Automated scoring and augmented reality visualization software program for evaluating tooth preparations

Statement of problem: Tooth preparation is an essential part of prosthetic dentistry; however, traditional evaluation methods involve subjective visual inspection that is prone to examiner variability. Purpose: The purpose of this study was to investigate a newly developed automated scoring and augmented reality (ASAR) visualization software program for evaluating tooth preparations. Material and methods: A total of 122 tooth models (61 anterior and 61 posterior teeth) prepared by dental students were evaluated by using visual assessments that were conducted by students and an expert, and auto assessment that was performed with an ASAR software program by using a 3-dimensional (3D) point-cloud comparison method. The software program offered comprehensive functions, including generating detailed reports for individual test models, producing a simultaneous summary score report for all tested models, creating 3D color-coded deviation maps, and forming augmented reality quick-response (AR–QR) codes for online data storage with AR visualization. The reliability and efficiency of the evaluation methods were measured by comparing tooth preparation assessment scores and evaluation time. The data underwent statistical analysis using the Kruskal–Wallis test, followed by Mann–Whitney U tests for pairwise comparisons adjusted with the Benjamini–Hochberg method (α=.05). Results: Significant differences were found across the evaluation methods and tooth types in terms of preparation scores and evaluation time (P<.001). A significant difference was observed between the auto- and student self-assessment methods (P<.001) in scoring both the anterior and posterior tooth preparations. However, no significant difference was found between the auto- and expert-assessment methods for the anterior (P=.085) or posterior (P=.14) tooth preparation scores. Notably, the auto-assessment method required significantly shorter time than the expert- and self-assessment methods (P<.001) for both tooth types. Additionally, significant differences in evaluation time between the anterior and posterior tooth were observed in both self- and expert-assessment methods (P<.001), whereas the evaluation times for both the tooth types with the auto-assessment method were statistically similar (P=.32). Conclusions: ASAR-based evaluation is comparable with expert-assessment while exhibiting significantly higher time efficiency. Moreover, AR–QR codes enhance learning and training experiences by facilitating online data storage and AR visualization.