Generative Artificial Intelligence for Dental Morphology Reconstruction and Prosthesis Design: A Scoping ReviewGenerative Artificial Intelligence for Dental Morphology Reconstruction and Prosthesis Design: A Scoping Review

Objectives: To systematically review the rapid advancement of Generative Artificial Intelligence (GenAI) in dental morphology reconstruction and prosthesis design, focusing on the methodological transition from Generative Adversarial Networks (GANs) to Diffusion Models and the clinical expansion from single crowns to complex removable prosthetics.
Data Sources: A scoping review was conducted following PRISMA-ScR guidelines. Major databases (PubMed, Scopus, Web of Science, IEEE Xplore) were searched for articles published up to January 2026.Study Selection: 87 studies meeting the inclusion criteria were analyzed and stratified into a novel four-category clinical framework: (1) Pre-design Automation, (2) Morphology Reconstruction, (3) Fixed Prosthesis Design, and (4) Removable \& Complex Prosthetics. Studies were further classified into Algorithm Development (Type A) and Clinical Evaluation (Type B).
Results: The analysis reveals a significant surge in research activity between 2024 and 2025, driven by the adoption of Diffusion Models and Implicit Representations, which offer enhanced geometric fidelity compared to earlier voxel-based GANs. Clinically, Fixed Prosthesis Design (Category 3) represents a relatively mature and competitive domain, where AI demonstrates a ~78\% efficiency gain over human experts while achieving comparable marginal fit (RMS < 50 µm). Conversely, Removable Prosthetics (Category 4) remains an emerging field, constrained by topological complexity but showing potential through continuous, resolution-independent representations.
Conclusions: Generative AI has transitioned from a theoretical concept to a clinically viable tool for fixed prosthodontics, exhibiting a notable balance between efficiency and quality, where computational speed excels but functional occlusion requires further refinement. Future research should prioritize the development of open-source benchmarks and the application of generative models to complex removable rehabilitations.
Keywords: Generative AI; Diffusion Models; Dental Prosthesis; Scoping Review; Morphology Reconstruction