Highly Cited Artificial Intelligence Research Studies Published in Neurosurgical Journals: A Bibliometric Analysis.

Neurosurgery is inherently suited for artificial intelligence (AI) integration due to its reliance on advanced technologies. Although AI research is expanding in neurosurgery, citation patterns of impactful publications remain underexplored. We identified AI-focused articles published in neurosurgical journals (NSJs) with ≥100 citations using PubMed and Google Scholar. A total of 66 articles were analyzed. Data collected included article age, country of first author, journal, article type, subspecialty, AI technology used, and application. Citation analysis was performed using mean difference testing across subgroups. The median article age was 8.5 years (range: 1-30 years). Most articles originated from the United States (34, 52%) and were published in Neurosurgery (23, 35%). Technical notes (26, 39%) and review articles (23, 35%) were the dominant formats. General neurosurgery and spine were the leading subspecialties (each 21, 32%). AI technologies included virtual reality (VR) (19, 29%), machine learning (ML) (13, 20%), and augmented reality (10, 15%). The primary applications were surgical planning/assistance (33, 50%) and training (12, 18%). Median citations per article were 173 (range: 100-406). Higher citation rates correlated with review articles, recent publication (within 10 years), and a general neurosurgery focus. AI type and application had no significant impact on citation count. Highly cited AI research in neurosurgery predominantly originates from the United States, focuses on general neurosurgery and spine, and employs VR and ML for planning and training. Citation impact is driven more by study type and recency than by AI modality. Continued original research is vital to integrate AI advancements into standard neurosurgical practice.