abstract


Abi Rimmer Sounds like a terrible gift Don’t worry, we’re not talking about wrapping up covid symptoms. This is the news that the UK government is looking into whether artificial intelligence (AI) can be used to detect covid-19 by analysing the sound of someone’s cough. And the box? I’m not quite sure where that fits in. The website Politico, which first reported the story, said that the Department of Health and Social Care has awarded two deals to Japanese tech firm Fujitsu to help it build an online “cough in a box” platform.1 Awhat? The health department said that the Joint Biosecurity Centre is working with the Alan Turing Institute and theRoyal Statistical Society to assess the feasibility of “applying algorithms to detect covid-19 based on vocal biomarkers including coughing.” So they listen to coughs? Essentially, yes. In a paper published in the IEEE Journal of Engineering in Medicine and Biology2 last year, a team of researchers from the Massachusetts Institute of Technology (MIT) in the US described an AI model they had developed which distinguishes the coughs of healthy patients from those with covid-19. Do I cough on amachine? No, the MIT researchers used forced cough recordings, which people voluntarily submitted through web browsers from their mobile phones and laptops.3 Before the pandemic, the researchers had been developing algorithms to diagnose conditions such as pneumonia and asthma from phone recordings of coughs. Somaybe it’s a phone box? Clever, but the technology is far more advanced than that. According to Politico, cough in a box will be a webplatform that can collect audio recordings andbe linked toQR codes of the kindusedby theUK’s covid-19 app. Futuristic! The health department seems to hope so. Commenting on the plans, it said that the UK “is at the forefront of innovative research to expand our collective understanding of covid-19.”

Volume 373
Pages None
DOI 10.1136/bmj.n1452
Language English
Journal BMJ

Full Text