Artificial Intelligence Applications in General Surgery in the United States of America: A Bibliometric Analysis

Artificial intelligence (AI) is rapidly transforming surgical practice, with applications spanning preoperative planning, intraoperative guidance, postoperative management, and surgical education. Despite accelerating research activity, the structure, thematic evolution, and funding landscape of AI research in general surgery remain incompletely characterized. This study aimed to systematically evaluate scientific production on AI in general surgery in the United States over the past 5 years using a bibliometric approach.

A bibliometric analysis was conducted following Preliminary Guideline for Reporting Bibliometric Reviews of the Biomedical Literature and Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines using Web of Science. English-language articles published between 2020 and 2025 with a U.S.-affiliated senior author and focused on AI use in general surgery were included. Publications were analyzed across five primary domains: authorship metrics, thematic endpoints, journal characteristics, country of origin, and funding patterns. Bibliometric indicators included H-index, citation counts, Article Influence Score (AIS), and Bradford's Law classification. Funding distribution across endpoints was evaluated using chi-square or Fisher's exact tests, with effect sizes estimated using Cramér's V and odds ratios. Temporal trends in endpoints and keywords were assessed using Poisson and negative binomial regression models.

Fifty-nine studies met inclusion criteria, comprising 20 reviews and 39 original investigations. Scientific production increased consistently from one study in 2019 to 17 in 2023 and 16 in 2024, demonstrating sustained growth. Surgical workflow recognition (n = 19) and clinical decision support (n = 18) were the predominant research domains, representing 63% of the included literature. Temporal analysis demonstrated significant annual growth in reviews (Incidence rate ratios [IRR] 2.09, P = .002) and workflow-focused studies (IRR 1.37, P = .031). Keyword analysis revealed sustained prominence of AI and machine learning, with limited emergence of new thematic directions. Most studies reported no funding (57.6%). Although overall funding distribution did not significantly differ across application categories (P = .846), clinically actionable AI applications were significantly more likely to receive funding compared with other research areas (OR 4.0, 95% CI 1.22-13.13; P = .029).

AI research in U.S. general surgery is growing but remains concentrated in workflow and decision-support domains. Funding favors clinically actionable applications, highlighting the need for broader, equity-focused AI development.