Artificial Intelligence-Assisted Cone-Beam Computed Tomography Analysis in Alveolar Ridge Preservation: A Narrative Review.
Authors
Affiliations (6)
Affiliations (6)
- Centre of Studies for Periodontology, Faculty of Dentistry, Universiti Teknologi MARA (UiTM), Sungai Buloh, Selangor, Malaysia.
- Department of Prosthodontics and Dental Implantology, College of Dentistry, King Faisal University, Al-Ahsa, Saudi Arabia.
- Department of Anatomy, Faculty of Dentistry, Center of Artificial Intelligence and Innovation (CAII), Center of Excellence for Dental Stem Cell Biology, Chulalongkorn University, Bangkok, Thailand.
- Department of Oral and Maxillofacial Surgery, Biruni University, Zeytinburnu/İstanbul, Türkiye.
- Orthodontic Department, Dental Regenerative Research Group, Faculty of Dental Medicine, Universitas Airlangga, Surabaya, Indonesia.
- Department of Dentomaxillofacial Radiology, Faculty of Dental Medicine, Universitas Airlangga, Surabaya, Indonesia.
Abstract
Alveolar ridge preservation (ARP) after tooth extraction depends on accurate three-dimensional bone assessment to optimize implant placement and prosthetic outcomes. Cone-beam computed tomography (CBCT) provides high-resolution volumetric imaging, yet manual interpretation remains operator-dependent and poorly standardized. Artificial intelligence (AI) may automate CBCT-based assessment, but its clinical role in ridge preservation remains insufficiently defined. This narrative review with structured thematic synthesis, reported in accordance with the Scale for the Assessment of Narrative Review Articles (SANRA), examined AI applications in CBCT analysis relevant to ARP, explicitly distinguishing technical performance from clinical usefulness. A structured search of PubMed/MEDLINE, Web of Science, and Scopus (January 2026) identified English-language studies published between 2012 and 2026. Articles were appraised using adapted Joanna Briggs Institute (JBI) criteria and synthesized thematically. Thirty-three articles were retained: 17 AI-focused primary studies forming the core synthesis, seven conventional (non-AI) CBCT-based ARP or implant studies, and nine reviews providing background context. AI-driven segmentation of teeth and alveolar bone consistently achieved Dice similarity coefficients above 0.90 with markedly reduced processing times, and selected architectures generalized across CBCT devices and populations. Multimodal CBCT-intraoral-scan fusion enabled crown-root-bone visualization, while deep-learning image enhancement supported diagnostic quality at reduced radiation doses. However, these findings reflect technical (geometric) accuracy: no included study linked AI output to patient-centered ARP outcomes such as implant survival, aesthetics, complications, or need for re-augmentation. External validation on independent datasets was reported in only four AI studies, and reporting metrics were heterogeneous. Current evidence therefore supports the technical feasibility of AI-assisted CBCT analysis for ridge preservation and implant planning, but not yet its clinical effectiveness. Translation will require prospective multicenter studies with patient-centered endpoints, standardized reporting, cost-effectiveness evaluation, and clear regulatory pathways.