THE USE OF ARTIFICIAL INTELLIGENCE IN DIAGNOSING AND MONITORING RHEUMATIC DISEASES

Abstract

Background: Rheumatic diseases such as rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), and psoriatic arthritis present diagnostic and monitoring challenges due to their complex pathophysiology and heterogeneous clinical manifestations. The integration of artificial intelligence (AI) into rheumatology has emerged as a promising approach to improve early diagnosis, personalize treatment strategies, and enhance disease activity monitoring.

Objectives: This systematic review aims to synthesize current empirical evidence on the use of AI techniques in diagnosing and monitoring rheumatic diseases, highlighting their diagnostic accuracy, clinical utility, and future potential.

Methods: The review followed PRISMA 2020 guidelines and included peer-reviewed studies published between 2010 and 2025. Eligible studies were identified through comprehensive searches in PubMed, Scopus, Web of Science, Embase, and Google Scholar. Data were extracted on AI methods, rheumatic disease types, outcome measures, and model performance. Quality assessment was conducted using the Newcastle-Ottawa Scale and the Cochrane Risk of Bias Tool.

Results: Fifteen studies were included, covering AI applications across RA, SLE, and PsA. AI techniques such as machine learning (ML), deep learning (DL), and convolutional neural networks (CNN) achieved diagnostic accuracies ranging from 82% to 95%. Applications in disease monitoring showed utility in predicting flares, tracking treatment responses, and enabling remote monitoring through wearable technologies. Despite promising outcomes, limitations included data heterogeneity, lack of interpretability, and challenges in clinical integration.

Conclusion: AI technologies hold considerable promise in rheumatology, particularly in early diagnosis and continuous monitoring. However, addressing technical, ethical, and infrastructural barriers is essential for widespread clinical adoption and equitable healthcare delivery.