Accuracy and reproducibility of virtual edentulous casts created by laboratory impression scan protocols

Statement of Problem. Although computer-aided design and computer-aided manufacturing (CAD-CAM) complete removable dental prostheses (CRDPs) have gained popularity, conventional impressions are still common for CAD-CAM CRDP treatment. These need to be digitized and converted into virtual edentulous casts with a laboratory impression scan protocol during prosthesis fabrication. How this can best be accomplished is unclear. Purpose. The purpose of this in vitro study was to compare the accuracy and reproducibility of virtual edentulous casts created by a dental laboratory laser scanner and a cone beam computed tomography (CBCT) scanner with a digitized master cast. Material and methods. A master cast was digitized as the virtual reference cast. Ten polyvinyl siloxane impressions were made on the master cast and scanned with the dental laboratory laser scanner and CBCT scanner. The impressions were sprayed with antiglare spray and rescanned. Four groups of virtual study casts (N=40) were created from the impression scans. All virtual study casts and the reference cast were registered with surface-matching software, and the root mean square (RMS) values (representation of overall accuracy) and percentage of measurement data points within 1 standard deviation (SD) of mean RMS values (%, representation of overall reproducibility) among the 4 study groups were measured. Additionally, 95 numeric distance differences (representation of accuracy at each region) were measured in 5 distinct regions: the apex of the denture border, 6 mm from denture border, crest of the ridge, palate, and posterior palatal seal. The repeated-measures ANOVA and post hoc test (t grouping) were used to determine statistical differences (alpha=.05). Results. The laboratory scanner group had a significantly larger RMS value (4.0 +/- 0.3 mu m, P<.001) and smaller percentage of measurement data points within 1 SD of mean RMS value (77.5 +/- 1.0%, P<.001). The RMS values between the CBCT scanner (1.2 +/- 0.3 mu m) and CBCT scanner-spray (1.1 +/- 0.2 mu m) groups were not significantly different (P=.968), and the percentage of measurement data points within 1 SD of mean RMS values (90.1 +/- 1.1% versus 89.5 +/- 0.8%) were also not significantly different (P=.662). The numeric distance differences across 5 regions were affected by the scanning protocols (P<.001). The laboratory scanner and laboratory scanner-spray groups had significantly higher numeric distance differences at the apex of the denture border and crest of the ridge regions (P<.001). Conclusions. The CBCT scanner created more accurate and reproducible virtual edentulous casts, and the antiglare spray only significantly improved the accuracy and reproducibility of virtual edentulous casts created by the dental laboratory laser scanner. The accuracy of the virtual edentulous casts was different across 5 regions and was affected by the scanning protocols.