Researchers at the University of Cambridge have developed an AI-based application that aims to improve the early-diagnostic processes of COVID-19 by automatically detecting its possible presence based on the unique properties of an affected person’s respiratory tract. COVID-19 patients have been found to have a few key distinguishable traits – such as shortness of breath, a dry cough and intervals in breathing patterns. The app will take the sounds of the user’s breathing and coughing as input to run the analysis.
Given the lack of information about the exact nature, symptoms and effects of the virus, the more the data available, the better the global pandemic will be tackled. The simplicity of this app promises to make widespread testing viable, and will enable quick collection of a large crowd-sourced dataset from both healthy and infected individuals. A greater volume of input will help develop refined machine learning algorithms for more accurate diagnosis of the disease. Additionally, the app will also collect demographic and location information from the users – along with their medical histories – to try and better understand factors such as disease progression, risk of contraction and relationship of respiratory complications based on medical history, among others.
Note: The COVID-19 app is now available as a web version for Chrome and Firefox users here. The iOS and Android versions will be available soon.