Keeping Current: Virtual Virus Mapping

Engineering professor Hui Yang is using data analytics and Artificial Intelligence to better understand how COVID-19 spreads.

dot matrix

“It has been over a hundred years since the last big pandemic outbreak, the 1918 influenza pandemic, and we are still doing many of the same things to mitigate the spread of the coronavirus. We’re staying at home, we’re distancing, and wearing masks. We have quarantines. There have been significant improvements, of course, in medical and pharmaceutical interventions, in designing and developing vaccines. But another tool we have today is artificial intelligence—and the large amount of data that’s been collected from diagnostic testing and contact tracing as the epidemic has evolved that we can use to develop analytical methods and tools to better understand the virus and how it spreads.

“We leverage the data with statistics and artificial intelligence to provide simulation models through which we can test different scenarios for virus spread mitigation. Right now, individual states and countries are trying different approaches; we would augment these real-world approaches in simulated models, test these different policies, and compare them to try and understand how activities impact the spread, for example, and whether the stay-at-home policy is effective. A simulated model could also combine a certain number of symptomatic and asymptomatic people to calculate the probability of infection, determine the effectiveness of testing in that particular population based on their movements, and refine contact tracing. For different scenarios of initial infections, we can model the virus spread in that population and help figure out how health care professionals can respond to it.

“We work very closely with front-line doctors, who look to us to model and compare alternative strategies they have for disease mitigation but cannot extensively experiment on the population because of practical and ethical limitations. Our simulated models offer flexibility to help them with health care delivery and optimize testing, tracing, and resources. We are also hoping that simulated models can help in optimizing the design of policies for epidemic control by informing policymakers who are also looking for ways to improve the resilience of health care systems. It’s become clear that epidemics mandate the redesign of hospitals, medical clinics, and emergency rooms for better treatment and patient accommodations. Simulation modeling and system informatics can help in spurring design and innovation.”  —SI