New ACR Research Highlights How Artificial Intelligence Could Strengthen Rheumatology Care, Education and Clinical Decision Support

Studies explore AI-enabled ultrasound interpretation, clinician education and the reliability of generative-AI platforms in rheumatology.

ORLANDO – New research presented at ACR Convergence 2026 examines artificial intelligence’s growing role in rheumatology—from helping clinicians interpret musculoskeletal ultrasound images to preparing the workforce to use AI responsibly and assessing whether widely available AI platforms provide guideline-concordant clinical information.

Together, the three studies highlight both the opportunity and the responsibility facing the field: AI may help make assessment of inflammatory arthritis more efficient and consistent, but its safe adoption depends on rigorous validation, clinician education, and careful evaluation of accuracy and hallucination risk.

In one study, researchers developed a deep learning-based tool designed to automate assessment of synovitis, power-Doppler activity, erosions, and osteophytes on ultrasound imaging, while also classifying arthritis. The work addresses a central challenge in musculoskeletal ultrasound: image interpretation can be highly dependent on expertise and experience, and detailed scoring can take substantial time in routine practice.

“Artificial intelligence has the potential to make ultrasound assessment more reproducible and more accessible, particularly by supporting clinicians in identifying the imaging features that distinguish inflammatory from degenerative disease,” said Maja Schlereth, MS, principal investigator of the study, “Development of a Deep Learning-Based Tool for Automated Assessment of Synovitis, Power-Doppler Activity, Erosions, and Osteophytes and Classification of Arthritis Using Ultrasound Imaging.” “Our goal is not to replace clinical expertise, but to develop a tool that can augment image interpretation and help clinicians make timely, informed decisions for their patients.”

The investigators emphasize that AI-driven imaging tools should be evaluated across diverse patient populations, ultrasound equipment, and clinical settings before widespread clinical implementation. Such validation is critical to ensuring that performance is reliable and that new technologies advance—not widen—equity in rheumatology care.

A second multicenter pilot study focused on the education needed to support the responsible use of AI in rheumatology. The investigators developed and tested an AI curriculum for continuing and graduate medical education, recognizing that clinicians, trainees, and educators increasingly encounter AI-based tools in clinical practice, research, and patient communication.

“As AI becomes more visible in medicine, rheumatology professionals need more than a basic awareness of these technologies—they need practical fluency,” said Niharika Goparaju, MD, pediatric emergency medicine physician at Dell Children’s Medical Center and principal investigator of the study, “Artificial Intelligence Curriculum for Continuing and Graduate Medical Education: A Multi-Center Pilot.” “That includes understanding how these systems work, recognizing their limitations, identifying bias, and knowing when an AI-generated output requires independent verification.”

The curriculum pilot underscores the importance of teaching AI literacy alongside clinical reasoning, evidence appraisal, and patient-centered communication. Researchers say educational programs should help learners distinguish between AI tools that may assist with administrative, educational, or analytic tasks and systems whose outputs could directly influence patient care.

The third study compared AI platforms for guideline concordance and hallucination risk in rheumatology. As clinicians and patients increasingly turn to generative AI for answers to medical questions, the analysis examined whether platform responses align with established rheumatology guidance and whether they introduce unsupported, incorrect, or fabricated information.

“Generative AI can produce responses that sound confident and persuasive even when they are incomplete, outdated, or inaccurate,” said Aakanksha Pitliya, MD, principal investigator of the study, “Artificial Intelligence in Rheumatology: A Comparative Evaluation of Guideline Concordance and Hallucination Risk Across Platforms.” “Our findings reinforce that these tools should not be treated as independent clinical authorities. Their output must be evaluated against current evidence, professional guidelines, and the clinical context of the individual patient.”

Taken together, the studies suggest that AI’s most productive role in rheumatology may be as a carefully governed partner to clinicians: supporting image analysis, improving education, and helping organize information, while leaving diagnosis, treatment decisions, and patient communication grounded in human clinical judgment.

“AI is rapidly moving from a theoretical topic to a practical reality in rheumatology,” said Howard H. Yang, MD, RhMSUS, chair of the ACR’s Communications and Marketing Committee. “These studies show why the field must pursue innovation and oversight in parallel. We need tools that are clinically meaningful, transparent, validated, and implemented in ways that preserve the physician-patient relationship.”

The American College of Rheumatology will continue to support research, education, and dialogue on the responsible integration of emerging technologies into rheumatology practice.

These abstracts will be presented at ACR Convergence 2026. Presentation dates, times, and session details are available through the official meeting program.

Media Contact
Teri Arnold
Director, Public Relations & Communications
757-272-7002
tarnold@rheumatology.org

###

About the American College of Rheumatology
Founded in 1934, the American College of Rheumatology (ACR) is a not-for-profit, professional association committed to advancing the specialty of rheumatology that serves nearly 10,500 physicians, health professionals, researchers and scientists worldwide. In doing so, the ACR offers education, research, advocacy and practice management support to help its members continue their innovative work and provide quality patient care. Rheumatology professionals are experts in the diagnosis, management and treatment of more than 100 different types of arthritis and rheumatic diseases.

About ACR Convergence
ACR Convergence, the annual meeting of the American College of Rheumatology, is where rheumatology meets to collaborate, celebrate, congregate, and learn. With hundreds of sessions and thousands of abstracts, it offers a superior combination of basic science, clinical science, business education and interactive discussions to improve patient care and advance the specialty. For more information about the meeting, visit the ACR Convergence page, or join the conversation on X by following the official hashtag (#ACR26).