1. Can deep learning-designed anterior tooth-borne crown fulfill morphologic, aesthetic, and functional criteria in clinical practice?
- Author
-
Çakmak G, Cho JH, Choi J, Yoon HI, Yilmaz B, and Schimmel M
- Subjects
- Humans, Male, Female, Adult, Software, Maxilla anatomy & histology, Dental Technicians, Young Adult, Crowns, Esthetics, Dental, Incisor anatomy & histology, Tooth Crown anatomy & histology, Deep Learning, Dental Prosthesis Design methods, Computer-Aided Design
- Abstract
Objectives: This study aimed to compare the design outcomes of anterior crowns generated using deep learning (DL)-based software with those fabricated by a technician using conventional dental computer-assisted design (CAD) software without DL support, with a focus on the evaluation of crown morphology, function, and aesthetics., Methods: Twenty-five in vivo datasets comprising maxillary and mandibular arch scans of prepared maxillary central incisors were utilized to design anterior crowns by using three methods: 1) a DL-based method resulting in as-generated outcome (DB), 2) a DL-based method further optimized by a technician (DM), and 3) a conventional CAD-based method (NC, control). Evaluations were conducted for crown morphology (total discrepancy volume (TDV), root mean square (RMS), positive average (PA) and negative average (NA) deviations), functional aspects (incisal path: deviations, length, and mean inclination), and aesthetics (crown width, height, width-to-height ratio, angular radius of mesioincisal line angle, proximal contact length, and tooth axis angle)., Results: Significant differences in TDV ratio were noted between the DB-NC (32.3 ± 8.5 %) and DM-NC (26.5 ± 5.4 %) pairs (P = 0.006). No significant differences were observed in TDV between the DB-NC (65.3 ± 24.4 mm
3 ) and DM-NC (54.3 ± 21.0 mm3 ) pairs (P = 0.095). For the entire palatal surface, significant differences in RMS and PA values were observed between the DB-NC and DM-NC pairs (P < 0.037). Significant differences in RMS values for the incisal half (P = 0.021) and in PA values for the cervical half (P = 0.047) of the palatal surface were also noted between these pairs. Significant differences in the deviation of the incisal path were observed between the DB-NC (290.4 ± 212.4 μm) and DM-NC (132.0 ± 122.3 μm) pairs (P < 0.001). However, no significant differences were found among the groups (DB, DM, and NC) in terms of the length and mean inclination of incisal paths or in aesthetic outcomes., Conclusions: A DL-based method can result in promising outcomes with clinically acceptable morphology and aesthetics for anterior crowns. Minor deviations in incisal path of the crowns may lead to anterior guidance discrepancies, which can be corrected by the dental technician at the design stage., Clinical Significance: With the potential of DL-based design methods in dental applications, integrating AI technology into dental CAD workflow can enhance the clinical efficiency and consistency of anterior crown design, although human intervention may be required to refine functional aspect., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests that could have appeared to influence the work reported in this paper., (Copyright © 2024 Elsevier Ltd. All rights reserved.)- Published
- 2024
- Full Text
- View/download PDF