Aim: Recording maximum intercuspal position (ICP) is critical for many dental procedures. Digital ICP from intraoral scanners (IOSs) produces variable results. This study investigated the sources of error in recording ICP using an IOS and a recently reported method.
Materials and methods: A set of dentate models was scanned three times in a Rexcan DS2 scanner. The models were then scanned six times with a Cerec Omnicam IOS. For each scan, 10 bilateral ‘bite’ scans were performed (n = 6 x 10 bite registrations). Three key points were identified on the first intraoral scan and automatically transplanted onto all subsequent scans. The key point method was validated by using a ‘secondary’ key point transplantation from each scan back to the three laboratory scans, where the location of each point was compared using one-way analysis of variance. Full-arch errors on the intraoral scans were identified by comparing the intermolar key point distances on all intraoral scans against the ‘gold standard’ model scans. Precision of the virtual occlusion was identified by comparing the distance between all upper-lower key point pairs for all intraoral scans using intraclass correlation.
Results: Automatic key points were transplanted to model scans with standard deviations (SDs) in location of ≤ 0.003 mm (upper [maxillary]) and ≤ 0.004 mm (lower [mandibular]) arch. The intermolar width of the intraoral scans had a mean error of 0.183 (± 0.061) mm (upper) and 0.017 (± 0.092) mm (lower) arch. Interocclusal key point separation showed poor reliability across groups, but good precision (SD < 0.022 mm) within groups.
Conclusion: Automatic key points allowed valid linear distance comparisons across repeated scans. Poor trueness and precision in the full-arch intraoral scans adversely affected interocclusal registrations. Bite scan precision had a less detrimental effect on interocclusal registration.