临床文档分类对于将非结构化医疗文本转换为标准化的ICD-10诊断至关重要，但面临着复杂医疗语言、隐私限制和标注数据有限等挑战。大型语言模型（LLMs）在该任务中提供了准确性和效率的提升。本研究评估了八种LLMs的性能和一致性：四种推理模型（Qwen QWQ、Deepseek Reasoner、GPT o3 Mini、Gemini 2.0 Flash Thinking）和四种非推理模型（Llama 3.3、GPT 4o Mini、Gemini 2.0 Flash、Deepseek Chat）在使用MIMIC-IV数据集对临床出院摘要进行分类的表现。使用cTAKES对临床叙述进行结构化处理，模型在三次实验运行中被评估，多数投票决定最终预测。结果显示，推理模型在准确率（71% vs 68%）和F1分数（67% vs 60%）上优于非推理模型，Gemini 2.0 Flash Thinking在准确率（75%）和F1分数（76%）上达到最高水平。然而，非推理模型在一致性（91% vs 84%）上表现更稳定。性能在ICD-10代码上有所不同，推理模型在复杂案例中表现优异，但在抽象类别中表现不佳。研究发现，准确率和一致性之间存在权衡，表明混合方法可以优化临床编码。未来研究应探索多标签分类、领域特定微调和集成方法，以提高模型在实际应用中的可靠性。
英文摘要：
Clinical document classification is essential for converting unstructured medical texts into standardised ICD-10 diagnoses, yet it faces challenges due to complex medical language, privacy constraints, and limited annotated datasets. Large Language Models (LLMs) offer promising improvements in accuracy and efficiency for this task. This study evaluates the performance and consistency of eight LLMs; four reasoning (Qwen QWQ, Deepseek Reasoner, GPT o3 Mini, Gemini 2.0 Flash Thinking) and four non-reasoning (Llama 3.3, GPT 4o Mini, Gemini 2.0 Flash, Deepseek Chat); in classifying clinical discharge summaries using the MIMIC-IV dataset. Using cTAKES to structure clinical narratives, models were assessed across three experimental runs, with majority voting determining final predictions. Results showed that reasoning models outperformed non-reasoning models in accuracy (71% vs 68%) and F1 score (67% vs 60%), with Gemini 2.0 Flash Thinking achieving the highest accuracy (75%) and F1 score (76%). However, non-reasoning models demonstrated greater stability (91% vs 84% consistency). Performance varied across ICD-10 codes, with reasoning models excelling in complex cases but struggling with abstract categories. Findings indicate a trade-off between accuracy and consistency, suggesting that a hybrid approach could optimise clinical coding. Future research should explore multi-label classification, domain-specific fine-tuning, and ensemble methods to enhance model reliability in real-world applications.
Guiding Clinical Reasoning with Large Language Models via Knowledge Seeds
Zhejiang University（浙江大学）；Tencent Youtu Lab（腾讯优图实验室）；Jarvis Research Center（贾维斯研究中心）；Brigham and Women’s Hospital（布莱根妇女医院）；Harvard Medical School（哈佛医学院）
Clinical Document Metadata Extraction: A Scoping Review
University of Minnesota（明尼苏达大学）；Mayo Clinic（妙佑医疗国际）；University of Texas Health Science Center at Houston（德克萨斯大学休斯顿健康科学中心）；Oregon Health & Science University（俄勒冈健康与科学大学）
University of Arizona（亚利桑那大学）；Boston Children’s Hospital（波士顿儿童医院）；Harvard Medical School（哈佛医学院）；University of Wisconsin–Madison（威斯康星大学麦迪逊分校）；Loyola University Chicago（芝加哥洛约拉大学）
Automatic phenotyping is a task of identifying cohorts of patients that match a predefined set of criteria. Phenotyping typically involves classifying long clinical documents that contain thousands of tokens. At the same time, recent state-of-art transformer-based pre-trained language models limit the input to a few hundred tokens (e.g. 512 tokens for BERT). We evaluate several strategies for incorporating pre-trained sentence encoders into document-level representations of clinical text, and find that hierarchical transformers without pre-training are competitive with task pre-trained models.
