Page 11 - 《中国医疗器械杂志》2025年第6期

Chinese Journal of Medical Instrumentation 2025年 第49卷 第6期

研 究 与 论 著

文章编号：1671-7104(2025)06-0597-07

医学人工智能研究伦理审查实施路径探析——基于

算法透明和数据安全维度

【作 者】 胡岚岚，刘丹，张玥，李园园，胡钰，张蒙，毛杨，代黔，周建云
陆军军医大学第二附属医院 临床医学研究中心，重庆市，400037
【摘 要】 目的 基于算法透明和数据安全的维度，建立一套切实可行的医学人工智能研究伦理审查机制，有效保障研
究参与者的生命健康、人格尊严及合法权益。方法 梳理国内外医学人工智能研究伦理审查的依据，分析相
关伦理审查实践中存在的问题和难点，结合工作实际，探讨医学人工智能研究伦理审查的程序和要点。结
果 明确了医学人工智能研究的伦理审查需要加强跨学科合作，邀请软件工程、计算机科学等多领域的专家
共同参与审查工作，优化伦理申请与受理的流程，规范伦理审查的方式，并结合医学人工智能研究的不同
类型，制定相应的审查要点，重点关注研究风险受益比的评估和知情同意程序的规范化程度。结论 该研究
在遵循传统伦理审查基本要求的同时，结合人工智能技术的特殊性，提出了规范有效的医学人工智能研究
伦理审查的实施路径，可操作性较强。
【关 键 词】 医学人工智能研究；伦理审查；风险受益评估；知情同意
【中图分类号】 R-052
【文献标志码】 A doi: 10.12455/j.issn.1671-7104.250287
Exploration of Implementation Paths of Ethical Review in Medical
Artificial Intelligence Research: from Perspectives of
Algorithmic Transparency and Data Security

【 Authors 】 HU Lanlan, LIU Dan, ZHANG Yue, LI Yuanyuan, HU Yu, ZHANG Meng, MAO Yang, DAI Qian,
ZHOU Jianyun
Clinical Research Center, The Second Affiliated Hospital, Army Medical University, Chongqing, 400037
【 Abstract 】 Objective From the perspectives of algorithmic transparency and data security to establish a practical
and feasible ethical review mechanism for medical artificial intelligence (AI) research, effectively
safeguarding the life, health, personal dignity, and legitimate rights and interests of research participants.
Methods To sort out the basis for ethical review of medical AI research at home and abroad, analyze the
problems and difficulties in relevant ethical review practices, and explore the procedures and key points of
ethical review of medical AI research based on practical work. Results It has been clarified that the
ethical review of medical AI research needs to strengthen interdisciplinary cooperation, invite experts from
multiple fields such as software engineering and computer science to participate in the review work,
optimize the process of ethical application and acceptance, standardize the way of ethical review, and
develop corresponding review points based on different types of medical AI research, focusing on the
evaluation of research risk-benefit ratio and the standardization of informed consent procedures.
Conclusion This study proposes a standardized and effective implementation path for ethical review of
medical AI research, while adhering to the basic requirements of traditional ethical review and taking into
account the particularity of AI technology. The path is highly practicable.
【Key words】 medical artificial intelligence research, ethical review, risk-benefit assessment, informed consent

收稿日期：2025-04-25
基金项目：陆军军医大学人文社会科学基金项目（2024XRW13）
作者简介：胡岚岚，E-mail: 35341579@qq.com
通信作者：周建云，E-mail: 35541017@qq.com

597