Help fight the pandemic

Record your cough sounds and answer a short survey to participate. Your contribution will help build an AI-powered cough analysis technology that could assist in rapid and early diagnosis of COVID-19

#CoughAgainstCovid

Here’s another way in which you can help. Help us reach people who have been tested for COVID-19. Please share widely in your networks.

     

Why participate?

Donate your cough sounds to science

We are collecting recordings of cough sounds, especially from those who have been tested for COVID-19. You can help from home, we need just five minutes of your time. You can use a mask while recording

Help train AI that can recognise Covid coughs

Once we have enough submissions, we will train an AI algorithm to analyze the cough recordings we have collected to try and find the early signs of COVID-19 in a cough sound

Help build an early diagnosis tool for
COVID-19

This technology could potentially help millions of people with limited access to testing facilities get an early diagnosis. All they will have to do is record their coughs and enter their symptoms

Contribute to an open access, anonymised dataset

We believe in open innovation. The data we collect will be open access and available to researchers across the world (after proper anonymisation). Any technology we develop will be open source and free-to-use

An initiative by
Supported by
In collaboration with

How to participate

Your contribution could help millions of people with limited access to testing facilities get an early diagnosis, and enable open innovation around COVID-19 from the medical and AI research community.

Who can contribute?

We are primarily interested in collecting cough sounds from those who have taken a COVID-19 test at a medical facility (both positive and negative test results). You can still contribute If you haven’t been tested yet — every cough sound submission is valuable.

How long will it take?

We need approximately 5 minutes of your time.

What you'll need to submit

Basic details

Answer a few questions about yourself, your symptoms and medical history

Picture of COVID-19 test result

If you have the result on-hand, please click a picture and share it with us. Don’t worry if you don’t have this, you can still contribute. (See our note on privacy.)

Sound recordings

We need 3 samples of cough sounds and a simple voice recording

Share this

     

About

The Wadhwani Institute for Artificial Intelligence is an independent, nonprofit research institute and global hub, developing AI solutions for social good.

To control the COVID-19 pandemic, it is crucial to identify positive cases early and break the chain of transmission. However, our testing capacity remains limited.

We want to build an AI-powered cough analysis technology that could potentially help in the early diagnosis of COVID-19. 

But first, we need to build a large dataset of cough sounds from COVID-19 positive patients as well as those who have matching symptoms. The anonymised dataset will be open access and available to the research community.

Team
We are a global initiative led by Wadhwani AI, a nonprofit research institute focused on AI for social good. This initiative is supported by the Bill and Melinda Gates Foundation, and collaborators include Stanford University. Collectively, we have world class capabilities in AI, medical technology, public health, and large-scale global deployment

Approach
In order to build this AI-powered technology, we are collecting cough sounds and other symptoms from people who have already been tested for COVID-19 or have some matching symptoms. We are also preparing to collect data from COVID-19 testing facilities and hospital isolation wards in multiple geographies

Open Innovation
Any technology we develop and the underlying algorithms will be open source and free-to-use. We will make the dataset we collect open access and available to researchers across the world, after suitable anonymisation

Share this

     

Privacy

To control the COVID-19 pandemic, it is crucial to identify positive cases early so that the chain of transmission can be broken. However, testing capacity across the world remains limited. There is a need for rapid early diagnosis tools that can help patients with mild or moderate symptoms get tested as well. 

One of the first symptoms of COVID-19 is a cough. Cough sounds carry vital information about the state of the respiratory tract. Doctors routinely use the acoustics of coughs as a diagnosis tool. Cough is a symptom of many other illnesses — from allergies to bronchitis — but anecdotal evidence suggests that a COVID-19 patient’s cough sounds different from other coughs. We want to study the acoustic signatures of coughs to try and find the early signs of COVID-19, for which we need to collect a large number of cough sounds.

We will use lab reports to verify if the samples came from a participant who is COVID-19 positive or not. Note that you may also submit a redacted version of your lab report (with private information blacked out or hidden), or the name of the facility where you were tested. The lab reports will only be accessible to select people from the core research team, with strict access control and access logging. These reports will not be a part of the open access dataset.

The service we use for data collection, JotForms, is HIPAA-compliant. Data is stored in a HIPAA-secure application, with access control and access logging.

It means that the datasets that we collect from this study will be made available to researchers across the world free of cost. We will validate and anonymise the data we collect before we make the dataset open. For example, we will use lab reports to verify if the participant is COVID-19 positive or not, but not include the reports themselves in the final dataset.

The only contact information we ask for is your email address, and it will not be shared with anyone. We will use it to follow up with you if you have not been tested or haven’t received your test results at the time of filling the survey form, or to send you a reminder email about recording your cough sounds every day so we can study the progression of cough sounds and symptoms.

Contact Back to top

Do you have any comments or feedback for us? Get in touch!

Or find us on: