© 2022 Innovation Trail

New system predicts likelihood of food poisoning outbreaks

Adam Sadilek

Upstate researchers have found a way to predict the likelihood of getting sick after visiting a particular restaurant. The system is called nEmesis and it monitors the tweets from restaurant patrons and detects likely cases of foodborne illness in close to real-time.

Many people tweet on devices that are GPS enabled, and nEmesis uses this to figure out which, of thousands of known restaurant locations, they ate at.

It then continues to track a person’s tweets for 72 hours after a restaurant visit, to detect whether or not they’ve become ill.

Researcher Vincent Silenzio, a member of the University of Rochester team to develop the nEmisis system, says it could help to address a major public health issue affecting millions of people every year.

““Food poisoning and food borne illness in and of itself is a huge, huge public health conundrum and in the US alone it’s responsible for almost $80 billion worth of economic costs per year, so it’s not a small problem by any stretch of the imagination,” Silenzio says.

“Literally millions of people experience food borne illness every year in the US. More than 100,000, well more than 100,000 are hospitalized and 3,000 die. So this is not a small issue by any stretch.”

Silenzio acknowledges that the system is not perfect - not everyone tweets and when they do it’s not always about their health.

But, he says the strength of the system lies in the large amount of data it can access.

“What’s nice about large amounts of data is that you can point back towards a potential problem spot and then correlate with other people who have passed through that problem spot.”

An individual case could be unrelated to a restaurant visit, but when multiple people have a similar experience, the numbers are revealing according to Silenzio.

The app is not currently available to the public, but Silenzio says they’re working on refining it for general release in the not-too-distant future.

He says the system has the potential to help the public make more informed choices. In addition Silenzio says it could be used by restaurant owners for self-monitored quality control, as well as to aid in traditional public health inspections.

“The biggest value I think is in helping to direct resources in a smart way. If you happen to be in the public health sector and you have the responsibility for doing this kind of monitoring, being able to look at the places that are more problematic today rather than using those resources to look at places that are less of a concern.”