Recent advances in Large Language Models (LLMs) and multi-agent systems have driven the rise of Agentic AI, showing promise for medical reasoning. However, open-ended conversational agents remain prone to two critical failure modes: premature diagnostic handoff and silent clinical hallucinations that may go undetected before reaching the patient. In this work, we propose a multi-agent framework that addresses both issues by replacing ``LLM-as-a-judge'' routing with deterministic orchestration constraints. The framework incorporates two safety mechanisms. First, a neuro-symbolic state-tracking gate enforces completeness of the OLDCARTS clinical protocol (Onset, Location, Duration, Character, Aggravating/Alleviating factors, Radiation, Timing, and Severity) by blocking diagnostic transitions until all required dimensions are collected. Second, an epistemic uncertainty quantification (UQ) gate computes semantic entropy (H) across K=5 independent diagnostic samples to identify and intercept divergent outputs before delivery. We evaluate the system using simulated patient agents powered by the llama-3.1-70b-instruct model on 150 test cases. The full architecture achieves 49.3% diagnostic precision, representing an absolute improvement of 11.3 percentage points over an unconstrained baseline. Additionally, we observe a statistically significant negative correlation (r = -0.181, p < 0.05) between OLDCARTS completeness (σ) and semantic entropy (H), suggesting that structured information gathering is associated with reduced diagnostic uncertainty.

翻译：近期大型语言模型（LLMs）与多智能体系统的进展推动了Agentic AI的兴起，其在医疗推理领域展现出巨大潜力。然而，开放式对话智能体仍存在两种关键失效模式：过早诊断交接与可能未被察觉即已触及患者的隐性临床幻觉。本文提出一种多智能体框架，通过以确定性编排约束替代"LLM裁判"路由机制，同时解决上述问题。该框架包含两项安全保障机制：其一，神经符号状态追踪门控通过阻断未完整收集必要维度的诊断过渡，强制完成OLDCARTS临床协议（发病、部位、持续时间、性质、加重/缓解因素、放射范围、时间模式、严重程度）的全维度采集；其二，认知不确定性量化门控计算K=5个独立诊断样本的语义熵，在结果输出前识别并拦截发散性输出。我们采用llama-3.1-70b-instruct模型驱动的模拟患者智能体，在150个测试案例上评估系统性能。完整架构实现49.3%的诊断精确率，较无约束基线提升11.3个百分点。此外，OLDCARTS完整度与语义熵之间存在统计显著的负相关关系（r=-0.181，p<0.05），表明结构化信息采集与诊断不确定性降低相关联。