Morphological comparison between artificial intelligence-driven and manual CAD design in single tooth restoration: a preliminary study
Bing-ying Xie, Xiaodong He, Li Hu, Shixian Guo, Jun-lan Chen, Jing Zhang, Xiaoqing Shen, Yuan‐ming Geng, Wei Li
Abstract
OBJECTIVE: The integration of artificial intelligence (AI) into CAD/CAM workflows has revolutionized dental prosthetics manufacturing, yet its morphological trueness compared to manual design remains underexplored. MATERIALS AND METHODS: This study evaluated 30 single-tooth restoration cases from 30 patients. For each case, the original clinically-approved designs were used as reference. AI designs (3Shape Automate) were compared to manual designs created by a technician (3Shape Dental System™). Morphological trueness was evaluated through 3D deviation analysis. Global surface deviations (RMSE) were compared using the Wilcoxon signed-rank test, and maximum discrepancies were compared with a paired Student's t-test, with significance set at p < 0.05. RESULTS: While AI demonstrated batch-processing efficiency, 6.7% of cases (2/30) with suboptimal preparation geometries required manual intervention. No significant difference was found in global surface deviation between AI (median = 79.8 μm) and manual designs (median = 68.6 μm; p = 0.1056). However, AI designs produced significantly greater maximum discrepancies (mean = 225.0 μm) compared to manual designs (mean = 184.4 μm; p = 0.0243). CONCLUSION: These findings validate AI's viability for routine restoration design but emphasize the necessity of case selection protocols and algorithm improvements for dynamic occlusion modeling to ensure comprehensive clinical adoption.