生成式人工智能对临床实践指南制订、评价和应用的影响

罗旭飞; 吕晗; 宋再伟; 刘辉; 王郅翔; 李昊东; 王晔; 朱迪; 张璐; 陈耀龙

doi:10.12290/xhyxzz.2024-0602

生成式人工智能对临床实践指南制订、评价和应用的影响

The Impact of Generative Artificial Intelligence on the Development, Evaluation, and Application of Clinical Practice Guidelines

摘要

摘要: 生成式人工智能(generative artificial intelligence，GAI)是指能够从训练数据生成全新内容(如文本、图像或音频等)的人工智能技术。GAI工具不仅在系统评价的文献筛选、数据提取和文献评价等方面显示出快速、高效的潜力，还可用于指南的评价和传播，提高指南的可读性和推广效率。但GAI工具生成内容的准确性、相关引证证据的合理性、证据等级以及数据的可靠性仍需验证，且数据的隐私保护和伦理问题也是面临的挑战。本文旨在概述GAI工具在指南制订、评价、传播与实施中的应用现状，探索GAI工具在指南领域应用的可行性和新模式，以期提高指南的制订效率和质量，更好地服务指南制订者和使用者。

Abstract: Generative artificial intelligence (GAI) refers to AI technology capable of generating new content such as text, images, or audio from training data. GAI tools not only demonstrate rapid and efficient potential in literature screening, data extraction, and literature appraisal in systematic reviews, but can also be used for guideline evaluation and dissemination, enhancing the readability and promotion efficiency of guidelines. However, the accuracy of content generated by GAI tools, the rationality of cited evidence, the level of evidence, and the reliability of data still need verification. Additionally, data privacy protection and ethical issues are challenges that need to be addressed. This article aims to overview the current status of GAI tools in the formulation, evaluation, dissemination, and implementation of guidelines, explore the feasibility and new models of GAI tools in the field of guidelines, and improve the efficiency and quality of guideline formulation to better serve guideline developers and users.

HTML全文

参考文献(51)

施引文献

资源附件(0)