• 文献检索
  • 文档翻译
  • 深度研究
  • 学术资讯
  • Suppr Zotero 插件Zotero 插件
  • 邀请有礼
  • 套餐&价格
  • 历史记录
应用&插件
Suppr Zotero 插件Zotero 插件浏览器插件Mac 客户端Windows 客户端微信小程序
定价
高级版会员购买积分包购买API积分包
服务
文献检索文档翻译深度研究API 文档MCP 服务
关于我们
关于 Suppr公司介绍联系我们用户协议隐私条款
关注我们

Suppr 超能文献

核心技术专利：CN118964589B侵权必究
粤ICP备2023148730 号-1Suppr @ 2026

文献检索

告别复杂PubMed语法，用中文像聊天一样搜索，搜遍4000万医学文献。AI智能推荐，让科研检索更轻松。

立即免费搜索

文件翻译

保留排版，准确专业，支持PDF/Word/PPT等文件格式，支持 12+语言互译。

免费翻译文档

深度研究

AI帮你快速写综述，25分钟生成高质量综述，智能提取关键信息，辅助科研写作。

立即免费体验

一种用于人工智能设计与验证的嵌套模型。

A nested model for AI design and validation.

作者信息

Dubey Akshat, Yang Zewen, Hattab Georges

机构信息

Center for Artificial Intelligence in Public Health Research (ZKI-PH) at Robert Koch Institute, Nordufer 20, 13353 Berlin, Germany.

Department of Mathematics and Computer Science, Freie Universität Berlin, Arnimallee 14, 14195 Berlin, Germany.

出版信息

iScience. 2024 Jul 30;27(9):110603. doi: 10.1016/j.isci.2024.110603. eCollection 2024 Sep 20.

DOI:10.1016/j.isci.2024.110603
PMID:39262809
原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC11387900/
Abstract

The growing AI field faces trust, transparency, fairness, and discrimination challenges. Despite the need for new regulations, there is a mismatch between regulatory science and AI, preventing a consistent framework. A five-layer nested model for AI design and validation aims to address these issues and streamline AI application design and validation, improving fairness, trust, and AI adoption. This model aligns with regulations, addresses AI practitioners' daily challenges, and offers prescriptive guidance for determining appropriate evaluation approaches by identifying unique validity threats. We have three recommendations motivated by this model: (1) Authors should distinguish between layers when claiming contributions to clarify the specific areas in which the contribution is made and to avoid confusion; (2) authors should explicitly state upstream assumptions to ensure that the context and limitations of their AI system are clearly understood, (3) AI venues should promote thorough testing and validation of AI systems and their compliance with regulatory requirements.

摘要

不断发展的人工智能领域面临着信任、透明度、公平性和歧视等挑战。尽管需要新的法规，但监管科学与人工智能之间存在不匹配，阻碍了统一框架的形成。一个用于人工智能设计和验证的五层嵌套模型旨在解决这些问题，简化人工智能应用设计和验证，提高公平性、信任度以及人工智能的采用率。该模型符合法规要求，解决了人工智能从业者的日常挑战，并通过识别独特的有效性威胁，为确定适当的评估方法提供了规范性指导。基于此模型，我们提出三项建议：（1）作者在声称做出贡献时应区分不同层次，以明确做出贡献的具体领域并避免混淆；（2）作者应明确阐述上游假设，以确保其人工智能系统的背景和局限性得到清晰理解；（3）人工智能相关场所应推动对人工智能系统进行全面测试和验证，并确保其符合监管要求。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/3660d2ac8613/gr10.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/02a148580484/fx1.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/7c1d485a5c57/gr1.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/3579d8d2da29/gr2.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/99c47c5c38d2/gr3.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/00f74ea6143f/gr4.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/76d80f9e8871/gr5.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/b96d3946d165/gr6.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/8cba8c50b0fe/gr7.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/d0f315ec1ba5/gr8.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/f4c439c4229c/gr9.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/3660d2ac8613/gr10.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/02a148580484/fx1.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/7c1d485a5c57/gr1.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/3579d8d2da29/gr2.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/99c47c5c38d2/gr3.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/00f74ea6143f/gr4.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/76d80f9e8871/gr5.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/b96d3946d165/gr6.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/8cba8c50b0fe/gr7.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/d0f315ec1ba5/gr8.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/f4c439c4229c/gr9.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c1fd/11387900/3660d2ac8613/gr10.jpg

