Effect of AI-Assisted Crown Design on Marginal And Internal Adaptation

Abstract

Background:
Artificial intelligence (AI) has emerged as a promising technology in digital dentistry, particularly in the automated design of dental crowns. Although AI-assisted systems have demonstrated potential in reproducing dental morphology and improving the efficiency of digital workflows, their effect on the marginal and internal adaptation of CAD/CAM-fabricated crowns requires further evaluation.

Aim:
The present in vitro study was conducted to evaluate the effect of AI-assisted crown design on the marginal and internal adaptation of CAD/CAM-fabricated dental crowns in comparison with conventional CAD crown design.

Materials and Methods:

Thirty standardized artificial maxillary first molar preparations were included in the study and randomly divided into two equal groups of 15 specimens each. In Group I, crowns were designed using an AI-assisted dental crown-design software, whereas in Group II, crowns were designed using conventional CAD software. The same scanning protocol, restorative material, CAD/CAM milling system, and manufacturing procedure were used for both groups. Marginal and internal adaptation was evaluated using a three-dimensional replica technique. Internal adaptation was assessed at the cervical, axial, and occlusal regions. The thickness of the replica material was measured using a digital microscope and recorded in micrometres (µm). The marginal gap was considered the primary outcome, while cervical, axial, and occlusal internal gaps were considered secondary outcomes. The data were analyzed using an independent-samples t-test, with the level of statistical significance set at p < 0.05.

Results:
The AI-assisted crown-design group demonstrated a mean marginal gap of 72.4 ± 8.6 µm compared with 86.7 ± 10.1 µm in the conventional CAD group (p < 0.001). The mean cervical internal gap was 78.3 ± 9.2 µm in the AI-assisted group and 91.5 ± 10.4 µm in the conventional CAD group (p = 0.001). At the axial region, the mean internal gap was 94.6 ± 11.3 µm and 106.8 ± 12.1 µm, respectively (p = 0.008). At the occlusal region, the corresponding values were 118.2 ± 15.4 µm and 132.5 ± 16.2 µm, respectively (p = 0.020). Thus, lower marginal and internal gap values were observed in the AI-assisted crown-design group at all evaluated locations.

Conclusion:
Within the limitations of the present in vitro study, AI-assisted crown design demonstrated lower marginal and internal gap values than conventional CAD crown design. The findings suggested that the crown-design approach influenced the adaptation of CAD/CAM-fabricated crowns. Further investigations using natural teeth, different restorative materials and AI systems, cementation, artificial aging, and clinical conditions are required to establish the clinical relevance of these findings.