A team of engineers at the University of Houston has published a study in the journal Scientific Reports on how international air travel has influenced the spread of COVID-19 around the world. By using a newly developed AI tool, the team identified hotspots of infection linked to air traffic, pinpointing key areas that significantly contribute to disease transmission.

The analyses identified Western Europe, the Middle East and North America as leading regions in fueling the pandemic due to the high volume of outgoing international flights either originating or transiting through these areas.

"Our work provides a robust deep learning-based tool to study global pandemics and is of key relevance to policymakers for making informed decisions regarding air traffic restrictions during future outbreaks," said Hien Van Nguyen, lead researcher and associate professor of electrical and computer engineering at UH.

The tools

Nguyen and team developed a computer program, called Dynamic Weighted GraphSAGE, that helps analyze big networks of constantly changing data, like flight schedules, to see patterns and trends.

"It looks at spatiotemporal graphs, or how things are linked across both space (different locations) and time to better understand how this affects things like the spread of diseases or transportation patterns," said Nguyen.

To understand how air travel affects the spread of infections, Van Nguyen and graduate students Akash Awasthi and Syed Rizvi tested small changes in their model (perturbation analysis) to see how sensitive it is to different factors and examined flight connections between different regions and countries.

This helped them analyze which parts of air traffic have the biggest impact on the spread of the virus and which flight reductions in highly sensitive areas would efficiently reduce predicted global cases.

The strategies

"We propose air traffic reduction strategies that can significantly impact controlling the pandemic with minimal disruption to human mobility," said Nguyen. "Policies including stringent reduction in the number of Western European flights are predicted to cause larger reductions in global COVID-19 cases.

This work represents a novel usage of perturbation analysis on spatiotemporal graph neural networks to gain insight on pandemic forecasting," he said.

Although the findings stem from the COVID-19 context, the insights gained are generalizable to any pandemic, said Nguyen.