Background : Accurate scanning of the implant position and emerging profile with intraoral scanner (IOS) is crucial for passive fit restoration. However, the accuracy of digital implant impression depends not only on the type of scanner, but also on scanned surface characteristics. Lack of reference points on edentulous areas can complicate the stitching of scanned images and lead to inaccuracies of digital implant impression(DII). Artificial landmarks in edentulous area should improve the scanning accuracy. Aim/Hypothesis : The aim of this in vitro study was to compare the trueness and precision of scanned partially and fully edentulous models with five different IOS when scan bodies and artificial landmarks were attached. Materials and Methods : Two types of maxilla models were printed with Asiga Max 3D printer. In the partially edentulous model instead of missing four posterior teeth, Straumann BL implants were inserted in the first premolar (straight) and the second molar area (tilted 20° mesially). In the fully edentulous model four implants were inserted symmetrically at second incisor (straight) and first molar (tilted 20° mesially) areas. With scan bodies attached, the models were scanned with coordinate measuring maCHINA to form the reference scans. DII was taken with Primescan (version 5.0.1), Trios3 (version 1.18.2.10), Trios4 (version 19.2.2), CS3600 (version 3.1.0), Medit i500 (version 2.0.3) IOS ten times each (n = 10). After that, the tablets of hardened glass-ionomer cement were attached at edentulous areas as artificial landmarks, and the models were scanned with five IOS again. The trueness and precision of distance, angle, and vertical shift parameters between scan bodies were compared in exported scanning data. Results : For partially edentulous model PS IOS showed statistically significantly the best trueness in distance (-25.13 ± 20.77 μm without landmark and-17.45 ± 16.11 μm with landmark), angle (0.22 ± 0.04° without landmark and 0.18 ± 0.04° with) and vertical shift (40.26 ± 12.08 μm without landmark and 36.73 ± 10.82 μm with) scanning without and with artificial landmark of all IOS used. While Trios3 and Trios4 showed similar results like PS IOS, the highest deviations were from CS3600 and Medit i500 data. For the fully edentulous model, PS IOS also showed the best results scanning with landmarks, but deviations were higher in the longest distances between scan bodies in all parameters. CS3600 IOS demonstrated the poorest results in distance, while Medit i500 in angle measurements. The best results of precision had PS IOS for both partially and fully edentulous models in all parameters measured (P < 0.05), followed by Trios4, Trios3, Medit i500, and CS3600. Conclusions and Clinical Implications : Artificial landmarks did not have a statistically significant impact on scanning accuracy nor in partially, nor in fully edentulous models. The longer the distance between scan bodies, the more deviations in the distance, angle, and vertical shift parameters arise, despite the usage of artificia...
