We used consumer smartwatches for the pre-symptomatic detection of coronavirus disease 2019 (COVID-19) in the work of Mishra, Wang, et.al, Nat. Biomed. Eng (2020). Both offline and online anomaly detection algorithms were developed for detecting physiological alterations due to the COVID infection. This work demonstrated the potential ability that activity tracking and health monitoring via consumer wearable devices can be used for the large-scale, real-time detection of respiratory infections, often pre-symptomatically. 

We developed a real-time smartwatch-based alerting system for detecting the COVID-19 and other stress events in the work of Alavi, Bogu, Wang, et al., Nat. Med (2022). In this paper, the alarming algorithms were developed to detect aberrant physiological and activity signals (heart rates and steps) associated with the onset of early infection. This paper showed that a real-time alerting system can be used for early detection of infection and other stressors and employed on an open-source platform that is scalable to millions of users.

In the work of Arias-Castro, Castro, Tánczos, Wang, JASA (2018), we investigated two distribution-free procedures – one based on the scan statistic calibrated by permutation and another one on a novel rank-based scan statistic – to detect structured anomalies such as abnormal intervals in one dimension. The paper showed that using one of these calibration procedures results in only a very small loss of power in the context of a natural exponential family. The procedure using the rank-based scan statistic has been applied to detect abnormal resting heart rates in the context of early detection of COVID from a smart watch.

Several signal detection methods were developed under the sparse mixture models in the works of Arias-Castro, Wang (2015, 2017).