MIT has developed software that detects COVID-19 through sound

Researchers at the Massachusetts Institute of Technology (MIT) have developed software that can analyze a person’s cough to see if he is infected with COVID-19, a development that could eventually lead to the development of an iPhone app for routine check-ups,media reported. Until now, the iPhone has helped users determine if they are at risk of contracting a new coronavirus by approaching someone who carries the virus.

MIT has developed software that detects COVID-19 through sound

If a new discovery by MIT researchers is further developed, the iPhone could eventually play a bigger role in further curbing the spread of the virus.

In a recent research paper, the team said they created an AI model that could distinguish asymptomatic and healthy groups by analyzing records of forced coughs. The model is said to be 98.5 per cent accurate when listening to recordings of patients diagnosed with COVID-19, and more accurate in recording asymptomatic coughs, at 100 per cent.

The team collected more than 70,000 recordings through a website where members of the public could record a series of coughs on their smartphones or other devices while filling out a survey of their symptoms. These records produced about 200,000 forced cough samples, including 2,500 people diagnosed with COVID-19 or asymptomatic infections with COVID-19.

The researchers say the results of the study reveal striking similarities between Alzheimer’s disease and COVID identification.

The AI framework is based on the existing Alzheimer’s research framework and determines that it can select four biomarkers associated with vocal band strength, mood, lung and respiratory responses, and muscle degeneration for COVID-19.

The team is currently working on creating a pre-trial application based on the AI model and intends to roll it out to the public free of charge, and will also work with hospitals to expand the cough recording pool so that further training can be carried out.

The team suggests that this cough analysis can be applied to smart speakers and digital assistants so that routine evaluations can be performed. Of course, it will depend on whether the device has a microphone of sufficient quality and handles the necessary privacy issues, not to mention the need for assistance from companies like Apple and Amazon, which, while although al-Alism, is unlikely to happen.