Artificial Intelligence in Endodontic Diagnosis and Treatment: A Bibliometric and Science-Mapping Analysis

This study aimed to map global research trends, intellectual structure, and emerging themes in artificial intelligence (AI) applications for endodontic diagnosis and treatment using bibliometric and science-mapping approaches. A comprehensive search of the Web of Science Core Collection was performed on July 20, 2025, without time restrictions. Records were independently screened by two reviewers according to predefined inclusion and exclusion criteria. Bibliometric data were analyzed using Microsoft Excel, RStudio (bibliometrix package), VOSviewer, and CiteSpace. Indicators assessed included annual scientific production, citation impact, leading journals and countries, collaboration networks, keyword co-occurrence, co-citation clusters, and citation bursts. Of the 121 records identified, 72 articles met the inclusion criteria. Scientific production increased markedly after 2022, with more than half of the publications appearing during 2024-2025. The Journal of Dentistry and the Journal of Endodontics demonstrated the highest citation impact. China, the United States, Germany, India, and Turkey were the most productive countries, with China leading in total citations. Four principal thematic clusters were identified, including review-based evidence synthesis, methodological workflows, convolutional neural network–based imaging applications, and future research perspectives. Citation burst analysis revealed emerging interest in segmentation techniques, cone-beam computed tomography–based diagnosis, and deep learning applications. Research on AI applications in endodontics has expanded rapidly, with a strong focus on imaging-driven diagnostic and clinical applications. Future research should emphasize multicenter validation, standardized methodological frameworks, and ethical integration into clinical practice.