Presymptomatic Early Warning System for COVID-19 Infection

A major challenge of the COVID-19 pandemic, which has tremendously impacted society, communities and the healthcare systems across the world, is the infectiousness of asymptomatic carriers of the virus. This challenge exposes major weaknesses in our US healthcare system’s approach to handling viruses in general, which currently depends on self-detection of symptoms by patients, and then subsequent reporting for care to the clinic, hospital, or emergency room as needed. For COVID-19, since a person can be infectious before any symptoms are perceived, the virus can be transmitted to any contacts directly or via surfaces for many days unknowingly. The ultimate goal of this research is to develop a multi-modal physiological sensing approach to providing presymptomatic early warning of COVID-19 infection. The central innovation lies in the combination of multiple modalities of cardio-mechanical, vascular, and electrophysiological signals with activity and environmental context. The following aims are proposed for the research: (1) using an existing database of intensive care unit patients wearing the existing sensing patch to calibrate the measured waveforms for accurate detection of pulse oximetry (SpO2) and temperature; and (2) conducting a prospective study in patients with COVID-19 and their family members to develop predictive analytics algorithms for presymptomatic early warning. Successful completion of this project would result in a set of algorithms and techniques that could be tested more broadly in a large population of patients and family members, providing feedback on standard vital signs such as SpO2 and temperature but also the derived presymptomatic warning, to assess whether spread of the virus could be curtailed in communities based on this knowledge. Such a study would likely be performed following commercialization of the technology, or licensing of the technology, via industry partners.