AI is making significant strides in the diagnosis and outcome prediction of vasculitides, rare inflammatory disorders that can be challenging to identify due to their diverse clinical presentations. A new systematic review analyzed AI applications in vasculitis care, revealing impressive accuracy in diagnosing Kawasaki disease and highlighting the need for broader research across other vasculitis types.
The study reviewed 46 research papers published between 2000 and 2024, assessing AI models across different data types, including clinical records, imaging, and textual data. The findings indicate that AI models excel in diagnosing Kawasaki disease, achieving sensitivities up to 92.5% and specificities reaching 97.3%. Additionally, predictive AI models demonstrated strong performance in identifying complications such as intravenous immunoglobulin (Ig) resistance, with areas under the curve (AUC) ranging from 0.716 to 0.834.
Despite these advances, AI applications in other vasculitis forms remain limited, often due to small datasets and a lack of external validation. The study underscores the potential of deep-learning and machine-learning models to improve diagnostic accuracy and outcome forecasting but stresses the importance of expanding research to include larger datasets and integrating newer AI models, such as large language models, to enhance clinical utility.
As AI continues to evolve, its integration into clinical practice could revolutionize how vasculitis is diagnosed and managed. For healthcare professionals, staying informed about these technological advancements may be key to improving patient outcomes in this complex field.
Reference: Omar M et al. Applications of Artificial Intelligence in Vasculitides: A Systematic Review. ACR Open Rheumatol. 2025;7(3):e70016.
Anaya Malik | AMJ