Mapping Artificial Intelligence Research in Oral and Maxillofacial Surgery: A Bibliometric Analysis.
Authors
Affiliations (2)
Affiliations (2)
- Department of Oral and Maxillofacial Surgery, Guanghua School of Stomatology, Hospital of Stomatology, Sun Yat-sen University, Guangzhou, Guangdong, China.
- Department of Oral and Maxillofacial Surgery, Guanghua School of Stomatology, Hospital of Stomatology, Sun Yat-sen University, Guangzhou, Guangdong, China. Electronic address: [email protected].
Abstract
Oral and maxillofacial surgery (OMFS) produces complex clinical data, while artificial intelligence (AI) applications in this field remain limited despite great potential. This study aimed to map the current status, hotspots, and emerging trends of AI in OMFS. Publications on AI and OMFS before January 10, 2025, were retrieved from the Web of Science Core Collection and Scopus. Data were analysed using CiteSpace and Carrot<sup>2</sup> to construct co-citation networks, perform structural variation analysis, and generate a decade-long term co-occurrence network. A total of 5267 articles were included. The co-citation network showed a research base dominated by AI-driven medical image analysis, with a shift from traditional machine learning to deep learning and transformer models. Carrot<sup>2</sup> clustering highlighted transfer learning and emerging applications in prognostic modeling for head and neck cancer. Structural variation analysis emphasised radiomics and pathology imaging as pivotal domains, while term co-occurrence networks revealed broad applications in radiomics, head and neck cancer, pathway analysis, maxillofacial surgery, and, more recently, orthognathic surgery. AI research in OMFS is currently centered on medical image analysis, particularly radiomics and pathology imaging. Methodological advances have shifted toward deep learning and transformer-based approaches, with ChatGPT as a representative model. Non-imaging applications, including pathway and prognostic analyses, represent promising directions for future integration of AI into OMFS. AI applications, particularly imaging-driven models and transformer-based architectures, offer practical tools for diagnosis, surgical planning, and prognostic assessment in OMFS. Transfer learning enables effective adaptation of AI models to institution-specific datasets, facilitating personalised patient management. By highlighting emerging opportunities in pathway analysis and outcome prediction, this study informs clinicians of actionable AI strategies, supporting evidence-based integration into routine practice and guiding the adoption of novel computational approaches for improved patient care.