Machine learning algorithm accurately predicts osteoarthritis progression in nearly 80 percent of cases.
Using a machine learning algorithm with MRI scans can pinpoint signs of osteoarthritis years before the symptoms of the condition actually appear, a new study has revealed.
Published recently in the Proceedings of the National Academy of Sciences of the United States of America, an article from a team of researchers from the University of Pittsburgh School of Medicine and Carnegie Mellon University College of Medicine outline how they used a homegrown algorithm to capture the sensitive cartilage phenotypes that can help predict how osteoarthritis will progress in patients.
Their algorithm is important, they said, because nothing else like it exists.
“Currently no reliable method exists for [osteoarthritis] detection at a reversible stage,” the team wrote. “We present an approach that enables sensitive [osteoarthritis] detection in pre-symptomatic individuals. Our approach combines optimal mass transport theory with statistical pattern recognition.”
In particular, they said, they trained a classifier to be able to tell the difference between progressors and non-progressors on baseline cartilage texture maps.
Related Content: New MRI Approach Helps Assess Knee Osteoarthritis Progression
To determine whether the algorithm worked, the team examined knee MRI scans from patients who were part of the National Institutes of Health Osteoarthritis Initiative, a longitudinal, mutli-center, prospective, observational knee osteoarthritis study. Ultimately, they concentrated on 86 patients who, at the beginning of the study, had little-to-no signs of cartilage damage.
Based on their analysis, using MRI scans, the algorithm predicted the development of osteoarthritis in 78 percent of cases three years before patients began feeling any symptoms. That means the algorithm detected signs that were too nascent and too subtle for the interpreting radiologist to catch.
“When doctors look at these images of the cartilage, there isn’t a pattern that jumps out to the naked eye,” said lead study author Shinjini Kundu, M.D., Ph.D., “but, that doesn’t mean there’s not a pattern there. It just means you can’t see it using conventional tools.”
While there is no medication available to stop osteoporosis at the pre-symptomatic stage from progressing, there are drugs that can stymy the development of rheumatoid arthritis, a condition that is related to osteoarthritis, the team said. The team’s ultimate goal, they said, is to facilitate the creation of similar drugs for osteoarthritis in the hopes that such medication will alleviate the need for more invasive therapies.
“Instead of recruiting 10,000 people and following them for 10 years, we can just enroll 50 people whom we know are going to be getting osteoarthritis in two or five years,” said study co-author Kenneth Urish, M.D., Ph.D. associate medical director of the bone and joint center at the University of Pittsburgh Medical Center Magee-Women’s Hospital. “Then, we can give them the experimental drug and see whether it stops the disease from developing.”
GE HealthCare Launches AI Mammography Platform with Key Applications from iCAD
November 30th 2023Offering an all-in-one platform of artificial intelligence (AI) applications, MyBreastAI Suite reportedly facilitates early breast cancer detection and enhances efficiency with breast imaging workflows.
Study: Black Patients Less Likely Than Others to Receive MRI Assessment of Cognitive Impairment
November 27th 2023In a four-year study of over 1,600 patients who had outpatient head CTs, head CT angiography and/or brain MRI to assess cognitive impairment, researchers found that Black patients were over 9 percent less likely than White patients and over 16 percent less likely than Hispanic patients to receive brain MRI.
Could an Emerging AI System Lead to Earlier Autism Detection with DT-MRI?
November 21st 2023Through assessment of diffusion tensor MRI of the brain, a new AI system reportedly offers a 97 percent sensitivity rate in diagnosing autism in children between two to four years of age, according to research to be presented at the annual Radiological Society of North America (RSNA) conference next week.