相似文献

1
A nested model for AI design and validation.一种用于人工智能设计与验证的嵌套模型。
iScience. 2024 Jul 30;27(9):110603. doi: 10.1016/j.isci.2024.110603. eCollection 2024 Sep 20.
2
A nested model for visualization design and validation.嵌套模型用于可视化设计和验证。
IEEE Trans Vis Comput Graph. 2009 Nov-Dec;15(6):921-8. doi: 10.1109/TVCG.2009.111.
3
Toward Fairness, Accountability, Transparency, and Ethics in AI for Social Media and Health Care: Scoping Review.迈向社交媒体和医疗保健领域人工智能的公平性、问责制、透明度和伦理：范围审查
JMIR Med Inform. 2024 Apr 3;12:e50048. doi: 10.2196/50048.
4
The Role of Artificial Intelligence Model Documentation in Translational Science: Scoping Review.人工智能模型文档在转化科学中的作用：范围综述
Interact J Med Res. 2023 Jul 14;12:e45903. doi: 10.2196/45903.
5
A multinational study on artificial intelligence adoption: Clinical implementers' perspectives.一项关于人工智能采用的跨国研究：临床实施者的观点。
Int J Med Inform. 2024 Apr;184:105377. doi: 10.1016/j.ijmedinf.2024.105377. Epub 2024 Feb 15.
6
Fairness of artificial intelligence in healthcare: review and recommendations.人工智能在医疗保健中的公平性：综述与建议。
Jpn J Radiol. 2024 Jan;42(1):3-15. doi: 10.1007/s11604-023-01474-3. Epub 2023 Aug 4.
7
A Mobile App That Addresses Interpretability Challenges in Machine Learning-Based Diabetes Predictions: Survey-Based User Study.一款应对基于机器学习的糖尿病预测中可解释性挑战的移动应用程序：基于调查的用户研究。
JMIR Form Res. 2023 Nov 13;7:e50328. doi: 10.2196/50328.
8
Can I trust my fake data - A comprehensive quality assessment framework for synthetic tabular data in healthcare.我能相信我的虚假数据吗——医疗保健中合成表格数据的综合质量评估框架。
Int J Med Inform. 2024 May;185:105413. doi: 10.1016/j.ijmedinf.2024.105413. Epub 2024 Mar 12.
9
Requirements for Trustworthy Artificial Intelligence and its Application in Healthcare.可信人工智能的要求及其在医疗保健中的应用。
Healthc Inform Res. 2023 Oct;29(4):315-322. doi: 10.4258/hir.2023.29.4.315. Epub 2023 Oct 31.
10
Trustworthy and ethical AI-enabled cardiovascular care: a rapid review.可信且合乎道德的人工智能赋能心血管护理：快速综述。
BMC Med Inform Decis Mak. 2024 Sep 4;24(1):247. doi: 10.1186/s12911-024-02653-6.

引用本文的文献

1
Protocol for implementing the nested model for AI design and validation in compliance with AI regulations.符合人工智能法规的人工智能设计与验证嵌套模型实施协议。
STAR Protoc. 2025 Apr 11;6(2):103771. doi: 10.1016/j.xpro.2025.103771.
2
AI-powered drug discovery for neglected diseases: accelerating public health solutions in the developing world.人工智能助力被忽视疾病的药物研发：加速发展中世界的公共卫生解决方案
J Glob Health. 2025 Jan 10;15:03002. doi: 10.7189/jogh.15.03002.
3
The Way Forward to Embrace Artificial Intelligence in Public Health.

本文引用的文献

1
Guideline for software life cycle in health informatics.健康信息学软件生命周期指南。
iScience. 2022 Nov 9;25(12):105534. doi: 10.1016/j.isci.2022.105534. eCollection 2022 Dec 22.
2
A unifying force for the realization of medical AI.实现医学人工智能的一股统一力量。
NPJ Digit Med. 2022 Nov 15;5(1):172. doi: 10.1038/s41746-022-00721-7.
3
Extending the Nested Model for User-Centric XAI: A Design Study on GNN-based Drug Repurposing.扩展以用户为中心的可解释人工智能的嵌套模型：基于图神经网络的药物重定位设计研究
公共卫生领域拥抱人工智能的前进道路。
Am J Public Health. 2025 Feb;115(2):123-128. doi: 10.2105/AJPH.2024.307888. Epub 2024 Nov 21.
IEEE Trans Vis Comput Graph. 2023 Jan;29(1):1266-1276. doi: 10.1109/TVCG.2022.3209435. Epub 2022 Dec 20.
4
Regulatory Issues and Challenges to Artificial Intelligence Adoption.人工智能采用的监管问题和挑战。
Radiol Clin North Am. 2021 Nov;59(6):1075-1083. doi: 10.1016/j.rcl.2021.07.007.
5
Fostering reproducibility, reusability, and technology transfer in health informatics.促进健康信息学中的可重复性、可重用性和技术转移。
iScience. 2021 Jul 1;24(7):102803. doi: 10.1016/j.isci.2021.102803. eCollection 2021 Jul 23.
6
Regulatory Frameworks for Development and Evaluation of Artificial Intelligence-Based Diagnostic Imaging Algorithms: Summary and Recommendations.基于人工智能的诊断成像算法的开发和评估的监管框架：总结和建议。
J Am Coll Radiol. 2021 Mar;18(3 Pt A):413-424. doi: 10.1016/j.jacr.2020.09.060. Epub 2020 Oct 20.
7
Addressing Bias in Artificial Intelligence in Health Care.应对医疗保健领域人工智能中的偏见问题。
JAMA. 2019 Dec 24;322(24):2377-2378. doi: 10.1001/jama.2019.18058.
8
Regulatory Approval versus Clinical Validation of Artificial Intelligence Diagnostic Tools.人工智能诊断工具的监管审批与临床验证
Radiology. 2018 Sep;288(3):910-911. doi: 10.1148/radiol.2018181310. Epub 2018 Jul 24.