University of Akron prof develops model to predict COVID-19 spread

AKRON On the commute from his home in Cuyahoga County to his job at the University of Akron earlier this year, assistant math professor Alex Hoover found himself pondering the novel coronavirus.

COVID-19 had come across his radar in January, but it hadn’t yet become the pandemic that knee-capped most of the world.

As the pandemic progressed, Hoover developed a idea. The roads he traveled and the roads connected to them acted as a network for the coronavirus, helping it to circulate in the region.

“[I] started thinking about all the commuter traffic that was being affected by lockdowns,” Hoover said. “[And] how that affected COVID.”

For an applied math expert, the idea was an exciting challenge. Hoover, who earned his doctorate in mathematics from the University of North Carolina in 2015, has applied mathematics to jellyfish — “Quantifying performance in the medusan mechanospace with an actively swimming three-dimensional jellyfish model” – crabs and larvaceans.

Using the idea he developed on his back-and-forth trips between Cuyahoga County and Akron, Hoover applied his math expertise to the problem.

If the region’s road network was functioning as a COVID superhighway, he reasoned, understanding the process could be quite helpful to health experts.

“Why not try to model how COVID is spreading?” he asked himself.

So Hoover, who grew up in Northeast Ohio and earned his bachelor’s degree at Case Western Reserve University, teamed up in early April with colleagues from Case to develop a method that would predict the spread of the disease. As the idea progressed, Hoover researched traffic data and census information on the movement of people from one community to another.

The team developed a COVID-19 forecasting tool that’s used by the Cuyahoga County Health Department and other health-related organizations to predict the course of the coronavirus. The forecasts help the department formulate a response to the expansion or contraction of new cases before they occur.

“Together, we created this COVID model [on] how it spreads from one county to another,” he said.

The forecasting tool incorporates current health data to predict COVID’s course in coming weeks and refines the data as it goes along. According to Case Western Mathematics Professor Daniela Calvetti, the predictions have proven “quite accurate” for two to three weeks ahead.

The question Hoover and his colleagues wanted to answer as they approached the problem was simple — but it was important in the battle against COVID-19.

“How is COVID going to affect our counties, our regions, in a week?” Hoover said.

The model looks at the counties that form the core of Northeast Ohio, and has proven reliable, he said. He compared its performance to a weather forecast, which is more dependable the closer it is to the forecast date.

In a paper authored by Hoover and his colleagues, data from Summit County is used in a discussion of results from their model. The article was published June 19 in Frontiers in Physics.

Hoover said the model is constantly updated to improve its accuracy.

“With every new data point, we can make better predictions,” he said.

Alan Ashworth can be reached at aashworth@thebeaconjournal.com. Follow him on Twitter at @newsalanbeaconj.

Learn more

To read about the COVID-19 forecast model developed by University of Akron assistant professor Alex Hoover and colleagues from Case Western Reserve University, go to https://bit.ly/3bvh9Mf.

To read the Frontiers in Physics article, go to https://bit.ly/3gZ5gzz.