SBIR Phase I: COVID-19 Cough Classifier Using Artificial Intelligence

The broader impact/commercial potential of this Small Business Innovation Research (SBIR) Phase I project is to develop a COVID-19 diagnostic tool using artificial intelligence. The proposed Cough Detector and Cough Classifier is able to "listen" to sounds in a given environment, then detects and classifies coughs. When a cough related to COVID-19 is identified, the individual and relevant personnel in a potential germ circle can be immediately notified. Functioning as an early warning system, the tool will work on a mobile device or laptop, and can be embedded in other technology, such as infrared cameras with microphones or other sound detection equipment. The tool will support ongoing outbreaks and mitigation of social distancing considerations.

This Small Business Innovation Research (SBIR) Phase I project will utilize deep learning and transfer learning to develop a COVID-19 cough classifier. The unique features of a COVID-19 cough require distinguishing between characteristics of widened airway, narrowed airway, fluid filled air sacs, airflow patterns of spirometry, stiff lungs, and others. The unique characteristics or features are learned while classify cough types on a training data set. A tuned deep learning model is able to distinguish COVID-19 cough from other types of cough in real-time.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.