This study aims to assess the impact of three different strategies of best-fit (BF) alignments of virtual casts on the estimation of the accuracy of intraoral scanner (IOS) in complete-arch scans. A maxillary typodont, modified with an Implant Scan Body (ISB) in the retroincisive area, was digitized using a desktop scanner (SW Optor Lab) to obtain a reference STL file. The typodont was then scanned 10 times using two IOSs (Trios4, Itero). Each STL file obtained from the IOS was superimposed onto the reference cast using three methods in CloudCompare: full-arch BF (BF-full), BF at the starting tooth (BF-tooth), and BF at the ISB (BF-ISB). Discrepancies from the reference were recorded, and trueness and precision were compared for each method. Statistical analysis with the Kruskall-Wallis nonparametric test was performed (α = 0.05). The Kruskall-Wallis test (p ⟨0.05) revealed statistically significant differences in trueness and precision among the alignment methods. Post Hoc multiple comparison test p-values were all below the critical alpha value. Differences in BF methods lead to significantly different accuracy values of IOS complete-arch scans as different virtual casts' alignment discrepancy. BF-full had the highest accuracy followed by BF-ISB and BF-tooth.
Read full abstract