上海口腔医学

上海口腔医学 ›› 2026, Vol. 35 ›› Issue (3): 319-324.doi: 10.19439/j.sjos.2026.03.016

• 论著 • 上一篇    下一篇

基于知识图谱的舌恶性肿瘤病案首页编码精准化体系的构建

周慧宁1, 杨益多1, 王衍鸣1, 刘苏泰1, 何柳2, 王延安3,*, 徐惠1,*   

  1. 1.上海交通大学医学院附属第九人民医院 病案管理办公室, 2.医务处,上海 200011;
    3.上海交通大学医学院附属第九人民医院 口腔颌面-头颈肿瘤科,上海交通大学口腔医学院,国家口腔医学中心,口腔疾病国家临床医学研究中心,上海市口腔医学重点实验室,上海市口腔医学研究所,上海 200011
  • 收稿日期:2025-12-09 修回日期:2026-02-08 发布日期:2026-07-02
  • 通讯作者: 王延安,E-mail：Wangyan_an@163.com;徐惠,E-mail：13761052512@163.com。*共同通信作者
  • 作者简介:周慧宁（1995—）,女,硕士,E-mail：ningzhhi@163.com
  • 基金资助:
    上海申康医院发展中心诊疗技术推广及优化管理项目（SHDC12026616）

Construction of a precision medical record homepage system for malignant tumors of the tongue based on knowledge graph

Zhou Huining1, Yang Yiduo1, Wang Yanming1, Liu Sutai1, He Liu2, Wang Yan'an3, Xu Hui1   

  1. 1. Medical Records Management Office, 2. Medical Affairs Department, Shanghai Ninth People's Hospital, Shanghai Jiao Tong University School of Medicine. Shanghai 200011;
    3. Department of Oromaxillofacial Head and Neck Oncology, Shanghai Ninth People's Hospital, Shanghai Jiao Tong University School of Medicine; College of Stomatology, Shanghai Jiao Tong University; National Center for Stomatology; National Clinical Research Center for Oral Diseases;Shanghai Key Laboratory of Stomatology; Shanghai Research Institute of Stomatology. Shanghai 200011, China
  • Received:2025-12-09 Revised:2026-02-08 Published:2026-07-02

PDF (PC)

4

摘要： 目的: 聚焦舌恶性肿瘤住院病案首页书写中存在的诊断术语不规范、手术编码不准确等质量问题,利用知识图谱技术,实现住院病案首页临床编码决策流构建。方法: 对2024年6月—2025年6月上海交通大学医学院附属第九人民医院口腔颌面-头颈肿瘤科的12 201份住院病案首页质量进行分析,汇总存在的问题及原因,基于知识图谱的抽取和融合,并结合舌癌临床诊治专家和编码专家意见,针对性设计临床编码决策流。结果: 12 201份住院病案中,8 117份病案首页出现错误（共12 889处）。通过柏拉图统计分析发现,主要诊断错填、其他诊断遗漏、主要手术错填以及其他手术错填是病案首页质量问题的前4位,占87.86%。基于临床实际痛点,构建涵盖流行病学与病理学、检查与诊断、治疗与用药以及预后与随访4个层面的知识图谱,通过向5名舌癌诊疗方面的临床专家和2名编码专家进行咨询,构建出以临床医师诊疗思路为导向的编码决策流。结论: 基于知识图谱构建的临床编码决策流,创新性地将临床医师融入病案质量管理中,将临床诊疗思路与首页编码进行有机结合,自源头提升住院病案首页填写质量,为医院高质量发展提供强有力的数据支撑。

关键词: 知识图谱, 舌恶性肿瘤, 临床编码决策流, 病案质量管理, 管理体系创新

Abstract: PURPOSE: To address quality issues in the documentation of medical record homepages for malignant tumors of the tongue, such as non-standard diagnostic terminology and inaccurate surgical coding, this study employs knowledge graph technology to construct a clinical coding decision-making workflow for inpatient medical record homepages. METHODS: The quality of 12 201 medical record homepages from the Department of Oromaxillofacial Head and Neck Oncology, Shanghai Ninth People's Hospital, Shanghai Jiao Tong University School of Medicine, between June 2024 and June 2025 was analyzed, and the existing problems and their causes were summarized. Based on knowledge graph extraction and fusion, combined with expert opinions from clinicians and coding specialists in tongue cancer treatment, a targeted clinical coding decision-making workflow was designed. RESULTS: Among the 12 201 inpatient medical records, totally 8 117 had errors on their homepages (12 889 errors in all). Through Pareto statistical analysis, it was found that incorrect principal diagnosis, omitted additional diagnoses, incorrect principal procedure, and incorrect other procedures were the top four medical record homepage deficiencies, accounting for 87.86%. Addressing these clinical challenges, a knowledge graph encompassing epidemiology and pathology, examination and diagnosis, treatment and medication, and prognosis and follow-up was constructed. Consultations with five clinical experts in tongue cancer treatment and two coding specialists informed the development of a clinician-oriented coding decision-making workflow. CONCLUSIONS: The knowledge graph-based clinical coding decision-making workflow innovatively integrates clinicians into medical record quality management, combining clinical diagnostic logic with homepage coding to improve documentation quality at its source, which provides robust data support for high-quality hospital development.

Key words: Knowledge graph, Malignant tumor of the tongue, Clinical coding decision-making workflow, Medical record quality management, Management system innovation

中图分类号: