基于人工智能的儿童专科医院门诊精准交互模式的构建与实践
高璇 , 叶成杰 , 钱玉萍 , 田俊华 , 史雨
复旦学报(医学版) ›› 2026, Vol. 53 ›› Issue (01) : 68 -74.
基于人工智能的儿童专科医院门诊精准交互模式的构建与实践
Construction and practice of an artificial intelligence-based precision interaction mode for children’s hospital outpatient services
目的 探索人工智能技术在儿童专科医院智能导诊、智能预问诊及云陪诊服务中的集成应用,构建高效、精准的智慧医疗服务体系。 方法 融合自然语言处理与机器学习技术,设计“DS-小布医生2.0”智能系统,利用患者诊前等待时间,通过微信公众号采集结构化病史信息,并与医院信息系统(Hospital Information System,HIS)实时对接;同步开发云陪诊导航功能,基于室内定位技术优化院内诊疗路径规划。通过试运行数据迭代优化模型算法,建立多层级验证机制以降低信息误差。 结果 复旦大学附属儿科医院智能导诊上线以来,使用量达19.08万人次,“DS-小布医生2.0”运行3个月以来,使用达到14 500余人次,85.17%的智能咨询集中于就诊流程和指引,人工热线相关咨询量分别下降31.32%和4.21%(P<0.05)。使用智能预问诊之后,患者就诊前平均候诊时长由(21.06±3.90) min缩短至(11.88±2.83) min (P<0.05)。运行云陪诊功能后,需要开具药品处方的患者全流程就诊时间由(149±23) min缩短至(134±20) min(P<0.05)。 结论 人工智能驱动的“预问诊-智能导诊-云陪诊”服务模式可优化资源配置、缩短候诊时间、提升医患交互效率,为儿科智慧门诊建设提供了可复制的技术方案。
Objective To explore the integrated application of artificial intelligence (AI) technologies in intelligent triage, pre-consultation, and online patient accompaniment navigation systems at a children’s hospital in order to build an efficient and precise smart medical service system. Methods Integrating natural language processing and machine learning technologies, the “DS-Dr. XiaoBu 2.0” intelligent system was designed. It utilized patients’ pre-consultation waiting time to collect structured medical history information via the hospital’s WeChat official account and interfaces in real-time with the Hospital Information System (HIS). Simultaneously, online patient accompaniment navigation function was developed, optimizing in-hospital diagnosis and treatment path planning based on indoor positioning technology. Operational trial data was used to iteratively optimize the algorithm, and a multi-level verification mechanism was established to minimize information errors. Results Since the implementation of the intelligent triage system in Children’s Hospital, Fudan University, usage reached 190 800 patient visits. Within 3 months of operating the “DS-Dr. XiaoBu 2.0” version, usage reached 14 500 visits, with 85.17% of intelligent inquiries focused on consultation procedures and guidance. Related inquiries to the manual hotline decreased by 31.32% and 4.21%, respectively (P<0.05). The implementation of intelligent pre-consultation significantly reduced the average waiting time before consultation from (21.06±3.90) minutes to (11.88±2.83) minutes (P<0.05). After deploying the online patient accompaniment navigation, the total consultation time for patients requiring medication prescriptions decreased from (149±23) minutes to (134±20) minutes (P<0.05). Conclusion The AI-driven service model encompassing “pre-consultation-intelligent triage-online patient accompaniment navigation function” can optimize resource allocation,reduce waiting times and enhance doctor-patient interaction efficiency, which provided a replicable technical solution for the construction of technology-enhanced pediatric outpatient clinics.
| [1] |
周锐,黄静,范嘉祺.突发公共卫生事件大数据画像构建研究[J].电子政务,2020(6):12-20. |
| [2] |
施曼.人工智能、新质生产力与产业链供应链安全[J].新疆社会科学,2024(6):54-64. |
| [3] |
代金平,覃杨杨.DeepSeek类生成式人工智能赋能中华文明发展传播研究[J].重庆大学学报(社会科学版),2025:1-14. |
| [4] |
吴静.DeepSeek应用下的技术可及性与社会治理挑战——基于生成式人工智能的大规模部署[J].苏州大学学报(哲学社会科学版),2025:1-9. |
| [5] |
上观新闻.跑步入场!上海多家三级医院宣布部署DeepSeek,AI深入医疗腹地[EB/OL].(2025-02-25)[2025-11-07]. |
| [6] |
上海静安.全市首个AI儿童健康管家上线!如何用?[EB/OL].(2025-03-04)[2025-11-07]. |
| [7] |
新闻晨报.国内首个!AI儿科医生“上岗”了[EB/OL].(2025-2-20)[2025-11-07]. |
| [8] |
吴毅剑,张永波,杨凯, |
| [9] |
清华大学.重磅!清华大学成立人工智能医院[EB/OL].(2025-04-27)[2025-11-07]. |
| [10] |
孙中海,赖小琴,颜怿炜.DeepSeek医院私有化部署的实现与应用探索[J].中国卫生信息管理杂志,2025,22(3):456-463. |
| [11] |
张小亮,景慎旗,戴作雷, |
| [12] |
宋彬彬,林君,冯镭, |
| [13] |
|
| [14] |
旷珊珊.面向科室分诊及医生推荐的智能导诊系统研究与设计[D].河北工程大学,2021. |
| [15] |
徐雪珍,陈建萍,顾跃英.门诊挂错号患者经历和心理体验的质性研究[J].护理与康复,2015,14(3):211-213. |
| [16] |
卢晓玲,曲晓童,何传荣.门诊导诊护士感知压力、心理一致感调查及相关性分析[J].齐鲁护理杂志,2022,28(16):39-43. |
| [17] |
李冬冬.护士在体检导诊过程中预防医患纠纷的措施研究[J].临床医药文献电子杂志,2020,7(1):93-100. |
| [18] |
陆康.基于知识图谱的医疗导诊问答系统的设计与实现[D].华中科技大学,2020. |
| [19] |
温源.北京市医疗资源信息整合与公众服务技术研究[D].北京建筑大学,2016. |
| [20] |
孙晓宁,景雨田,刘思琦, |
| [21] |
沈阳,余梦珑.重构智能交互范式:基于DeepSeek的提示强化与人机协同[J].新疆师范大学学报(哲学社会科学版),2025,46(4):90-98. |
| [22] |
房华,许惠翔,杨锐, |
| [23] |
徐挺玉,王兵,索海燕, |
| [24] |
江燕,吕鸣,胡婉婷, |
| [25] |
张靓囡,冯文,李璟媛.我国不同类型医疗机构信息化功能建设现状分析[J].中国卫生信息管理杂志,2022,19(3):400-405,435. |
| [26] |
张晓波,冯瑞,杨睿, |
| [27] |
吴丹麦,张晓丽,侯旭敏.基于大语言模型的智慧医院热点主题演变研究[J].中国卫生质量管理,2025,32(2):1-5. |
| [28] |
王敏怡,何智敏,吴绮常, |
| [29] |
卢子月.基于患者体验的公立医院门诊全流程智慧服务优化研究[D].华中科技大学,2022. |
| [30] |
|
国家卫生健康委医院管理研究所医疗质量(循证)管理研究项目(YLZLXZ24G108)
上海市医院协会医院管理课题(X2024121)
国家儿童医学中心儿童健康政策研究课题(2025RCCHP05)
/
| 〈 |
|
〉 |