DiReCT: Diagnostic Reasoning for Clinical Notes via Large Language Models
Premium Research Institute for Human Metaverse Medicine (WPI-PRIMe)（人类元宇宙医学高级研究机构（WPI-PRIMe））；Osaka University（大阪大学）；Department of Cardiology, The First Affiliated Hospital of Dalian Medical University（大连医科大学第一附属医院心内科）；Institute of Computing Technology, Chinese Academy of Science（中国科学院计算技术研究所）；D3 Center, Osaka University（大阪大学D3中心）；Agency for Science, Technology and Research (A*STAR)（科技研究局（A*STAR））
KREL: Automatic Medical Coding via Knowledge-Guided Reasoning over Clinical Evidence with LLMs
The University of New South Wales（新南威尔士大学）；University of Göttingen（哥廷根大学）；Macquarie University（麦考瑞大学）；Beijing Intelligent Decision Medical Technology Co. Ltd（北京智决医疗科技有限公司）
Limitations of Large Language Models in Clinical Problem-Solving Arising from Inflexible Reasoning
Department of Neurology and Neurologic Sciences（神经病学与神经科学系）；Stanford University（斯坦福大学）；Image Guided Therapy and Research Facility（影像引导治疗与研究设施）；University of Dundee（邓迪大学）；Weill Institute of Neurology and Neurosciences（韦尔神经病学与神经科学研究所）；University of California, San Francisco（加州大学旧金山分校）；Department of Systems and Enterprises（系统与企业系）；Stevens Institute of Technology（史蒂文斯理工学院）；Department of EECS（电子工程与计算机科学系）；University of California Berkeley（加州大学伯克利分校）
Coding-Free and Privacy-Preserving Agentic Framework for Data-Driven Clinical Research
Infmedix, Co., Ltd.（Infmedix公司）；Department of Transdisciplinary Studies, Seoul National University（首尔国立大学跨学科研究系）；Healthcare AI Research Institute, Seoul National University Hospital（首尔国立大学医院医疗人工智能研究所）；Department of Medicine, Seoul National University College of Medicine（首尔国立大学医学院医学系）；Department of Transdisciplinary Medicine, Seoul National University Hospital（首尔国立大学医院跨学科医学系）；Department of Radiology, Massachusetts General Hospital and Harvard Medical School（麻省总医院放射科及哈佛医学院）
A rubric landscape for evaluating clinical reasoning in large language models: what exists, what is missing, and what needs to be combined
DRIVE-Health CDT（DRIVE-Health 博士训练中心）；Department of Biostatistics and Health Informatics（生物统计学与健康信息学系）；Institute of Psychiatry, Psychology and Neuroscience（精神病学、心理学与神经科学研究所）；King’s College London（伦敦国王学院）；Cleveland Clinic London（伦敦克利夫兰诊所）
Cognitive bias in LLM reasoning compromises interpretation of clinical oncology notes
Mayo Clinic College of Medicine and Science（梅奥诊所医学与科学学院）；School of Computing and AI, Arizona State University（亚利桑那州立大学计算与人工智能学院）；Mayo Clinic Comprehensive Cancer Center（梅奥诊所综合癌症中心）；Department of Radiation Oncology, Mayo Clinic（梅奥诊所放射肿瘤科）；Department of Oncology, Mayo Clinic（梅奥诊所肿瘤科）；Department of Artificial Intelligence and Informatics, Mayo Clinic（梅奥诊所人工智能与信息学系）；Dana-Farber Cancer Institute, Harvard Medical School（哈佛医学院丹娜-法伯癌症研究所）
Medical Reasoning with Large Language Models: A Survey and MR-Bench
University of Science and Technology of China（中国科学技术大学）；The First Affiliated Hospital of USTC（中国科学技术大学附属第一医院）；National University of Singapore（新加坡国立大学）
Lingdan-V2: Large language models for clinically grounded reasoning in traditional Chinese medicine
School of Computer Science & Technology, Beijing Jiaotong University（北京交通大学计算机科学与技术学院）；Tianjin Tasly Digital Chinese Medicine Technology Co., Ltd.（天津天士力数字中医药科技有限公司）；Tasly Biopharmaceuticals Co., Ltd.（天士力生物制药有限公司）；Affiliated Hospital of Hubei University of Chinese Medicine（湖北中医药大学附属医院）；Da’an Health Technology (Beijing) Co., Ltd.（大安健康科技（北京）有限公司）；Institute of Chinese Medicine Epidemic Disease, Beijing University of Chinese Medicine（北京中医药大学中医疫病研究所）；Guang’anmen Hospital, China Academy of Chinese Medical Sciences（中国中医科学院广安门医院）；Xiyuan Hospital, China Academy of Chinese Medical Sciences（中国中医科学院西苑医院）
Large language models are vulnerable to incidental information in clinical documentation and reasoning
NYU Langone Health（纽约大学朗格尼健康中心）；Johns Hopkins University School of Medicine（约翰霍普金斯大学医学院）；Washington University School of Medicine in St. Louis（圣路易斯华盛顿大学医学院）；University of Alabama at Birmingham Heersink School of Medicine（阿拉巴马大学伯明翰分校希尔斯金医学院）；Perelman School of Medicine at the University of Pennsylvania（宾夕法尼亚大学佩雷尔曼医学院）；New York University（纽约大学）
Berta: an open-source, modular tool for AI-enabled clinical documentation
