Combined ultrasound and AI model best predicts rheumatoid arthritis activity

Living with rheumatoid arthritis means watching for signs that your disease is flaring up. Doctors often rely on blood tests like ESR and CRP, but these can be slow to change. A new study offers a sharper way to see what is happening inside the joints right now. Researchers tested a combined model that uses musculoskeletal ultrasound and a machine learning algorithm to assess disease activity. They compared this to ultrasound alone and a different imaging method called radiomics. The results showed the combined model was the most accurate at predicting when the disease was active. It found specific signs like high blood flow in the joint lining that signal a flare. This approach helps clinicians monitor patients more dynamically and choose the right medical strategies for their specific needs. While the study involved patients from two hospitals in China, the potential to improve routine care is clear. By providing a reliable imaging basis, this tool supports better decisions without adding unnecessary risk or cost.