If an outbreak of a disease like the flu is striking where you live, wouldn’t you want to know about it right away?
Scientists at Sandia Labs are working to get the word out even faster.
They say it’s especially important here in New Mexico where we still see uncommon diseases like Hantavirus and the plague.
“You can kind of think of it as a more effective way of finding the needle in the haystack,” explained computer scientists Patrick Finley.
He and fellow Sandia Labs scientist Drew Levin are finding success in their new research.
They’re working to improve how we detect infectious diseases around the country.
“To try to get an advance warning on when a particular disease may hit anywhere in the United States,” Finley said.
Right now, the Centers for Disease Control and Prevention uses traditional math to detect outbreaks.
They crunch the numbers on things like emergency room visits and prescription sales to determine where a disease like the flu, mumps or measles may be spiking.
Now, Finley and Levin are taking that approach one step further using artificial intelligence.
“Taking that idea we went ahead and constructed a computer system that behaves in many ways like your immune system,” Finley explained.
It is a concept straight from the human body.
Our immune system learns what is normal and can detect when something is not right.
“We’re taking a very similar approach with immune system on a national scale level we are trying to create a model of what is normal that we see across the country and then when something different happens we should be able to better detect it more rapidly,” Levin said.
Fast detection also means a quicker response, which is crucial.
The Sandia researchers’ approach can also better detect diseases in specific areas. Like Zika in Florida.
“Here in New Mexico, of course, we’re interested in rather obscure disease like the plague or hantavirus and I think that these methods will give us the capability to hopefully be able to find those more accurately and more quickly,” Finley said.
These researchers, who are also working with the University of New Mexico say being able to detect outbreaks faster and more accurately means hospitals can prepare a response and people can take preventative measures.
On the other hand, flagging non-outbreaks can waste resources and that’s why being able to spot diseases in specific areas is important.