Effect of registration methods on accuracy and efficiency of dynamic computer-assisted implant surgery: An in vitro...

Despite rapid development in dynamic computer-assisted implant surgery (d-CAIS) registration methods, a systematic comparison of their accuracy and efficiency is lacking.

The purpose of this in vitro study was to compare the implant placement accuracy and efficiency of 5 different registration methods in d-CAIS.

A standard partially edentulous mandibular model with a replaceable component for inserting a single implant into a polyurethane foam block was designed. D-CAIS was performed using 5 registration methods (n=12 per group): the standard registration (S), the standard screw registration (SS), the orthodontic screw registration (OS), the tooth cusp registration (TC), and the tooth surface registration (TS). The duration of registration was recorded. A postoperative cone beam computed tomography (CBCT) scan was acquired to evaluate the positional accuracy of the implants. Statistical analysis was performed using the Shapiro-Wilk test, 1-way ANOVA, Kruskal-Wallis test, and Bonferroni-adjusted Spearman correlation (α=.05).

For both platform and apex 3-dimensional (3D) deviations, the TC group showed significantly lower accuracy than the S (P<.001 and P<.001 respectively), SS (P<.001 and P<.001 respectively), and OS (P=.009 and P=.008 respectively) groups, exhibiting 1.5 to 2 times of the mean 3D deviation at the platform level than the S group. No significant difference in the angular deviation was observed among the 5 groups (P=.354). The registration duration of the OS and TC groups was significantly longer than that of the S group (P<.001 and P<.001 respectively), with a median duration of more than 2 minutes.

Caution should be exercised when using the cuspate or surficial registration for d-CAIS because of the increased systematic error. Both standard screw and orthodontic screw registration demonstrated accuracy close to the standard registration.