UF researchers using artificial intelligence to detect Parkinson's disease more accurately
Using AI to detect Parkinson's sooner
FOX 13's Kylie Jones reports on researchers at UF using AI to detect Parkinson's disease sooner and more accurately, which they say can be a game changer.
GAINESVILLE, Fla. - Researchers at the University of Florida are using artificial intelligence to detect Parkinson's disease more accurately.
Researchers at UF have developed Automated Imaging Differentiation for Parkinson’s, an AI software designed to detect the disease sooner and more precisely.
Researchers at UF using artificial intelligence to detect Parkinson's disease more accurately.
"If you speak to patients, many of them have told me that their journey to get a diagnosis takes a long time," said Dr. David Vaillancourt, a professor and the Chair of the UF Department of Applied Physiology and Kinesiology.
What they're saying:
Vaillancourt says it can take patients multiple years, and to see multiple providers to simply get diagnosed with Parkinson's disease.
A brain scan showing Parkinson's disease.
"They're basically kind of spinning their wheels sometimes and looking for that outcome and looking for that decision," Vaillancourt said.
Vaillancourt says the statistics surrounding Parkinson's diagnoses are striking, and that this new AI technology could drastically change that.
"In the first five years of a patient's journey, the diagnostic accuracy is between 55-78% accurate," Vaillancourt said. "Which means, if you had a room of 100 people who have been diagnosed with Parkinson's within the last five years, you would say 55 to 75, or so, would have the right diagnosis and the rest wouldn't."
Big picture view:
He says this software could detect the disease sooner and be over 96% more precise.
Vaillancourt says the technology is unique and can work across all types of MRI software.
Brain scans from JAMA Neurology.
"A typical patient who has Parkinson's would see a neurologist," Vaillancourt said. "A neurologist would order an MRI."
Vaillancourt says a neurologist would use the AI software in conjunction with existing technology to help study those scans and look for changes in degeneration in the brain.
"Looking, almost like, for a fingerprint in the brain," Vaillancourt said. "So, you know, when you open up a phone, and it might read your thumbprint for opening a phone to say, 'Yes, open it', or 'No, keep it locked', that's basically what we're doing."
What's next:
He says this doesn't tell a physician what to do, but it could help them see things more accurately and make decisions about a patient's treatment or needs.
Vaillancourt says it could help patients get the proper treatment for their diagnosis.
CLICK HERE:>>> Follow FOX 13 on YouTube
He says the technology could also be a big step forward in advancing drug development against the disease in the long term.
"There's no disease-modifying medications for Parkinson's; none have ever been approved. Part of the reason could be the medications, but also part of the reason could be that the patients that are in the trials might not have the right diagnosis."
"If this tool could be used to say, well, maybe 10% of the patients in that trial didn't actually have the disease that you thought they had, then maybe the medication could actually become seen as effective," Vaillancourt said. "That might actually help more medications get approved."
Researchers say the software is waiting for approval from the U.S. Food and Drug Administration, and they hope doctors can start using it in conjunction with existing technology.
The Source: FOX 13's Kylie Jones collected the information in this story.
WATCH FOX 13 NEWS:
STAY CONNECTED WITH FOX 13 TAMPA:
- Download the FOX Local app for your smart TV
- Download FOX Local mobile app: Apple | Android
- Download the FOX 13 News app for breaking news alerts, latest headlines
- Download the SkyTower Radar app
- Sign up for FOX 13’s daily newsletter