Mapping the global landscape of artificial intelligence in infertility research: A bibliometric study of publications from 2005 to 2025
Hui Yang, Dan Li, Lijuan LiuObjective
Despite growing use of artificial intelligence (AI) in infertility management, systematic bibliometric mapping of its evolution remains scarce. This study maps the research landscape and tracks AI applications in infertility over two decades.
Methods
Publications on AI and infertility (2005–2025) were retrieved from the Science Citation Index Expanded (SCI-Expanded) of the Web of Science Core Collection (WoSCC). Bibliometric analysis used VOSviewer, CiteSpace, Bibliometrix, and Microsoft Excel. Publication trends, countries, authors, journals, funding, keywords, and citation bursts were analyzed.
Results
A total of 513 eligible publications were retrieved, accumulating 6,479 citations (average 12.63 per article). Annual output surged after 2018, peaking at 161 articles in 2025. China ranked first with 150 publications (29.24%), followed by the United States with 129 publications (25.15%). Keyword co-occurrence analysis identified four thematic clusters: diagnostic models for female infertility, AI-assisted male infertility diagnosis, assisted reproductive technology (ART) outcome prediction, and deep learning-based reproductive morphology assessment. Three phases emerged: traditional machine learning for male infertility diagnosis (2011–2017), deep learning-driven morphological evaluation (2018–2020), and clinical translation for ART outcome prediction (2021–2025). Citation burst analysis identified computer vision, chronic endometritis, and non-obstructive azoospermia (NOA) as hotspots.
Conclusion
AI research on infertility has shifted from methodological development toward clinical translation, with growing emphasis on personalized diagnostics and outcome prediction. This mapping underscores the need for external validation, international collaboration, explainable artificial intelligence (XAI), and standardized reporting to accelerate translation into patient-centered reproductive care. Future directions may include AI-driven tools for polycystic ovary syndrome (PCOS), NOA, and dynamic embryo assessment.