A CiteSpace-based analysis of the hotspots and trends in the machine learning applications in sarcopenia research
10.3760/cma.j.cn115822-20241217-00234
- VernacularTitle:基于CiteSpace的机器学习应用于肌少症研究热点与趋势分析
- Author:
Jing ZHANG
1
;
Yu WANG
;
Qiong WANG
;
Yu ZHANG
;
Kang YU
Author Information
1. 中国医学科学院 北京协和医学院 北京协和医院临床营养科，北京 100730
- Publication Type:Journal Article
- Keywords:
Sarcopenia;
Machine learning;
CiteSpace;
Bibliometric analysis;
Visualization
- From:
Chinese Journal of Clinical Nutrition
2025;33(5):377-386
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To summarize the applications of machine learning (ML) in sarcopenia research through bibliometric analysis.Methods:Literature on ML applications in sarcopenia research was retrieved on the Web of Science database from its inception to December 2024, along with quality assessment. The CiteSpace software was used to conduct bibliometric analysis of the included literature in terms of publication volume, journals, countries, institutions, authors, co-citations, and keywords.Results:A total of 753 articles were obtained, with 480 being of high quality. There was an overall upward trend in the publication volume from 1999 to 2024. The United States led in total publications (119), with China ranking second (97). At the institutional level, Harvard Medical School led globally (14 publications). Kyung Won Kim had the highest publication count (10), and Alfonso J. Cruz-Jentoft was the most cited author (173 citations). His 2019 paper in Age and Ageing achieved the highest co-citation count. J Cachexia Sarcopenia (with a citation frequency of 208) and Age and Ageing (with a centrality of 0.22) were journals with high impacts in this field. "Body composition" (78 occurrences) was the most frequently occurring keyword apart from the search terms, followed by "computed tomography" (54 occurrences) and "mortality" (47 occurrences). The keyword clustering labels include "prognostic value" (first) and "sarcopenia prediction" (second). Conclusions:Current research hotspots involve applying ML to the diagnosis, risk stratification, prognostic prediction, and health management of sarcopenia. Future trends will focus on its deep integration into comprehensive sarcopenia management and on innovating multimodal data-driven algorithms to improve performance and validation.
