The Pentagon experimented with an algorithm that uses data collected by fitness trackers to predict whether a user has COVID-19 two days before they show symptoms. Politics reported Feb. 23.
The Pentagon’s Department of Innovation Unit tracked 11,500 users who wore the Garmin Phoenix 6 watch and an Oura ring from June 2020 to September 2021. The watch tracked the user’s respiratory rate, oxygen saturation, and heart rate while the ring collected heart rate and skin temperature. .
The researchers used the data from these devices to calculate a score for the user. The score ranged from 1 to 100 – the higher the number, the higher the probability of infection.
The results were used to warn users who were otherwise presymptomatic to be tested.
Out of the users who carried the trackers, 491 tested positive for COVID-19, with 73 percent of these cases detected by the algorithm. The algorithm also identified COVID-19-positive individuals 2.3 days before diagnostic testing.