Abstract :
Ankylosing spondylitis (AS), a chronic inflammatory disease affecting the spine, is marked by gradual structural damage that often remains undetected in early stages. Predicting radiographic progression is crucial for timely intervention and personalized treatment. Traditional scoring systems, though validated, are limited by subjectivity, delayed detection, and poor sensitivity to change. Recent advances in artificial intelligence (AI), particularly machine learning and deep learning, offer novel tools for early and accurate prediction of progression. These AI models leverage imaging, clinical, laboratory, and genetic data to identify high-risk patients and stratify disease phenotypes. Studies have shown that AI-based systems can outperform traditional approaches in sensitivity and efficiency. Despite promising results, challenges remain in model generalizability, interpretability, and clinical integration. Future research must focus on explainable, multi-modal AI systems validated across diverse populations to fully harness their potential in improving AS management. Clinicians and researchers should now focus on integrating these validated AI tools into real-world care pathways to enable early intervention and data-driven treatment planning.Key Messages: 1. AI models predict radiographic progression in AS more accurately than traditional scoring systems; 2. Integration of imaging, clinical, and genetic data enhances predictive power and personalization; 3. Explainable and validated AI tools are essential for real-world clinical adoption in AS care.
Keyword :
Ankylosing spondylitis, Radiographic progression, Artificial intelligence, Machine learning, Deep learning, Predictive modelling, mSASSS.