指南与共识技术与安全

人工智能与医学影像2018:法国放射学界白皮书

Artificial intelligence and medical imaging 2018: French Radiology Community white paper.
指南 / 共识 · Diagn Interv Imaging · 2018-11-22

法国放射学界发布关于人工智能(AI)应用于放射学的立场白皮书,内容涵盖可用算法及术语表、影像数据与协变量及元数据等医疗数据问题、研究与创新、当前与未来应用、AI教育及伦理问题,并在阿西洛马有益AI会议原则基础上提出十项原则以规范AI工具的使用与开发。

要点

  • 白皮书提出十项原则,旨在以患者获益为中心、协调各方方式治理AI工具的使用与开发。
  • 需关注影像数据及协变量、元数据等医疗数据相关问题。
  • 高质量放射学诊疗与大型数据集管理是发展精准、个性化、参与式放射学的两条相关路径。

核心结论

  • 法国放射学界发布AI立场白皮书,提出十项原则以协调方式治理AI工具的使用与开发,确保以患者获益为中心并良好融入临床工作流程。

研究设计与样本

  • 法国放射学界发布的立场白皮书,内容涵盖可用算法与术语表、医疗数据问题、研究与创新、当前与未来应用、AI教育及伦理问题,并在阿西洛马有益AI会议原则基础上提出十项原则。

局限性

  • 原文未报告局限性

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

英文原文(摘要)· 点击展开

The rapid development of information technology and data processing capabilities has led to the creation of new tools known as artificial intelligence (AI). Medical applications of AI are emerging, and the French radiology community felt it was therefore timely to issue a position paper on AI as part of its role as a leader in the development of digital projects. Essential information about the application of AI to radiology includes a description of the available algorithms with a glossary; a review of the issues raised by healthcare data, notably those pertaining to imaging (imaging data and co-variables, metadata); a look at research and innovation; an overview of current and future applications; a discussion of AI education; and a scrutiny of ethical issues. In addition to the principles set forth at the Asilomar Conference on Beneficial AI, the French radiology community has developed ten principles aimed at governing the use and development of AI tools in a manner that will create a concerted approach centered on benefits to patients, while also ensuring good integration within clinical workflows. High-quality care in radiology and opportunities for managing large datasets are two avenues relevant to the development of a precision, personalized, and participative radiology practice characterized by improved predictive and preventive capabilities.

来源:Diagn Interv Imaging,查看原文。版权归原作者所有。

引用

  1. Artificial intelligence and medical imaging 2018: French Radiology Community white paper. Diagn Interv Imaging. 2018. PMID: 30470627. doi:10.1016/j.diii.2018.10.003

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。