You are viewing a preview of...

Digital COVID-19 Testing with Machine Learning Analysis of Cough Sounds

A low cost, non-invasive digital method of COVID-19 testing with high specificity and sensitivity for better management of COVID-19

Background

More than 40% of COVID-19 infected people show no to very moderate symptoms, which has significantly contributed to the non-intentional spread of the disease. This situation mandates prompt and precise identification of COVID-19 through frequent and widespread testing to prevent community outbreaks. The world health organization (WHO) has identified and updated several symptoms of COVID-19, such as high temperature, coughing, and breathing difficulties. However, these symptoms are common for several respiratory diseases and not necessarily unique to COVID-19, rendering it difficult for patients to self-assess. The gold-standard method for diagnosing COVID-19 uses reverse transcription-polymerase chain reaction (rRT-PCR) in nasopharyngeal (NP) swabs. However, sample collection with the NP swab is an invasive method and is not ideal for screening, prognostics,

Log in or create a free account to continue reading