arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2602.01297cs.AI
收藏

RE-MCDF:闭环多专家LLM推理用于知识驱动的临床诊断

RE-MCDF: Closed-Loop Multi-Expert LLM Reasoning for Knowledge-Grounded Clinical Diagnosis

School of Informatics, Xiamen University, China(厦门大学信息学系, 中国);National Institute for Data Science in Health and Medicine, Xiamen University, China(健康医学数据科学国家研究院, 厦门大学, 中国);Key Laboratory of Intelligent Manufacturing Equipment and Industrial Internet Technology, Fujian Provincial Universities, the School of Information Science and Technology, Xiamen University Tan Kah Kee College, and also with the Department of Informatics and Communication Engineering, Xiamen University, China(福建省智能制造装备与工业互联网技术重点实验室, 福建省高校, 厦门大学信息科学系, 厦门大学坦克 Kee 学院, 以及厦门大学信息与通信工程系, 中国);School of Medicine, Xiamen University, China(厦门大学医学院, 中国);School of Electronic and Information Engineering, Tongji University, China(同济大学电子与信息工程学院, 中国);department of Electrical Engineering, University at Buffalo-SUNY, Buffalo, NY, USA(University at Buffalo-SUNY 电气工程系, Buffalo, NY, 美国)

Shaowei Shen, Xiaohong Yang, Jie Yang, Lianfen Huang, Yongcai Zhang, Yang Zou, Seyyedali Hosseinalipour

AI总结:

针对神经科电子病历异质性、稀疏性和噪声问题,提出RE-MCDF框架,通过闭环架构整合三个互补组件,强化疾病间逻辑约束,提升复杂诊断场景性能。

摘要:

电子病历(EMRs),特别是在神经科,本质上是异质的、稀疏的和嘈杂的,这对大语言模型(LLMs)在临床诊断中的应用构成了重大挑战。在这样的环境中,单体系统容易产生自我强化的错误,因为其预测缺乏独立验证且可能偏向于虚假结论。尽管最近的多代理框架试图通过协作推理来缓解这一问题,但它们的交互往往浅显且结构松散,无法反映临床专家所使用的严谨、证据驱动的过程。更根本的是,现有方法大多忽略了疾病之间丰富的逻辑依赖关系,如互斥性、病理兼容性和诊断混淆。这种限制使得它们在有足够证据的情况下也无法排除临床不合理的假设。为了解决这些问题,我们提出了RE-MCDF,一种关系增强的多专家临床诊断框架。RE-MCDF引入了一种生成-验证-修订闭环架构,整合了三个互补组件:(i)主要专家生成候选诊断和支撑证据,(ii)实验室专家动态优先处理异质临床指标,(iii)多关系意识和评估专家组显式执行疾病间的逻辑约束。在医学知识图谱(MKG)的指导下,前两个专家自适应地重新加权EMR证据,而专家组验证并修正候选诊断以确保逻辑一致性。在神经科CMEMR(NEEMRs)子集和我们整理的数据集(XMEMRs)上的大量实验表明,RE-MCDF在复杂诊断场景中始终优于最先进的基线模型(https://github.com/shenshaowei/RE-MCDF)

英文摘要:

Electronic medical records (EMRs), particularly in neurology, are inherently heterogeneous, sparse, and noisy, which poses significant challenges for large language models (LLMs) in clinical diagnosis. In such settings, single-agent systems are vulnerable to self-reinforcing errors, as their predictions lack independent validation and can drift toward spurious conclusions. Although recent multi-agent frameworks attempt to mitigate this issue through collaborative reasoning, their interactions are often shallow and loosely structured, failing to reflect the rigorous, evidence-driven processes used by clinical experts. More fundamentally, existing approaches largely ignore the rich logical dependencies among diseases, such as mutual exclusivity, pathological compatibility, and diagnostic confusion. This limitation prevents them from ruling out clinically implausible hypotheses, even when sufficient evidence is available. To overcome these, we propose RE-MCDF, a relation-enhanced multi-expert clinical diagnosis framework. RE-MCDF introduces a generation--verification--revision closed-loop architecture that integrates three complementary components: (i) a primary expert that generates candidate diagnoses and supporting evidence, (ii) a laboratory expert that dynamically prioritizes heterogeneous clinical indicators, and (iii) a multi-relation awareness and evaluation expert group that explicitly enforces inter-disease logical constraints. Guided by a medical knowledge graph (MKG), the first two experts adaptively reweight EMR evidence, while the expert group validates and corrects candidate diagnoses to ensure logical consistency. Extensive experiments on the neurology subset of CMEMR (NEEMRs) and on our curated dataset (XMEMRs) demonstrate that RE-MCDF consistently outperforms state-of-the-art baselines in complex diagnostic scenarios (https://github.com/shenshaowei/RE-MCDF).

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