Purpose: Sufficient data are not currently available on how the various geometries of scan bodies and different scan strategies affect the quality of digital impressions of implants. The purpose of this study was to present new data on these two topics and give clinicians a basis for decision making. Materials and Methods: A titanium master model containing three Nobelreplace Select TM implants (Nobelbiocare Services AG, Zurich, Switzerland) was digitized using an ATOS industrial noncontact scanner. Digitization was repeated three times with different types of scan bodies integrated into the implants: ELOS A/S, nt-trading GmbH, and TEAMZIEREIS GmbH. These three scans served as virtual master models. The titanium master model was then scanned with the TRIOS3 C digital intraoral scanner (ELOS A/S, Copenhagen, Denmark), which was used for two different scanning strategies. Strategy A was a one-step procedure that included both the titanium master model and the integrated scan bodies. Strategy B comprised two steps. First, a digital overlay was performed with a scan of the titanium master model without integrated scan bodies. A second scan was performed with the titanium master model and integrated scan bodies. By repeating both strategies 10 times for each type of scan body, 60 scans were generated and the corresponding standard tessellation language data sets overlaid with the corresponding virtual master model. Deviations in the resulting superimpositions were calculated and evaluated separately in the individual axes (x, y, z) and in three-dimensional space (Euclidean distance). Statistical evaluation was performed using the R-project software. Level of significance was determined at p ࣘ 0.05. Results: With regard to the geometry of the scan bodies, strategy A significantly influenced the accuracy of the digital implant impression in regards to Euclidean distance (p = 0.003). No significant difference was found for strategy B in this context. Comparing the two scan strategies revealed that strategy A achieved significantly higher accuracy overall (p = 0.031). Conclusion: The quality of digital intraoral impressions seems to be influenced by both the geometry of the scan body and the scan strategy. For clinical practice, the one-step scan strategy seems beneficial. Furthermore, the scan bodies of ELOS A/S showed a potential clinical advantage.

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.

General rightsThis document is made available in accordance with publisher policies. Please cite only the published version using the reference above. Abstract. Machine learning methods can be used for estimating the class membership probability of an observation. We propose an ensemble of optimal trees in terms of their predictive performance. This ensemble is formed by selecting the best trees from a large initial set of trees grown by random forest. A proportion of trees is selected on the basis of their individual predictive performance on out-of-bag observations. The selected trees are further assessed for their collective performance on an independent training data set. This is done by adding the trees one by one starting from the highest predictive tree. A tree is selected for the final ensemble if it increases the predictive performance of the previously combined trees. The proposed method is compared with probability estimation tree, random forest and node harvest on a number of bench mark problems using Brier score as a performance measure. In addition to reducing the number of trees in the ensemble, our method gives better results in most of the cases. The results are supported by a simulation study.

No abstract

No abstract