Diagnostic Performance of Artificial Intelligence in Detecting Periodontal Diseases Using Cone-Beam Computed Tomography: A Systematic Review.
Authors
Affiliations (3)
Affiliations (3)
- Shanghai Engineering Research Center of Tooth Restoration and Regeneration & Tongji Research Institute of Stomatology & Department of Implantology, Shanghai Tongji Stomatological Hospital and Dental School, Tongji University, Shanghai, China.
- Division of Applied Oral Sciences and Community Dental Care, Faculty of Dentistry, The University of Hong Kong, Hong Kong, China.
- Division of Computer Science, School of Computing and Data Science, The University of Hong Kong, Hong Kong, China.
Abstract
Recent advances in artificial intelligence (AI) have shown promise for the detection of periodontal disease. This study aimed to synthesize current evidence on the diagnostic performance of AI models applied to cone-beam computed tomography (CBCT) for periodontal disease. A comprehensive literature search of MEDLINE, Scopus, Web of Science, and Embase was conducted to identify studies on AI-assisted diagnosis of periodontal disease using CBCT published up to December 2025. The Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) tool was used to assess the risk of bias. Among 10,099 records screened, eight studies met the inclusion criteria. These studies indicate that AI models, predominantly based on convolutional neural network (CNN) architectures applied to CBCT images, have substantial potential for detecting periodontal bone loss and related conditions. However, diagnostic performance metrics varied markedly across studies, and direct comparisons with human clinicians produced inconsistent findings, underscoring the need for more rigorous study designs and standardized datasets. Despite limitations arising from heterogeneous methodologies and non-standardized data collection, CBCT-based AI models show considerable promise for enhancing periodontal disease diagnosis. Further high-quality, methodologically robust research is warranted to validate and optimize these approaches for clinical use. PROSPERO: CRD420251273846.