Related Experiment Video
Updated: Jun 16, 2026
Author Spotlight: 3D Movement Assessment of Maxillary Posterior Teeth in Clear Aligner Treatment
Published on: February 23, 2024
A comparison between commercially available artificial intelligence-based and conventional human expert-based digital
Jun-Ho Cho1, Gülce Çakmak2, Eun-Byeol Jee3
1Assistant Professor, Department of Prosthodontics, School of Dentistry and Dental Research Institute, Seoul National University, Seoul, Republic of Korea.
Statement Of Problem:
Studies comparing digital workflows driven by commercially available deep learning (DL)-based design software solutions with conventional digital workflows to explore their potential for ensuring baseline design quality and achieving consistency are lacking.
Purpose:
The purpose of this study was to evaluate whether digital workflows driven by commercially available DL-based design solutions can produce outcomes comparable with those of a dental laboratory technician.
Material And Methods:
Three different anterior complete crowns (n=25) were virtually designed on prepared abutments using 3 distinct digital workflows: 2 commercially available DL-based software solutions, DD (Dentbird Crown) and DA (Automate), and a conventional human-based method (CC). Designed complete crowns from each workflow were evaluated in terms of function (deviations in the incisal path, length, and mean inclination) and esthetics (width-to-height ratio, mesio-incisal line angle, proximal contact length, and tooth axis angle). A noninferiority analysis was performed to investigate whether the 2 DL-based workflows were not inferior to the CC workflow for each parameter of functional and esthetic outcomes. The noninferiority margin was set to ensure clinical relevance, balancing it with statistical sensitivity. The Welch t and Mann-Whitney U tests were used for pairwise comparisons (α=.05).
Results:
The incisal path deviations for the DD-CC and DA-CC pairs were statistically similar (P=.054). Both DD and DA crowns were noninferior to CC crowns in terms of the length of the incisal path (P≤.037). When the mean inclination of the incisal path was considered, DD crowns were noninferior to CC crowns (P=.003). Both DD and DA crowns were statistically noninferior to CC crowns in terms of width-to-height ratio, proximal contact length, and tooth axis angle (P≤.032). For the angular radius, DD crowns were found to be noninferior to CC crowns (P=.015).
Conclusions:
Anterior crowns designed using either the DD or DA workflow achieved similar functional outcomes, including incisal path length and deviation and esthetic outcomes, including width-to-height ratio, proximal contact length, and tooth axis angle, compared with those designed using the CC workflow.
