# Comparison of the superimposition accuracy in the integration of different digital dental models into the cone-beam computed tomography: An ex vivo study

## TL;DR

This study compares the accuracy of integrating digital dental models into CBCT scans using two scanning methods, finding that intraoral scanning may offer better alignment accuracy.

## Contribution

The study provides a novel ex vivo comparison of superimposition accuracy between intraoral and extraoral scanning methods in dental CBCT integration.

## Key findings

- Both intraoral (IS) and extraoral (PM) scanning methods showed clinically acceptable superimposition accuracy.
- IS models demonstrated lower registration errors and smaller deviations compared to PM models, particularly in the mandible.
- Deviations were mostly within -0.5 to 0.5mm, with 80% of values falling within this range for both methods.

## Abstract

Accurate virtual planning for orthognathic surgery and dental implants requires integrating cone-beam computed tomography (CBCT) images into digital dental models, but the impact of different scanning methods on the superimposition accuracy remains unclear. The aim of this study was to compare the superimposition accuracy of dental models obtained from two scanning methods using CBCT scans of dental surfaces.

The maxilla (MX) and mandible (MD) of 4 dry skulls were scanned using CBCT and converted into 3D meshes. Dental arches were obtained using intraoral devices (IS) and by digitizing plaster models using an extraoral scanner (PM). Sixty-four digital models were produced per group. Each 3D mesh and corresponding model were imported into the Geomagic software for alignment and deviation analysis. Registration errors and mean deviations (MD+ and MD-) were assessed quantitatively, and the qualitative evaluation was performed through color maps.

Deviations ranged from -0.4 to 0.3mm in both groups, with 80% of the values distributed between -0.5 and 0.5mm. In both groups, Registration error, MD+ and 3D Error were below 0.5mm, and MD- was higher than -0.5mm. Statistically significant differences were found between IS and PM models for MD+, as well as for registration error and MD+ in mandibular comparisons.

Superimpositions were not affected by the model acquisition method, and both were clinically acceptable. However, PM superimpositions showed greater deviation magnitudes, suggesting an inclination for IS to produce more accurate alignments, especially in the mandible.