Table 3.

Summary of the studies of artificial intelligence (AI) in disease management of rheumatoid arthritis (RA), including authors, data types, methods, and main findings

Main findings	Data types	Methods	Author (Month/Year)
ML is being used in the field of RA to support dis-	Observational cohort	Bayes, random forests	Gossec et al.[60] (Oct. 2019)
ease management. Applications include detecting flares based on physical activity data, predicting flares using ultrasound and blood test data,	Ultrasound images, blood test	Logistic regression, random forest, and XGBoost	Matsuo et al.[63] (May. 2022)
extracting results from clinical notes using natural	EHRs	NLP	Humbert-Droz et al.[64] (Mar. 2023)
language processing, and developing AI-based
flare prediction systems. These approaches have the potential to improve disease monitoring in RA.	Clinical cohort	AI-powered RA clinical decision support tool	Labinsky et al.[65] (Jan. 2023)

NLP, natural language processing; EHRs, electronic health records; XGBoost, eXtreme gradient boosting; ML, machine learning.