PITTSBURGH (KDKA) — More than 3 million Americans and almost 25 percent of Pennsylvanians have osteoarthritis, a debilitating joint disease.
Currently, there is no cure. Osteoarthritis contributes to the hundreds of thousands of joint replacement surgeries performed in the U.S. each year.
Local researchers are now using artificial intelligence to identify the disease much sooner, along with a possible cure.
“This is allowing us to follow along with what the computer is seeing,” said Dr. Ken Urish, a researcher at UPMC.
Researchers from the University of Pittsburgh and Carnegie Mellon University have designed computer software that detects osteoarthritis years before symptoms develop.
“The biggest lesson I have learned from my job is that people completely take for granted the ability to walk a block and not have it hurt,” said Dr. Urish.
The disease is typically diagnosed using a standard grayscale MRI. It’s only visible once the cartilage has already deteriorated.
The new technique uses color with an MRI that visualizes good and bad cartilage.
Using this, doctors can determine pre-symptomatic joint disease in a matter of seconds and with 78 percent accuracy, according to studies.
“You’re getting these very dark areas where the arthritis is starting to develop,” said Dr. Urish as he explained how the visualizations work.
The computer software generates the images using thousands of lines of code observing each individual pixel. It’s like analyzing a single tree from an entire forest.
“Generating thousands of features that we’d never been able to understand by looking at the lines of code,” said Dr. Urish.
The doctor said knowing who is at risk may uncover an even bigger scientific breakthrough.
“If we can pick out the patients who are going to get this, the next step becomes how can we develop different treatments to actually keep them from developing the arthritis?” said Dr. Urish.
Though the system is years away from being used in the doctor’s office, researchers hope this technique will allow patients faster access to clinical trials for a preventative drug.