人工智能应用于糖尿病的文献计量学分析

Bibliometric Analysis of Artificial Intelligence Applications in Diabetes Mellitus

目的:通过文献计量学分析人工智能(AI)在糖尿病(DM)领域的应用情况,阐明AI在DM领域的研究现状、热点和趋势,为未来的研究提供参考。方法:以Web of Science数据库为来源,检索建库至2024-05-10的AI应用于DM领域的相关研究,运用VOSviewer和CiteSpace软件对纳入研究的发文量、国家、作者、机构和关键词进行文献计量学分析。结果:共获得8007篇文献,2014年1月至2024年5月发文量总体呈上升趋势,美国(2076篇)发文量最多。809名核心作者,共计发文4449篇;发文量最高的作者是Acharya,U. Rajendra (24篇)和被引频次最高的作者是Uelmen,Sacha (2709次)。发文量最高的机构为哈佛医学院(120篇)。AI应用于DR研究的最新热点为machine learning (机器学习)和deep learning (深度学习),其他共现频次较高的关键词是分类、诊断、自我管理、危险因素。结论AI在DM的研究热点为机器学习和深度学习,未来研究可重点关注机器学习应用于DM的分类、诊断、自我管理、危险因素预测。

Objective: To summarize the application of Artificial Intelligence (AI) in the field of Diabetes Mellitus (DM) through bibliometric analysis, to illustrate the current status, hotspots, and emerging trends of AI related research in the field of DM, and to provide references for future research. Methods: Using the Web of Science database as a source, we searched for studies related to the application of AI in the field of DM from the establishment of the database to 2024-05-10. Using VOSviewer and CiteSpace software to conduct a bibliometric analysis of the number of articles, countries, authors, institutions, and keywords included in the studies. Result: A total of 8007 documents were obtained, and the number of publications from January 2014 to May 2024 showed an overall upward trend, with the United States having the largest number of publications (2076). There are 809 core authors who have published a total of 4449 documents;the author with the highest number of publications is Acharya, U. Rajendra (24 publications), and the author with the highest number of citations is Uelmen, Sacha (2709 citations). The institution with the highest number of publications was Harvard Medical School with 120 publications. The newest hotspots for AI in DR research are machine learning and deep learning, and other keywords with high co-occurrence are classification, diagnosis, self-management, and risk factors. Conclusion: The research hotspot of AI in DM is machine learning and deep learning, and future research could focus on applying machine learning to the classification, diagnosis, self-management, and risk factor prediction of DM.