Optimizing complete arch implant digital scans: Influence of scan body height, scanner type, and implant location on...

Intraoral scanning for complete arch implant-supported prostheses has been associated with directional deviations that may compromise prosthetic fit. However, the clinical relevance of these deviations and their relationship with scan body height, scanner system, and implant location remain unclear.

The purpose of this in vitro study was to evaluate the directional deviations (X-axis, Y-axis, and Z-axis) of multiple implants in intraoral scans of edentulous maxilla, depending on scan body height, scanner type, and implant location.

A 3-dimensionally (3D) printed model of an edentulous maxilla with 6 implants, multi-units, and cylindrical polyetheretherketone (PEEK) scan bodies of 3 heights (4.5 mm, 6 mm, and 8 mm) was scanned with 3 intraoral scanners (TRIOS 5 [TR], Primescan [PS], and Infinite [IF]). A total of 144 scans were obtained, and a high-resolution industrial scanner (Artec Micro II) was used to acquire the reference model. Trueness evaluation was conducted using the Geomagic X software program. Directional deviations along the X-, Y-, and Z-axes were calculated for each scan body location. The data were stratified according to scan body height, scanner type, and implant location (molar, premolar, and lateral incisor). Statistical analysis with nonparametric tests (Shapiro-Wilk, Kruskal-Wallis, and post hoc pairwise comparisons) were used to assess differences in absolute deviations among groups (α=.05). Outliers were identified using the z-score method (|z|>3) and removed prior to analysis. Deviations were dichotomized (<90 µm or ≥90 µm), and simple and multiple binary logistic regression models were applied to the X- and Y-axis to evaluate the effects of scanner type, scan body height, and implant location on the likelihood of deviations ≥90 µm. Logistic regression was not performed for the Z-axis because of the absence of sufficient events.

Shorter scan bodies showed significantly lower deviations across all scanners (P<.001). Scan bodies of 4.5 mm produced the smallest deviations on the X- and Z-axes, whereas 6 mm yielded the lowest deviations on the Y-axis (P<.001). Scanner performance was axis-dependent, with IF showing the largest deviations on the X- and Y-axes and increased odds of deviations ≥90 µm. Deviations increased with scan body height for all scanners. Implant location showed greater deviations at posterior sites, but this effect was limited after multivariable adjustment. Z-axis deviations remained consistently low.

Scan body height and scanner type were the main factors influencing directional deviations in complete arch implant scans. Shorter scan bodies improved trueness, whereas taller scan bodies and IF increased the likelihood of clinically relevant deviations. Implant location had a limited effect after adjustment, and Z-axis deviations were minimal. Selecting appropriate scanning strategies may improve accuracy in complete arch implant rehabilitations.