New AI Tool Can Quickly Spot Coronavirus Outbreaks
Post# of 281
Researchers from Georgia Tech, Harvard Medical School and the Boston Medical Center have developed a new online learning-based tool that allows the health sector to detect coronavirus outbreaks earlier in various counties in the U.S.
The Coronavirus outbreak detection tool is usually updated 2-3 times every week. It forecasts how fast a coronavirus outbreak will spread in a county by approximating the doubling rate of cases. The doubling time refers to the total number of days it takes for the cases in any given county to double. The tool takes into consideration the reported coronavirus deaths and cases, social distancing measures, face masks mandates, the rates of positive tests, changes in tests performed as well as the CDC’s Social Vulnerability Index.
This generates an interactive map that lets the users see a specific county’s population, the total number of new coronavirus cases in the county in the last week, the average number of cases recorded every day in the last week and also the coronavirus doubling rate.
Turgay Ayer, who is an associate professor in Georgia Tech H. Milton Stewart School of Industrial and Systems Engineering, says that it is important to identify an outbreak promptly to ensure that the area in question can be isolated. This will in turn help limit the spread to neighboring locations, and this will in turn assist in effectively containing the pandemic. The tool provides an upper hand in early detection of an outbreak as it is automated, thus speeding up the process.
An assistant professor at Harvard Medical School who is also the associate director at the Massachusetts General Hospital Institute for Technology Assessment, Jagpreet Chhatwal, states that while the earlier measures focused on reducing the spread of COVID-19 at the state level, detecting outbreaks at a local level allows the legislators to execute the necessary measures at the local level. This will help in effectively containing the pandemic. The presence of the automated system helps speed up the whole process as it can detect current outbreaks within a few days.
Last week, the research team used the coronavirus outbreak tool to confirm an outbreak in Johnson County, Iowa. The outbreak is thought to have originated from the University of Iowa. Furthermore, the tool also pointed out a few counties where outbreaks might occur. These include Wheeler County in Georgia, Whitman County in Washington, Harrisonburg County in Virginia and Monroe County in Indiana.
Analysts say the use of artificial intelligence (“AI”) to detect or deal with major disease conditions isn’t very new to many players in the biomedical field, such as Predictive Oncology (NASDAQ: POAI), so it isn’t hard to see how hugely useful the COVID-19 outbreak detection tool can be at this time.
Please see full terms of use and disclaimers on the BioMedWire website applicable to all content provided by BMW, wherever published or re-published: http://BMW.fm/Disclaimer