Effect of Scan Body Splinting on Geometric Scan-Body Measurements in Complete-Arch Digital Implant Impressions: An In Vitro Study

Background/Objectives: This in vitro study evaluated the effect of PEEK scan body splinting on geometric scan-body measurements obtained from complete-arch digital implant scans using two intraoral scanners under two simulated clinical scenarios. Methods: Two 3D-printed edentulous maxillary models, each containing six implants at positions #16, #14, #12, #22, #24, and #26 (FDI notation), were evaluated. Model 1 contained parallel implants, whereas Model 2 included selected implants angulated 20° buccally or distally. Each model was scanned with the CS 3600 and TRIOS 3 under non-splinted (NSS) and PEEK-splinted (SS) conditions. Ten repeated scans were acquired per experimental condition (80 scans in total). After alignment to a common reference coordinate system, the scan-body longitudinal-axis vector magnitude (Vector, mm) and a composite angular metric (Δφ, °) were calculated at each implant position. Position-specific differences among NSS, SS, controlNSS, and controlSS groups were evaluated with Kruskal–Wallis tests followed, when appropriate, by Bonferroni-adjusted pairwise comparisons (α = 0.05). Results: Statistically significant NSS–SS differences were observed only at selected implant positions. For the CS 3600, significant differences involved Vector and/or Δφ, whereas for the TRIOS 3 they were confined to Δφ. In every statistically significant NSS–SS comparison, the SS group showed a lower mean value of the corresponding metric. More statistically significant NSS–SS differences were observed in Model 2 for the CS 3600 and in Model 1 for the TRIOS 3. These findings describe differences in the analyzed geometric metrics and should not be interpreted as direct test–reference trueness estimates. Conclusions: PEEK scan body splinting was associated with changes in scan-body geometric measurements at selected implant positions, but the effect was not uniform across scanners or experimental conditions. Because Vector and Δφ were not calculated as direct test–reference deviations, the present data do not establish improved trueness, precision, or overall scanning accuracy. Further studies using direct reference-based error metrics are required to determine clinical relevance.