MIT Researchers to develop tool that uses AI to detect Covid by listening to humans cough

As the world continues to battle against the COVID-19 pandemic, the number of asymptomatic cases have been rising significantly. In order to detect these asymptomatic Coronavirus cases, researchers at the Massachusetts Institute of Technology (MIT) have said that an artificial intelligence (AI) tool can detect if there are any COVID-19 indications in a human's cough even before seeking medical help and further testing. The researchers informed that asymptomatic people are different from healthy individuals in the way they cough and while differences are not decipherable to the human ear, AI can always detect the difference. 

READ | AI Camera Follows Soccer Ref’s Bald Head Mistaking It As Ball, Netizens Say ‘has A Crush’

MIT Researchers: AI tool can help detect COVID-19

MIT Researchers further stated that they aim to make this AI tool available to the users in their phones as a mobile application. This mobile application will be FDA approved as well. As per the information given in the paper published in the IEEE Journal of Engineering in Medicine and Biology, the researchers claim that the AI tool can detect 'induced or forced cough recordings' via mobile or a laptop. The researchers also asserted that so far, the AI tool has successfully identified 98.5 per of coughs from people who were confirmed to have contracted Coronavirus. 

READ | US Election 2020: Trump's COVID-19 Vaccine Before Nov 3 Promises Blasted By Health Experts

Co-author Brian Subirana who is a research scientist in MIT's Auto-ID Laboratory said, 'The effective implementation of this group diagnostic tool could diminish the spread of the pandemic if everyone uses it before going to a classroom, a factory, or a restaurant.' Jordi Laguarta and Ferran Hueto, of MIT’s Auto-ID Laboratory, are the other authors of the research. According to the authors of the study, the mobile app would allow a user to log in daily, record the sound of their cough on the phone and then get immediate results in whether the person is infected with the virus or not.

READ | COVID-19: 177 New Cases In Ahmedabad, 2 Deaths, 219 Recover

Researchers of the study said that the AI tool's model was first trained on a general machine-learning algorithm called ResNet50 so that it can easily distinguish and separate sounds associated with different degrees of vocal cord strength. This model then worked with over 10,000 of speech so that it could pick out specific words. The second level of training was done on a neural network that enabled the tool to detect different emotional states that often show up in speech. The third level training consisted of the neural network that has a huge database of coughing recordings to differentiate between the different lung and respiratory responses. Lastly, the final step is an algorithm that can distinguish between strong and weak coughing as the latter one is a sign of muscle weakness. 

READ | Delhi Doctor Shares Experience Of First Day Working In COVID-19 Ward: 'Duty Comes First'