Accuracy of complete-arch digital implant impression with intraoral optical scanning and stereophotogrammetry: an in vivo prospective comparative study

Objectives: To assess accuracy of intraoral optical scanning (IOS) and stereophotogrammetry (SPG), complete-arch digital implant impressions in vivo. Materials and methods: Consecutive patients needing implant-supported screw-retained zirconia complete-arch fixed-dental prostheses (ISZ-FDP) were recruited. For each patient, three impressions were taken: IOS, SPG (tests), and open-tray plaster (reference). Linear (ΔX, ΔY, and ΔZ), three-dimensional (ΔEUC), and angular deviations (ΔANGLE) were evaluated and stratified according to scanning technology for each implant. Potential effects of impression device (IOS and SPG), arch (maxilla and mandible), and implant number (4 and 6) were evaluated through multivariable analysis. Significance level was set at .05. Results: A total of 11 complete arches (5 maxillae, 6 mandibles) in 11 patients were rehabilitated with ISZ-FDPs supported by 4 (n = 8) and 6 implants (n = 3). A total of 50 implants and 100 implant positions were captured by two investigated devices and compared to respective reference (mean ΔEUC IOS 137.2, SPG 87.6 μm; mean ΔANGLE 0.79, 0.38°). Differences between measurements (SPG-IOS) were computed for each implant, with negative values indicating better SPG accuracy. Significant mean ΔEUC difference of -49.60 μm (p = .0143; SD 138.15) and mean ΔANGLE difference of -0.40° (p < .0001; SD 0.65) were observed in favor of SPG. Multivariable analysis showed significant effect on ΔEUC (p = .0162) and ΔANGLE (p = .0001) only for impression devices, with SPG performing better. Conclusions: SPG experienced significantly higher linear and angular accuracy. No effect of type of arch or implant number was detected. Higher extreme deviations were experienced for IOS. SPG can be feasible for complete-arch digital impressions with caution, and rigid prototype try-in is recommended before screw-retained prosthesis manufacturing.