A group of engineers has released a study detailing how international air travel has affected the spread of COVID-19, pinpointing Western Europe, the Middle East, and North America as major contributors to the pandemic.
At the University of Houston, a team of engineers has published a study in the journal Nature exploring the impact of international air travel on the global spread of COVID-19. Utilizing a newly created AI tool, the researchers were able to identify infection hotspots associated with air travel, highlighting significant areas that contribute to disease transmission.
The study found that Western Europe, the Middle East, and North America are the primary regions driving the pandemic, largely due to the high number of international flights departing from or passing through these areas.
Hien Van Nguyen, the lead researcher and an associate professor of electrical and computer engineering at UH, stated, “Our work provides a powerful deep learning-based tool for examining global pandemics and is essential for policymakers to make informed choices regarding air traffic limitations during future outbreaks.”
The tools
Nguyen and the team created a computer program named Dynamic Weighted GraphSAGE, designed to analyze expansive networks of frequently changing data, such as flight schedules, to uncover patterns and trends.
“The program examines spatiotemporal graphs, which show how various elements are connected through both space (different locations) and time, giving a clearer understanding of how this influences factors like disease spread or travel trends,” Nguyen explained.
To assess the influence of air travel on infection rates, Nguyen along with graduate students Akash Awasthi and Syed Rizvi conducted perturbation analysis on their model to determine its sensitivity to various factors and investigated flight connections among different regions and countries.
This analysis enabled them to identify which segments of air traffic most significantly affect the transmission of the virus, as well as which flight reductions in sensitive areas could effectively lower projected global case numbers.
The strategies
Nguyen suggested air traffic reduction strategies that could greatly aid in controlling the pandemic while minimizing disruptions to public mobility. “Policies that involve significant cuts to flights originating from Western Europe are anticipated to yield larger decreases in global COVID-19 cases,” he added.
This research showcases a novel approach using perturbation analysis within spatiotemporal graph neural networks to provide insights into pandemic prediction,” he mentioned.
Though the conclusions are based on the context of COVID-19, Nguyen emphasized that the findings can be applied to any pandemic scenario.
Additional contributors to this project are affiliated with the Houston Methodist Research Institute.
