Autoimmune rheumatic diseases (AIRD) are heterogeneous, relapsing–remitting disorders in which early diagnosis, flare prediction, and individualized treatment selection remain critical unmet needs. Recent advances in multimodal biomarkers—including serological and inflammatory markers, quantitative imaging (ultrasound/MRI), -omics signatures (e.g., interferon- and B-cell–related programs), and digital phenotypes from wearables and smartphones—can now be fused through AI pipelines to enhance phenotyping, risk stratification, and treatment-response modeling. This review synthesizes recent advances across three interconnected domains: (i) imaging artificial intelligence (AI), which standardizes the quantification of synovitis, erosions, and microvascular changes; (ii) omics-based stratification approaches in systemic lupus erythematosus (SLE) and related AIRD; and (iii) remote, patient-generated data streams that extend and complement traditional clinic-based assessments. We emphasize implementation science, highlighting registry-enabled infrastructures (e.g., ACR RISE), federated learning to preserve privacy across sites, and modern validation standards (TRIPOD+AI, PROBAST+AI, CONSORT-AI/SPIRIT-AI). Finally, we address equity and drift-monitoring, underscoring the need for continuous recalibration across ancestry, sex, age, and care settings. Collectively, these innovations are transitioning precision rheumatology from conceptual promise toward pragmatic, clinic-embedded deployment.
Al-Ewaidat et al. (Mon,) studied this question.