Sotiris Tsiodras – Koronovios: The artificial intelligence system that detects asymptomatic travelers

Work concerning the analysis of a method of traveler control using techniques artificial intelligence and entitled “Improving COVID-19 border control with reinforcement learning” is published in the scientific journal Nature, in which the Sotiris Tsiodras.

The work is co-authored by, among others, the main creator of Eva Kimon Drakopoulos (University of Southern California-Los Angeles) and the professors Sotiris Tsiodras, Gikas Majorkinis, Dimitris Paraskevi, Pagona Lagiou and Christos Hadjichristodoulou (the first four from the Medical School of the University of Athens and the last from the Medical School of the University of Thessaly).

“During his pandemic coronavirus“Countries have imposed various travel restrictions to limit the spread of the virus, from quarantine to all travelers to a ‘blockade’ of tourists from specific countries based on the number of cases, deaths or positive tests,” the scientific journal said.

Eva: The “smart” artificial intelligence system that treats asymptomatic travelers

The enhanced learning system, called “EVA»Was applied in the summer of 2020 to travelers who visited Hellas to limit the influx of asymptomatic travelers infected with SARS-CoV-2 and to inform border authorities immediately.

The researchers say that Eva – which allowed more targeted controls based on specific criteria – identified almost twice as many asymptomatic travelers infected with coronavirus (1.85 times more) than if controls at the border had become completely in luck. If the tests had been done randomly, it is estimated that about 54% of the cases that the “smart” system had detected would have been detected.

In addition, Eva identified 1.25 to 1.45 times more asymptomatic infected travelers than they would have found if the test policy had been based on epidemiological criteria alone, as other countries have done. In fact, at the peak of the tourist season, he identified two to four times more asymptomatic than expected.

Also, Eva’s estimates of its spread coronavirus in other countries were used to issue early warnings about high-risk areas, which the Greek government used to adjust its travel protocols accordingly, requiring travelers from certain countries to have a negative molecular test (PCR) upon entering Hellas. In this way, the study estimates that Eva prevented an additional 6.7% of coronavirus-infected travelers from entering our country during last year’s tourist season.

In an analysis by Eva in ‘Nature’, Distinguished Professor of Health Policy Ziad Obermayer of the School of Public Health at the University of California, Berkeley points out that the system has faced political and legal challenges, as well as – in order to comply with the Data Center (GDPR) of the EU – the creators of the algorithm deliberately limited the data with which they fed it, thus inevitably limiting its accuracy somewhat. This shows, as he points out, that well-meaning legislation such as the GDPR can have both negative and positive consequences, as it limits a government’s ability to protect the health of its citizens.

The results of Eva are impressive

However, as he emphasizes, Eva’s results have been “impressive”, doubling the number of cases detected per test performed. At the same time, as he states, “the success of this algorithm highlights the inadequacy of the border policies of almost all other countries”, which often resorted to more “coarse” policies, for example by imposing the complete closure of borders to travelers or testing in all travelers from a particular country, “ignoring the huge differences between people within countries”.

As he underlines, “if the (Greek) border officials had denied entry to all passengers from countries with alarming epidemiological data, then they would have prevented people with Covid-19 from entering Greece, but at the cost of destroying a pillar of the its economy (of tourism) ”. That’s why, he concludes, the Eva algorithm “will be remembered as one of the best examples of data usage in the fight against Covid-19. “It’s a success story of how a team of researchers collaborated with enlightened policymakers to produce a tool that has enormous social value.” It also highlights, he says, “the great promise of artificial intelligence to make good decisions, which in many cases can make the difference between life and death.”