Accuracy of intraoral scanners for full-arch implant scanning in the Maxilla using horizontal scanbody

This laboratory study evaluated the accuracy of six intraoral scanner models (Primescan; Trios 4; Trios 5; Medit i700; i900; Itero Lumina) compared to a desktop scanner (E4) used as a control. Additionally, the influence of the scan body design, horizontal (SBH) and short cylindrical (SBC), was investigated through two different software. A master model containing six implants was used, and the scanning workflow was standardized and repeated ten times (N = 10). The data were analyzed in terms of trueness and precision using two distinct types of analysis: one performed in Blender software (Stichting Blender Foundation) and the other in Evalumap software (Claronav). The lower the number the better precision and accuracy were noted. The position deviation (3D_dev) and angular deviation (Ang_dev) were evaluated. Statistical analyses were conducted using the Kruskal-Wallis test, Bonferroni correction, and multiple comparisons. Trueness (n = 1,437) showed a significant effect on 3D_dev (p < 0.001) and Ang_dev (p < 0.001), as well as for precision (n = 6,478) in 3D_dev (p < 0.001) and Ang_dev (p < 0.001). The lowest median values for trueness in 3D_dev were observed for the i900L (35.25) and in Ang_dev for Primescan (0.35), whereas the lowest precision values for 3D_dev were found for the iTero Lumina (18.81μm) and Ang_dev in i900L (0.19 μm). Regarding the scan body type, a significant difference was observed between SBH and SBC for both trueness and precision, with SBH presenting lower angular deviation values and SBC exhibiting lower 3D position deviations. Multiple comparisons revealed significant differences for the E4 scanner across all analyses. The choice of scanner impacts both the trueness and precision of the generated models, as does the scan body type and the software.