This comprehensive survey examines how LLM-based agents are transforming healthcare through intelligent assistance systems that understand, reason about, and support medical tasks.

• Analyzes key components of medical agent systems including profiles, planning mechanisms, reasoning frameworks, and external tools
• Covers diverse applications across clinical decision support, documentation, and healthcare service optimization
• Addresses critical challenges including ethical considerations and hallucination management
• Provides a structured framework for understanding the architecture of medical AI assistants

This research matters because it establishes a foundation for developing more capable, trustworthy AI systems that could revolutionize healthcare delivery while highlighting the current gaps between theoretical potential and practical implementation.

A Survey of LLM-based Agents in Medicine: How far are we from Baymax?