[ad_1]
Researchers at MIT have used machine learning to create software capable of detecting whether a person has contracted COVID-19 by analyzing their cough, a development that could eventually result in an iPhone app for daily checks.
Until now, the iPhone has been provided to help users determine if they are at risk of contracting coronavirus by approaching someone who carries the virus. If a new discovery by MIT researchers is further developed, the iPhone may end up being able to do more to further slow the spread of the virus.
A team article published in the IEEE Journal of Engineering in Medicine and Biology states that an artificial intelligence model was created that could differentiate between asymptomatic people and those who are healthy, by analyzing the recordings of forced coughing. The model is claimed to be accurate 98.5% of the time when listening to recordings of people confirmed to have COVID-19, as well as 100% of recordings of asymptomatic cough.
The team collected more than 70,000 recordings through a website where the public could record a series of coughs via their smartphone or other devices, while also completing a survey on their symptoms, if they were confirmed to have the virus. and others. details. The recordings resulted in around 200,000 samples of a forced cough, including 2,500 of which were confirmed to have COVID-19 or were asymptomatic.
By combining the 2500 confirmed samples with another 2500 randomly selected from the dataset, the AI model was trained and then tested. The researchers say the results revealed “a striking similarity between Alzheimer’s discrimination and COVID.”
The AI framework built on one that existed for Alzheimer’s research and determined that it could detect four biomarkers related to vocal cord strength, feeling, lung and respiratory response, and COVID-19-specific muscle breakdown.
The team is now working to create a pre-screening application that it intends to distribute for free, based on the artificial intelligence model, as well as working with several hospitals to expand the pool of cough records, for further training.
The team suggests that such a cough analysis could be implemented in smart speakers and digital assistants for daily assessments. This, of course, would depend on the devices involved having good enough quality microphones and handling the necessary privacy issues, not to mention assistance from companies like Apple and Amazon to implement, which makes it unlikely if altruistic.
