Jason Mayberry is the stroke coordinator for Texas County Memorial Hospital in Houston, Missouri.
They began using RapidAI, an AI software, in October 2023 that helps their doctors more quickly assess, diagnose and begin treatment for people experiencing stokes.
For the month of October, we’re focusing on the intersection of healthcare and artificial intelligence.
He said the hospital also developed a relationship with CoxHealth, a Level I stroke center in Springfield that also uses the software.
Mayberry said the software can more quickly read scans and provide feedback – giving ER doctors in Texas County more confidence in their decisions and allows them to share scans, in real time, with Springfield neurologists to determine the best course of treatment, if needed.
“It definitely gives our doctors, once we've admitted somebody, it gives them a little sense of relief that within a couple minutes of the test being done, they can have an interventional neurologist looking at it and helping them,” Mayberry said. “So, they just have that backing, and I think that sense of or that awareness that they're not alone, they're not on an island, and they can get the information needed to them quickly.”
Mayberry added that within the first few months of using the software, they were able to keep six patients at Texas County that otherwise would have been transferred to a larger hospital.
"The AI will read it in two minutes, with, I think the last I heard was like a 99.97% accuracy or something like that."Jason Mayberry, Texas County Memorial Hospital
He said the FDA-approved software is HIPPA compliant, as patient records are largely de-identified – keeping patients anonymous, and are only available on the app for 72 hours.
Mayberry did stress that the AI software is not solely relied on. Instead, it's an addition to the process to improve confidence and streamline process.
If the program makes an error, Mayberry said, the hospital reaches out to the company so the system can be updated and, ideally, learn from its mistakes.
He spoke about how this software has impacted his staff – and his patients.
Jason Mayberry: Rural healthcare is – man, it's kind of the front lines of what's going on, you know, we have people that drive 45 minutes from their house just to get to our hospital here at Texas County. We're located about an hour and a half from any hospital in Springfield.
Our job is mainly to get those patients seen as quickly as possible, get them diagnosed, and then get them on to definitive care.
If we can treat them in the emergency department and be able to discharge them back home, that's great. If we're able to admit them, you know, that's phenomenal, too.
But if they need that specialized care, then that's where our hospital comes in to facilitate those transfers to where those people can get their total end care.
When I came over to Texas County, and I first started looking through their past stroke history and logs, they were transferring around 100-120 stroke patients a year to a Level I or Level II center. We're a Level III center here at Texas County.
But out of those people, only about 30 to 40 of them were getting treatment that we could not do here.
"Even though we are a smaller hospital, we can do a lot of great things, and we can – it's kind of cliche, but we can save your life."Jason Mayberry, Texas County Memorial Hospital
So, RapidAI is simply an app that has been developed, again, it's artificial intelligence that understands what like a hemorrhage would look like in the brain on a CT. On your CTAs, it understands what a large vessel occlusion looks like, or when there's decreased blood flow to certain areas of the brain.
What's phenomenal is it gives my ER doctors that backing. So, RapidAI – normally, when we do a head CT, we always have a radiologist read that.
Well, our radiologists aren't in-house, they're off-site, and it takes, on average, about 22 to 25 minutes for them to get all the images and look at them and then send a report to us. The AI will read it in two minutes, with, I think the last I heard was like a 99.97% accuracy or something like that.
So, that gives my ER doctors a little bit of, “Okay, I'm right. I can go ahead, and I can pull the trigger on the lytic, if needed, or I need to go ahead and contact neurology a little bit quicker, because, you know, it is a bleed somewhere, and I don't, we can't treat that here.”
So, within two minutes of those being done, we've got a reading back on our tablets. They kind of let us know what the AI is thinking.
Even though we are a smaller hospital, we can do a lot of great things, and we can – it's kind of cliche, but we can save your life.
You know, you don't have to go to the city, necessarily, to get medication that will prevent these things from happening, you know, we can offer that help to them and get them where they need to go and hopefully not have lasting effects from that injury.
