Efficient Computer-Aided Design of Dental Inlay Restoration: A Deep Adversarial Framework

Area of Science:

• Biomedical Engineering
• Computer Science
• Dental Restoration

Background:

• Restoring masticatory function in teeth with large defects is challenging.
• Existing image-to-image methods struggle to generate realistic dental inlay surfaces with critical details.

Purpose of the Study:

• To design a computer-aided framework for automatic reconstruction of realistic defective tooth surfaces.
• To improve the generation of dental inlays with accurate anatomical morphology and occlusal details.

Main Methods:

• Development of a Deep Adversarial-driven dental Inlay reStoration (DAIS) framework.
• Utilizing a Wasserstein generative adversarial network (WGAN) with specialized loss functions.
• Implementation of a local-global discriminator mechanism for enhanced surface reconstruction.

Main Results:

• DAIS efficiently reconstructs realistic occlusal surfaces for teeth with large, arbitrarily shaped defects.
• The generated inlay prostheses exhibit accurate anatomical morphology.
• DAIS demonstrates higher clinical applicability compared to state-of-the-art methods.

Conclusions:

• The DAIS framework provides a novel and effective solution for dental restoration.
• The method successfully generates anatomically correct and clinically applicable dental inlays.
• This approach addresses limitations of current methods in restoring complex tooth defects.