Artificial Intelligence in MSK Radiology: Current Applications, LLMs, Agentic AI, and Future Directions

The intersection of musculoskeletal (MSK) radiology and artificial intelligence (AI) represents a profound technological evolution in modern medicine. Confronted with exponentially increasing imaging volumes, a shortage of subspecialized radiologists, and the anatomical complexity of osteoarticular pathologies, the radiological community increasingly relies on computational solutions. This comprehensive review delineates the trajectory of AI in MSK radiology, tracing the transition from rudimentary computer-aided diagnosis systems to the sophisticated deep learning networks, multimodal architectures, and large language models defining the present era. Contemporary applications span the entire imaging value chain. In diagnostics, AI algorithms automate fracture detection with high sensitivity comparable to human readers, classify bone tumors via high-throughput radiomics, and execute opportunistic screening for systemic conditions such as osteoporosis. In procedural contexts, AI substantially enhances MSK interventions through advanced needle tracking, thermal mapping, and robotic-assisted orthopaedic surgery. Operationally, AI optimizes workflows by accelerating magnetic resonance imaging acquisition times by up to 75%, intelligently triaging urgent cases, and drafting structured reports in half the traditional time. Looking forward, the paradigm is rapidly shifting toward agentic AI and digital twins, promising highly personalized patient care. Alongside these technological triumphs, this report critically examines the ethical, legal, and educational frameworks required to safely integrate these transformative tools into clinical practice, ensuring AI serves as an augmentative instrument rather than a replacement for human expertise.