By Emily Gersema, University of Southern California — MedicalXPress — 22 September 2021
A prescriptive computer programme developed by USC Marshall School of Business and the Wharton School of Business at the University of Pennsylvania, and used by Greece to identify asymptomatic COVID-19 infections in travellers, may have slowed the virus’s spread through its borders, according to a study published in the journal Nature. “It was a very high-impact artificial intelligence project, and I believe we saved lives by developing a cutting-edge, novel system for targeted testing during the pandemic,” said USC Marshall assistant professor Kimon Drakopoulos.
In July 2020 Greece reopened its borders to spare its tourism-dependent economy. Working with USC Marshall and Wharton researchers, Greece developed “Eva” — an AI algorithm that uses real-time data to identify high-risk visitors for testing. The algorithm analyses traveller-submitted data including prior travel locations, demographics and travel itinerary to identify those most likely to be asymptomatically infected. To prevent blind spots, the system also identified travellers for whom limited data was available, reinforcing the algorithm’s accuracy over time. With Eva, Greece tested about 17% of the estimated 41,830 households arriving daily and nearly doubled the number of infections that a typical randomised testing approach would have captured. “Given that randomised testing requires a large testing supply, Eva offers an impressive alternative,” said Drakopoulos. The joint study was published in Nature.