Artificial intelligence in shoulder and elbow surgery: a bibliometric analysis of affiliation-based collaboration patterns
No Thumbnail Available
Date
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Description
<p>Background: Despite growing interest, artificial intelligence (AI) applications in shoulder and elbow surgery remain underdeveloped. While adoption is accelerating and shows promise in addressing complex clinical problems, substantial technical and clinical barriers persist. Collaborative research may be relevant for generating high-quality datasets and more robust, generalizable, and clinically relevant algorithms. This study aimed to 1) analyze trends in AI research productivity and impact, 2) map collaboration patterns among affiliations and regions, and 3) assess the relationship between affiliation-based collaboration and research outcomes. Methods: We conducted a bibliometric analysis of Scopus-indexed articles published between January 2000 and November 2024, focusing on peer-reviewed studies involving AI applications in shoulder or elbow surgery. Data collected included number of publications, citation metrics, author affiliations, and index keywords. These variables were used to calculate composite metrics and to examine the geographic distribution of research and collaboration patterns using network analysis. Two linear regression models assessed the relationship between affiliation-based collaborations and publication volume and citation impact. Results: Of 181 identified scholarly documents, 119 met eligibility criteria. These articles were published across 63 journals and cited a total of 1,519 times. The Journal of Shoulder and Elbow Surgery contributed the highest number of articles (n = 20), while Acta Orthopaedica had the highest average citations per article (n = 310), although in a single article. The annual publication rate increased rapidly after 2014, peaking at 49 in 2024. A small group of affiliations disproportionately influenced output and citations. Collaboration networks were sparse (density 0.03) yet showed distinct geographic clusters. Most research output originated from the United States (48%), followed by South Korea (12%) and China (8%). A total of 1,010 collaborations were identified among 260 affiliations. The network showed low density (0.03) but high modularity (0.81), indicating sparse overall connectivity yet tightly clustered communities. Regression models indicated that each additional collaboration established between affiliations was associated with 5 more publications (R<sup>2</sup> = 1.0) and increased average citations per article by 0.2 (R<sup>2</sup> = 0.77). Conclusion: Affiliation-based collaboration was strongly associated with both the volume and citation impact of AI research in shoulder and elbow surgery. Strengthening and expanding these networks may enhance global research participation, foster innovation, and improve the clinical applicability of future work.</p>
Keywords
Affiliation-based collaboration, Artificial intelligence, Bibliometric analysis, Elbow surgery, Literature, Machine learning, Shoulder surgery, Survey Study