目的 采用文献计量学分析人工智能(artificial intelligence，AI)在医院感染(healthcare-associated infection，HAI)领域中应用的研究热点及发展趋势，为进一步开展相关研究提供参考。方法 计算机检索Web of Science Core Collection 的 Science Citation Index Expanded 数据库中有关 AI 与 HAI 的文献，检索时限为 1994 年 1 月1 日-2024年1月22日。采用VOSviewer(v1。6。19)和CiteSpace(v6。1。R6)软件绘制科研合作网络、关键词等知识图谱，进行文献计量学分析。结果 共纳入文献305篇，早期文献发表数量与被引频次长期维持在很低的水平;2018年后，二者均呈现逐年上升趋势。纳入文献来自50个国家/地区，文献发表数量最高的国家为美国。文献发表数量最多的科研机构为哈佛大学，文献发表数量最高的学者是来自南卡罗来纳州医科大学的Evans HL教授。根据关键词共现与聚类分析，AI在HAI领域的研究主要聚焦在AI算法及技术、HAI的监控与预测、HAI诊断和预测的准确性3个方面。机器学习算法自动化和智能化发展、模型解释性提升和跨领域融合，脓毒血症以及传播机制是未来的研究重点。结论 AI在HAI领域的相关研究已迈入新阶段。我国在这一领域的学术成果较世界先进水平仍有差距，未来可针对热点方向进行更深入的研究。
Bibliometric analysis of the application of artificial intelligence in the field of healthcare-associated infection
Objective To use bibliometrics to identify research hotspots and emerging trends in the use of artificial intelligence(AI)in healthcare-associated infections(HAI),as well as to offer a resource for more relevant research.Methods The literature on AI and HAI from the Science Citation Index Expanded database of the Web of Science Core Collection was retrieved through computer searches,covering the period from January 1,1994,to January 22,2024.VOSviewer(v1.6.19)and CiteSpace(v6.1.R6)software were utilized for bibliometric analysis,creating knowledge maps that include research cooperation networks and keyword analysis.Results A total of 305 documents were included,and both the number of early publications and the frequency of citations were at a very low level for a long time before showing an annual increase trend after 2018.The United States had the most published documents among the 50 countries/regions from where they were sourced.Harvard University was the scientific research institution with the most publications,while Professor Evans HL of the Medical University of South Carolina was the scholar with the most publications.Research on AI in the field of HAI primarily focused on three aspects:AI algorithms and technologies,monitoring and prediction of HAI,and the accuracy of HAI diagnosis and prediction.These findings were based on keyword co-occurrence and clustering analysis.Conclusions A new phase of AI research in the subject of HAI has begun.More in-depth research can be done in the future for the hot direction,as there is still a gap between China's academic accomplishments in this subject and the advanced level of the world.
NETL
NSTL
万方数据
