Artificial intelligence in orthopaedic research: a bibliometric analysis of publication trends in high-impact journals (2016–2024)

Abstract

This bibliometric study quantified the integration of artificial intelligence (AI) into orthopaedic research by analyzing publication trends and thematic evolution in high-impact journals over the past decade. We systematically analyzed articles published between 2016 and 2024 in ten leading orthopaedic journals (SCI Q1) using Web of Science, PubMed, and Scopus databases. Publications were categorized into pre-emergence (2016–2022) and post-emergence (2023–2024) periods based on AI adoption patterns, with the temporal boundary informed by documented ChatGPT release timing and the observed inflection point in AI-related publication volume. AI-related articles were identified through validated keyword searches of titles and abstracts. Publication trends were assessed using time-series analysis including third-order polynomial regression modeling, while Latent Dirichlet Allocation modeling identified thematic shifts. Specific AI methodologies (e.g., convolutional neural networks, random forests, support vector machines) were classified where reported. Statistical significance was determined using chi-square and trend analyses. Total publications increased from 1,487 (2016) to 2,089 (2024), with AI-related content rising from 0.7% to 9.8% (p < 0.001); polynomial regression confirmed an exponential growth pattern (R² = 0.98). Machine learning applications showed 15-fold growth, while deep learning studies increased 12-fold post-2022. Among methodologically specified AI studies (n = 412), convolutional neural networks were most prevalent (38%), followed by random forests (22%), support vector machines (15%), and gradient boosting methods (12%). Topic modeling revealed the emergence of AI-driven themes, including automated diagnostic imaging (14% of AI studies), predictive analytics (12%), and surgical navigation systems (8%). Traditional research domains largely maintained their absolute volume but decreased proportionally due to diversification into AI-related topics. Clinical implementation studies comprised approximately 60% of recent AI publications, suggesting a shift from primarily methodological work toward more application-focused research. AI appears to be increasingly shaping orthopaedic research, contributing to both quantitative growth and qualitative thematic shifts in premier journals. The observed transition from methodological development to clinically oriented studies indicates that AI integration is becoming a recurrent component of orthopaedic research practice. These findings suggest that orthopaedic professionals may benefit from developing AI literacy to keep pace with evolving research paradigms and emerging clinical applications.

Ähnliche Arbeiten