As the COVID-19 pandemic rages on unabated in countries around the world, there’s a shared desire among those forced to shelter in place to see the extent to which social distancing is slowing the disease’s spread. It’s understandable — collateral damage from government-imposed business closures threatens to devastate entire industries. As of this week, 26 million Americans have filed for unemployment claims, according to the U.S. Bureau of Labor Statistics, and the International Monetary Fund predicts a global financial crisis rivaling the Great Depression.
Fortunately, a preprint study published by researchers at the University of Texas, the Southwest Research Institute, and the University of Texas Health Science Center in San Antonio strongly implies that quarantining and physical distancing are having the intended effects. Using a hybrid AI system dubbed SIRNet and several epidemiological models, which were trained on smartphone location data along with population-weighted density and other data points from the startup Safe Graph, World Health Organization, the U.S. Centers for Disease Control and Prevention, and elsewhere, the coauthors claim they managed to accurately predict the outcomes of various social distancing policies.
People can check their state’s projections on a website published by the University of Texas’ COVID-19 Modeling Consortium.
The country-, state-, and country-level location data from tens of millions of smartphones ingested by the researchers’ system was used to predict contact rate, a function of population density as well as movement and interactions among people in a region. This was plotted against COVID-19 case count data, specifically a time series set that captured active, recovered, and fatal cases of COVID-19 at varying levels of geographical granularity, to which the researchers applied a 10-day lag time to account for the delay between infectiousness and receiving a positive test confirmation.
The researchers report that, based on projected forecasts three weeks into the future (the system’s maximum), only a continuation of “quarantine-level mobility” will result in low COVID-19 case counts. If restrictions were to be reduced by around 50%, the system projects that some communities would reach the edge of stable peak cases, where the death curve would either stay at a low peak or quickly, sharply increase. And if 75% of a population was able to move about as freely as they normally would, the system predicts the result would be a slightly delayed peak approximately 2/3 of the maximum peak during 100% mobility (except in South Korea).
In Bexar County, Texas, where mobility as of April 11 was approximately 50% of normal, relaxing social distancing measures could result in a “runaway growth” in deaths and hospitalizations. the system shows. By contrast, in King County, Washington, where heavy mobility restrictions remain in place, the system predicts that continuing those measures would depress the number of new deaths to close to zero by June.
The system agrees with an MIT model detailed in an early April preprint paper, which found that in places like South Korea, where there was immediate government intervention, the virus spread plateaued more quickly. Trained on data collected from Wuhan (China), Italy, South Korea, and the U.S. after the 500th case was recorded in each region, it learned to predict patterns in the infection spread, drawing a correlation between quarantine measures and a reduction in the virus’ effective reproduction number.
A separate model — one published earlier this year by researchers at Microsoft, the Indian Institute of Technology, and TCS Research (the R&D division of Tata Consultancy Services) — learned policies automatically as a function of disease parameters like infectiousness, gestation period, duration of symptoms, probability of death, population density, and movement propensity. Over the course of 75 simulations with simulations lasting 52 weeks (364 days), it showed that governments that locked down 5% to 10% of communities experienced a lower peak of COVID-19 infections.
Elsewhere, an international team of researchers used human mobility data supplied by Baidu to elucidate the role of COVID-19 transmission in Chinese cities. They found that, following the implementation of control and containment measures, the correlation between the geographic distribution of COVID-19 cases and mobility dropped and growth rates became negative in most locations, indicating that the measures mitigated the spread of COVID-19.
As encouraging as the predictions might be, it’s important to keep in mind that even the best algorithms — like those developed by HealthMap, Metabiota, and BlueDot, which were among the first to accurately identify the spread of COVID-19 — can only learn patterns from historical data. As the Brookings Institution noted in a recent report, while some epidemiological models employ AI, epidemiologists largely work with statistical models that incorporate subject-matter expertise.
“[A]ccuracy alone does not indicate enough to evaluate the quality of predictions,” wrote the Brookings report’s author. “If not carefully managed, an AI algorithm will go to extraordinary lengths to find patterns in data that are associated with the outcome it is trying to predict. However, these patterns may be totally nonsensical and only appear to work during development.”
Nevertheless, the models provide a preponderance of evidence in support of quarantining and distancing policies — even as those policies come under fire from protesters.
For AI coverage, send news tips to Khari Johnson and Kyle Wiggers and AI editor Seth Colaner — and be sure to subscribe to the AI Weekly newsletter and bookmark our AI Channel.
Thanks for reading,
Kyle Wiggers
AI Staff Writer
Author: Kyle Wiggers.
Source: Venturebeat
