ClinicalAgents: Multi-Agent Orchestration for Clinical Decision Making with Dual-Memory

arXiv cs.CL / 3/30/2026

💬 OpinionSignals & Early TrendsIdeas & Deep AnalysisModels & Research

共有:

Key Points

ClinicalAgents is presented as a multi-agent framework to improve LLM-based clinical decision making by modeling iterative, hypothesis-driven reasoning rather than static symptom-to-diagnosis mappings.
The system uses a dynamic orchestrator based on Monte Carlo Tree Search (MCTS) to generate hypotheses, verify evidence, and backtrack when key information is missing.
It introduces a Dual-Memory architecture with a mutable Working Memory for evolving patient state and a static Experience Memory that retrieves clinical guidelines and historical cases through an active feedback loop.
Experiments reported in the paper claim state-of-the-art results, improving diagnostic accuracy and explainability versus both strong single-agent and other multi-agent baselines.

Abstract

While Large Language Models (LLMs) have demonstrated potential in healthcare, they often struggle with the complex, non-linear reasoning required for accurate clinical diagnosis. Existing methods typically rely on static, linear mappings from symptoms to diagnoses, failing to capture the iterative, hypothesis-driven reasoning inherent to human clinicians. To bridge this gap, we introduce ClinicalAgents, a novel multi-agent framework designed to simulate the cognitive workflow of expert clinicians. Unlike rigid sequential chains, ClinicalAgents employs a dynamic orchestration mechanism modeled as a Monte Carlo Tree Search (MCTS) process. This allows an Orchestrator to iteratively generate hypotheses, actively verify evidence, and trigger backtracking when critical information is missing. Central to this framework is a Dual-Memory architecture: a mutable Working Memory that maintains the evolving patient state for context-aware reasoning, and a static Experience Memory that retrieves clinical guidelines and historical cases via an active feedback loop. Extensive experiments demonstrate that ClinicalAgents achieves state-of-the-art performance, significantly enhancing both diagnostic accuracy and explainability compared to strong single-agent and multi-agent baselines.