Accuracy of registration between digitized extraoral scan bodies and virtual casts: Effect of the edentulous area, tooth anatomy, and registration method

Abstract

Statement of problem

Digitized analog records have been used for the superimposition of intraoral and facial scans. However, the discrepancy in the registration between the digitized occlusal records contained on extraoral scan bodies and the maxillary virtual cast remains uncertain.

Purpose

The purpose of this in vitro study was to evaluate the effect of the registration method, edentulous area, and tooth anatomy on the accuracy of the alignment between the digitized extraoral scan body (ESB) and the maxillary virtual cast.

Material and methods

A scannable ESB and a set of 8 maxillary casts (2 completely dentate simulating unworn and worn tooth anatomy, 5 partially edentulous, and 1 completely edentulous were printed (Pro 95S; SprintRay). Four zirconia markers were attached to the index of the ESB and each of the evaluated casts. Each cast was positioned into the tray of the ESB using occlusal registration material (O-Bite; DMG). The ESB and each corresponding evaluated cast were digitized by using a calibrated laboratory scanner (T710; Medit). Then, each cast and index of the ESB were scanned separately by using the same scanner (n=10). Using a CAD software program, each virtual cast was attached to the ESB 10 times by using 3 different alignment methods: an analog using an iterative closest points (ICP) algorithm and 2 semimanual alignments using a best-fit algorithm for the entire data set with or without including the edentulous areas. A metrology software program was used to measure the linear distance between the corresponding gauge balls and the angulation between the planes defined by the markers on the cast and the ESB. The measurements from the scans of the casts attached to the tray were used as a reference to calculate the discrepancies in each experimental group. Α P value threshold of <.05 was used to determine statistical significance.

Results

The best-fit algorithm registration method produced better trueness and precision than the manual point-to-point registration (P<.05). When the edentulous areas were not included in the analog surface record, the trueness and precision of the best-fit algorithm were significantly worse (P<.05). In respect of tooth anatomy, no significant difference in trueness and precision was found among the investigated groups (P>.05). The completely dentate groups presented significantly better trueness than the edentulous groups (P<.05).

Conclusions

The accuracy of the registration between digitized occlusal surface scans and digital casts was influenced by the registration method, as well as by the location and extent of the edentulous areas.