Update: CoughVid app

In response to the COVID-19 pandemic, five researchers at EPFL's Embedded Systems Laboratory (ESL) have created CoughVid, an artificial intelligence (AI)-based app that can detect from the sound of your cough whether you have COVID-19. In fact, 57% of patients with COVID-19 experience a cough as a symptom¹. Historically, listening to patients' coughs has been an essential part of in-person diagnosis by doctors in detecting respiratory illnesses.

The app, launched in April 2020, has already gained traction with the media and the public. Thanks to the high number of participants who submitted audio recordings of their cough, Dr. Tomás Teijeiro, an ESL postdoctoral researcher, and his team have gathered a dataset of more than 19,000 audio recordings from which to base their analysis.

With the help of three medical doctors with experience with COVID-19 patients, the EPFL researchers have developed an algorithm that can automatically assess the recordings.

As a next step, the team will continue to test different AI-based models to analyze cough sounds. Also, once the data curation and statistical analysis of the data are completed, Teijeiro and his team will publish all recordings as an open dataset for others to experiment with.

If you wish to contribute to the development of the app, please send a recording of your cough via CoughVid.

HelpfulETH: an initiative bringing engineers together to fight the pandemic. Coughvid is one of around a dozen projects being carried out under HelpfulETH, a joint initiative of EPFL and ETH Zurich aimed at developing solutions to help hospitals and other health-care providers fight COVID-19. The 100 students, PhD researchers and professors taking part in the initiative are putting their engineering knowledge to work to support the medical community. Additional projects could soon be added to the mix.²