Survey Maps Medical Embodied AI for Next-Gen Healthcare.

Foundation models have significantly improved efficiency across various healthcare applications. However, their limited ability to interact with the physical world restricts their utility in real-world clinical settings, where critical decisions and physical actions are closely linked. This gap is being addressed by the emerging field of embodied artificial intelligence (AI), which enables agents to operate within complex medical environments. As research in medical embodied AI rapidly expands, there's a growing need to understand how these intelligent agents function as integrated, end-to-end systems in clinical workflows. Existing surveys often focus on individual components, lacking a holistic view. This new survey aims to consolidate recent advancements by systematically reviewing the core components of medical embodied AI, emphasizing the coordinated integration of perception, decision-making, and action. The paper further reviews key medical applications and relevant datasets, while also analyzing the significant challenges encountered in practical clinical use. It concludes by discussing crucial directions for future research in this evolving domain, providing a valuable resource for researchers and practitioners.