AI in rheumatology: Diagnostic gains, clinical gray zones in OA care
Industry Buzz
AI is a powerful adjunct to an orthopedist's practice, allowing the focus to be on taking care of the patient.
—Stephen Stache, Jr., MD, FAMSSM
Al allows us to move from ‘standard’ positioning to optimized alignment tailored to that individual's unique anatomy and activity goals. We’re restoring specific lifestyles with a level of accuracy that was previously impossible.
—Harpreet Bawa, MD
Over the last few years, artificial intelligence (AI) has undoubtedly found its place in medicine—from supporting clinical documentation to imaging interpretation.
With nearly 500 million people affected by osteoarthritis (OA) globally, the time is now to advance OA care. [] So, how is AI showing up in the OA space?
Cory Calendine, MD, an orthopedic surgeon, put it succinctly: “The honest answer is we are further along than most clinicians realize and nowhere near as far along as the headlines suggest.”
Related: 2026's most anticipated AI advances—and how docs are navigating the promise and pitfallsAI in OA diagnosis, information extraction, and treatment
First, AI may aid in the diagnostic process. OA is generally diagnosed using symptomology, physical exams, and imaging, but these methods have limitations. In fact, early disease may be missed entirely.
Now, AI may be able to identify subtle joint changes and improve diagnostic accuracy, according to Bone Research. []
AI can also help with information extraction. For example, a study abstract presented at the most recent Osteoarthritis Research Society International (OARSI) World Congress examined how AI-prompt engineering can help sift through patient notes to identify key information about pain. []
The study found that out of 159 OA patients seeking a total knee arthroplasty, AI extracted 133 internal medicine notes and 66 orthopedic notes centered on pain. []
Although there were limitations, the researchers found that when patient pain was explicitly documented in the chart, AI was able to accurately extract this information. This can act as a time-saving tool for abstracting data from clinical notes. []
And then there’s PROBE—a large research project using AI and machine learning to improve how knee osteoarthritis is understood and treated. []
PROBE analyzes data from millions of patients to identify disease subtypes, while also working on clinical trial design and customized treatment plans. []
What do the experts have to say about AI in OA care?
Stephen Stache, Jr., MD, FAMSSM, a non-operative sports medicine physician, says he uses AI in all the aforementioned ways—to analyze imaging, assist in decision-making, and help with billing and other non-clinical tasks. “AI is a powerful adjunct to an orthopedist's practice, allowing the focus to be on taking care of the patient,” he says.
“While AI is still finding its footing in non-surgical treatments…it is already revolutionizing the surgical management of osteoarthritis,” Harpreet Bawa, MD, a fellowship-trained orthopedic surgeon and joint replacement specialist, tells MDLinx. “AI integrated into robotic-assisted joint replacement surgery has allowed us to go beyond simple 2D templates to dynamic, patient-specific modeling.”
He says AI helps him understand how an implant will behave when in specific positions, like a yoga pose or athletic movement.
“Al allows us to move from ‘standard’ positioning to optimized alignment tailored to that individual's unique anatomy and activity goals. We’re restoring specific lifestyles with a level of accuracy that was previously impossible,” he says.
Dr. Celandine says he’s eager to learn more about AI in shared decision-making. “Patients now use LLMs routinely to get generic information, but how can AI be used to enhance personal decision-making and improve the physician-patient collaboration instead of damaging it?” he asks.
It’s a good question. Patients and clinicians alike will still need to use the technology judiciously, according to a letter published in JMIR Biomedical Engineering. []
The authors suggested clinicians use ChatGPT “cautiously and as a screening tool prior to their own validation.” They also recommended that clinicians educate patients on the risks of using AI for self-diagnosis. []
Jonathan Bruner, DO, an osteopathic physician board-certified in neuromusculoskeletal medicine and osteopathic manipulative medicine, also remains wary.
“Open AI engines pull information from unknown sources and can provide misinformation,” he says. “It’s more important than ever for patients to have a strong support group of healthcare providers with whom they can have these discussions.”