Accuracy of complete-arch intraoral scans based on confocal microscopy versus optical triangulation: A comparative in vitro study

Abstract

Statement of problem: Fully digital dentistry is contingent on an accurate digital scan of the complete arch; however, the dimensional accuracy of different scanners for digitizing a complete arch is unclear.

Purpose: The purpose of this in vitro study was to compare the accuracy of 2 intraoral scanners, TRIOS 3 (TR) and CEREC Omnicam (OC). Accuracy was assessed from changes to reference distances defined along the complete arch of a reference cast including 3 precision balls and 3 prepared teeth. The local accuracy (trueness and precision) of the scanned surface of each prepared tooth was also assessed.

Material and methods: Each intraoral scanner was used to scan a metal mandibular reference cast 20 times in a randomized sequence. The complete dental arch of a mandible in which the second left premolar was missing contained 2 prepared teeth, the first left premolar (LP) and the first left molar (LM) to accommodate a fixed partial denture (FPD) with complete crowns. The arch also contained an inlay preparation on the right second premolar (RP). Stainless-steel precision balls (ball centers P₁, P₂, and P₃), which were welded onto the left second molar (P₁), the incisal contact point of the central incisors (P₃), and the right first molar (P₂), were used to determine dimensional changes over the complete arch. Powder (CEREC Optispray) was applied to the reference cast to reduce reflections. Deviations between the reference cast (digitized with high precision before the tests) and the intraoral scans were calculated using Matlab R2015a and Geomagic Design X. To evaluate dimensional changes, the centers of the balls were determined, and their absolute changes in distance (ΔP₁P₂, ΔP₁P₃, and ΔP₂P₃) were calculated. To calculate accuracy (trueness and precision) in relation to the prepared teeth, the mesh deviation between the intraoral scans and the superimposed reference teeth was determined. The data were statistically analyzed by using a nonparametric rank-based 2-way repeated-measures ANOVA, and differences in least square means for pairwise comparisons were calculated (α=.05).

Results: The following mean absolute changes in distance were determined: ΔP₁P₂, TR: 74.4 μm, OC: 119.6 μm; ΔP₁P₃, TR: 24.7 μm, OC: 17.2 μm; ΔP₂P₃, TR: 68.6 μm, OC: 41.2 μm. The scanner did not have a statistically significant effect (P=.118) for distance, and the different distances differed significantly from each other (P<.001). Both scanners provided results acceptable for the fabrication of inlays and short-span FPDs. A complete-crown scan was more accurate than an inlay scan (P<.001). Accuracy and precision were better for TR than for OC (P<.001).

Conclusions: With maximum discrepancies of 192.5 to 294.6 μm across the dental arch, complete-arch scans cannot yet be recommended for the fabrication of long-span FPDs.