Empowering Locally Deployable Medical Agent via State Enhanced Logical Skills for FHIR-based Clinical Tasks

2026-03-06

AI Summary

This work addresses the challenge of deploying medical agents locally under constraints of scarce clinical data and stringent privacy requirements, which hinder effective handling of FHIR-based electronic health record (EHR) tasks. To overcome this, the authors propose SELSM, a training-free framework that distills simulated clinical trajectories into entity-agnostic operational rules and employs a query-anchored two-stage retrieval mechanism. During inference, SELSM dynamically invokes logical priors to guide step-by-step reasoning, thereby mitigating state ambiguity. The approach constructs a dynamically updatable, state-augmented cognitive scaffold that enables efficient local adaptation while preserving privacy. Evaluated on MedAgentBench, SELSM achieves a 100% task-chain completion rate with the Qwen3-30B-A3B model and improves overall success rate by 22.67%, substantially outperforming existing memory-augmented baselines.

Technology Category

• Multiagent Systems: Adversarial Agents
• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Machine Learning: Privacy

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

While Large Language Models demonstrate immense potential as proactive Medical Agents, their real-world deployment is severely bottlenecked by data scarcity under privacy constraints. To overcome this, we propose State-Enhanced Logical-Skill Memory (SELSM), a training-free framework that distills simulated clinical trajectories into entity-agnostic operational rules within an abstract skill space. During inference, a Query-Anchored Two-Stage Retrieval mechanism dynamically fetches these entity-agnostic logical priors to guide the agent's step-by-step reasoning, effectively resolving the state polysemy problem. Evaluated on MedAgentBench -- the only authoritative high-fidelity virtual EHR sandbox benchmarked with real clinical data -- SELSM substantially elevates the zero-shot capabilities of locally deployable foundation models (30B--32B parameters). Notably, on the Qwen3-30B-A3B backbone, our framework completely eliminates task chain breakdowns to achieve a 100\% completion rate, boosting the overall success rate by an absolute 22.67\% and significantly outperforming existing memory-augmented baselines. This study demonstrates that equipping models with a dynamically updatable, state-enhanced cognitive scaffold is a privacy-preserving and computationally efficient pathway for local adaptation of AI agents to clinical information systems. While currently validated on FHIR-based EHR interactions as an initial step, the entity-agnostic design of SELSM provides a principled foundation toward broader clinical deployment.

Problem

• Medical Agent
• Data Scarcity
• Privacy Constraints
• FHIR-based Clinical Tasks
• Local Deployment

Innovation

• State-Enhanced Logical-Skill Memory
• Entity-Agnostic Reasoning
• Query-Anchored Two-Stage Retrieval
• FHIR-based EHR
• Zero-Shot Medical Agent