• 中国科学论文统计源期刊
  • 中国科技核心期刊
  • 美国化学文摘(CA)来源期刊
  • 日本科学技术振兴机构数据库(JST)

临床输血与检验 ›› 2026, Vol. 28 ›› Issue (5): 667-675.DOI: 10.3969/j.issn.1671-2587.2026.05.001

• 专家论坛 •    下一篇

人工智能在临床用血中的应用与展望:从基础研究迈向临床实践

徐祝帆, 黄龙, 屠嘉欣, 朱长太   

  1. 上海交通大学医学院附属第六人民医院输血科,上海 201300
  • 收稿日期:2026-09-02 修回日期:2026-09-23 出版日期:2026-10-20 发布日期:2026-11-11
  • 通讯作者: 朱长太,主要从事临床输血与实验诊断研究,(E-mail)zct101@163.com。
  • 作者简介:徐祝帆,主要从事临床输血与实验诊断相关研究,(E-mail)18275427463@163.com。

Applications and Prospects of Artificial Intelligence in Clinical Blood Use: From Basic Research to Clinical Practice

XU Zhufan, HUANG Long, TU Jiaxin, ZHU Changtai   

  1. Department of Transfusion Medicine, The Sixth People's Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai 201300
  • Received:2026-09-02 Revised:2026-09-23 Online:2026-10-20 Published:2026-11-11

摘要: 输血医学支撑围手术期、重症医学、血液病及创伤急救等多学科救治。人工智能(AI)技术在此领域展现出广阔前景。本文梳理国内外研究现状,涵盖技术演进、数据基础、临床场景、管理应用及伦理合规。首先回顾AI从数据统计到深度学习的演变,阐述输血数据标准化、特征筛选及知识图谱构建等关键技术;其次总结AI在输血前配血筛查、术中风险监控、输血后预后预判中的应用,及其在血液供需调度、不良反应预警、临床决策辅助系统中的价值;最后探讨数据隐私、算法黑箱及监管适配等伦理法律问题,分析现有模型落地短板,为AI在国内输血科及采供血机构的转化提供理论参考,助力提升用血精准度、效率与安全水平。

关键词: 人工智能, 输血医学, 临床应用, 管理优化, 伦理挑战

Abstract: Transfusion medicine underpins multidisciplinary clinical care in perioperative care, critical care, hematology, and trauma emergencies. Artificial intelligence (AI) has shown broad potential in this field. This review examines the current domestic and international research, covering technological evolution, data infrastructure, clinical applications, management implementations, and ethical and regulatory considerations. First, it outlines the evolution of AI from statistical analysis to deep learning and describes key technologies, including the standardization of transfusion data, feature selection, and knowledge graph construction. Second, it summarizes AI applications in pre-transfusion blood compatibility screening, intraoperative risk monitoring, and post-transfusion prognosis prediction, as well as its value in blood supply-demand coordination, adverse reaction early warning, and clinical decision support systems. Finally, it discusses ethical and legal issues such as data privacy, the "black box" nature of algorithms, and regulatory adaptation; analyzes the limitations in the practical implementation of existing models; and provides a theoretical reference for the translation of AI into transfusion departments and blood collection and supply institutions in China. This review is expected to inform the clinical translation of AI in transfusion medicine and to promote safer, more efficient, and more precise blood utilization in China.

Key words: Artificial intelligence, Transfusion medicine, Clinical applications, Management optimization, Ethical challenges

中图分类号: