Using Artificial Intelligence Technology to Explore the Effectiveness of Acupuncture in Improving Symptoms of Patients With Rheumatoid Arthritis

This study proposes an artificial intelligence-based multidimensional analysis framework to address the lack of quantitative evaluation and individualized prediction tools for evaluating the efficacy of acupuncture at improving symptoms of rheumatoid arthritis. By integrating multicenter clinical data, biomarkers, and acupuncture parameters, machine learning models were built to analyze the relationship between patient features, treatments, and outcomes. Key predictive factors were identified via feature importance and interaction analysis. A dynamic modeling approach predicted efficacy probabilities, and molecular pathway analysis revealed underlying mechanisms. The model achieved an area under the curve of 0.87 and 82.3% accuracy, and it identified 89.6% of tumor necrosis factor alpha dynamic patterns. This method supports precision acupuncture and advances data-driven traditional medicine.