Department of Medicine, University of Alberta, Edmonton, Alberta, Canada（阿尔伯塔大学医学系，埃德蒙顿，阿尔伯塔，加拿大）；Department of Emergency Medicine, University of Alberta, Edmonton, Alberta, Canada（阿尔伯塔大学急诊医学系，埃德蒙顿，阿尔伯塔，加拿大）；Alberta Health Services, Edmonton, Alberta, Canada（阿尔伯塔省健康服务，埃德蒙顿，阿尔伯塔，加拿大）；Alberta Machine Intelligence Institute, Edmonton, Alberta, Canada（阿尔伯塔人工智能研究所，埃德蒙顿，阿尔伯塔，加拿大）；Department of Computer Science, University of Alberta, Edmonton, Alberta, Canada（阿尔伯塔大学计算机科学系，埃德蒙顿，阿尔伯塔，加拿大）
CLR-voyance: Reinforcing Open-Ended Reasoning for Inpatient Clinical Decision Support with Outcome-Aware Rubrics
ASUS Intelligent Cloud Services (AICS)（ASUS智能云服务（AICS））；Rehabilitation Research Institute of Singapore, Nanyang Technological University（新加坡康复研究院，南洋理工大学）；Department of Family Medicine, Taipei Veterans General Hospital（台北荣民总医院家庭医学部）；School of Medicine, National Yang Ming Chiao Tung University（国家阳明交通大学医学院）；Yong Loo Lin School of Medicine, National University of Singapore（新加坡国立大学 Yong Loo Lin 医学院）
Effect of Document Packing on the Latent Multi-Hop Reasoning Capabilities of Large Language Models
Chandar Research Lab（Chandar研究实验室）；Mila – Quebec AI Institute（魁北克人工智能研究所）；Université de Montréal（蒙特利尔大学）；Microsoft Research（微软研究院）；Polytechnique Montréal（蒙特利尔理工学院）；Canada CIFAR AI Chair（加拿大CIFAR人工智能主席）
MaLei at MultiClinSUM: Summarisation of Clinical Documents using Perspective-Aware Iterative Self-Prompting with LLMs
University of Manchester, UK（曼彻斯特大学）；Modul University Vienna, Austria（维也纳Modul大学）；Leiden Institute of Advanced Computer Science (LIACS), Leiden University, The Netherlands（莱顿先进计算机科学研究所（LIACS）、莱顿大学）；Leiden University Medical Center, The Netherlands（莱顿大学医学中心）
ER-Reason: A Benchmark Dataset for LLM Clinical Reasoning in the Emergency Room
University of California, Berkeley（加州大学伯克利分校）；University of California, San Francisco（加州大学旧金山分校）；Duke University（杜克大学）；University of Alberta（阿尔伯塔大学）；Beth Israel Deaconess Medical Center（贝斯以色列德aconess医疗中心）；UC Berkeley and UCSF（伯克利大学和旧金山分校）
Retro*: Optimizing LLMs for Reasoning-Intensive Document Retrieval
University of Science and Technology of China（中国科学技术大学）；Beijing Academy of Artificial Intelligence（北京人工智能研究院）；Beijing University of Posts and Telecommunications（北京邮电大学）；State Key Laboratory of Cognitive Intelligence（认知智能国家重点实验室）；Hong Kong Polytechnic University（香港理工大学）；Hong Kong University of Science and Technology（香港科技大学）
InformGen: An AI Copilot for Accurate and Compliant Clinical Research Consent Document Generation
University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）；University of Edinburgh（爱丁堡大学）；Carle Illinois College of Medicine（卡尔伊利诺伊医学院）；Seoul National University Bundang Hospital（首尔大学盆唐医院）
ClinDEF: A Dynamic Evaluation Framework for Large Language Models in Clinical Reasoning
ZJU-Hangzhou Global Scientific and Technological Innovation Center, Zhejiang University（浙江大学杭州全球科技创新中心）；ZJU-UIUC Institute, Zhejiang University（浙江大学UIUC研究院）；College of Computer Science and Technology, Zhejiang University（浙江大学计算机科学与技术学院）；The Polytechnic Institute, Zhejiang University（浙江大学 polytechnic 院）；AntGroup（蚂蚁集团）
ClinOCR-Bench: A Comprehensive Clinical Scanned Document Dataset for Optical Character Recognition Model Evaluation
McWilliams School of Biomedical Informatics, University of Texas Health Science Center at Houston（德克萨斯大学健康科学中心休斯顿分校麦威廉斯生物医学信息学学院）；Enterprise Development & Integration, University of Texas MD Anderson Cancer Center（德克萨斯大学MD安德森癌症中心企业开发与集成）
Advances in LLM Reasoning Enable Flexibility in Clinical Problem-Solving
Weill Institute of Neurology and Neurosciences（威尔神经与神经科学研究所）；Carle Illinois College of Medicine（卡乐伊利诺伊医学院）；University of California, San Francisco（加州大学旧金山分校）；University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）；Stanford University（斯坦福大学）；University of Dundee（邓迪大学）；Stevens Institute of Technology（史蒂文斯理工学院）；University of California Berkeley（加州大学伯克利分校）
Mind the Gap: Benchmarking LLM Uncertainty and Calibration with Specialty-Aware Clinical QA and Reasoning-Based Behavioural Features
Department of Medical Informatics, Amsterdam University Medical Center, University of Amsterdam, Amsterdam, The Netherlands（医学信息学系，阿姆斯特丹大学医学中心，阿姆斯特丹大学，阿姆斯特丹，荷兰）；Amsterdam Public Health, Methodology, Amsterdam, The Netherlands（阿姆斯特丹公共健康，方法学，阿姆斯特丹，荷兰）
