A new method for the evaluation of the accuracy of full-arch digital impressions in vitro

Introducing a new approach to evaluate the accuracy of digital impression methods for full-arch scans, avoiding "best-fit alignment." A lower jaw model with a straight metal bar between the second molars of both quadrants was directly digitized using an intraoral scanner (True Definition, TRD, n = 12) and indirectly digitized (D810, CON, n = 12) after impression and plaster cast. A dataset of the bar from a coordinate measuring machine served as reference (REF). Datasets obtained from test groups were analyzed using inspection software to determine the aberration of the bar length, the linear shift (in X-, Y-, Z-axis) and the angle deviation (alpha (overall), alpha (coronal), alpha (horizontal)) caused by the digitalization method. Mann-Whitney U and unpaired two-sample Student's t test were implemented to detect differences. The level of significance was set at 5 %. Concerning the bar length, no significant differences were found between groups. In view of the linear shift, CON showed significantly higher values than TRD in Y-axis (p = 0.003) and in Z-axis (p = 0.040). Regarding the angle measurement, TRD showed significant smaller values than CON for alpha (overall) (p = 0.006) and for alpha (coronal) (p = 0.005). This in vitro study shows that intraoral scanning systems seem to show the same or even higher accuracy than the conventional impression with subsequent indirect digitalization. Intraoral scanners have proven excellent accuracy for single teeth or small spans. However, insufficient data is available about their accuracy for full-arch scans. The presented new approach seems to be suitable to precisely analyze differences in the accuracy of different digitalization methods without using best-fit alignment.