# Dimensional Accuracy Assessment of 3D-Printed Edentulous Jaw Models: A Comparative Analysis Using Three Laboratory Scanners

## TL;DR

This study compares the accuracy of 3D-printed jaw models using three scanners, finding that all showed clinically significant deviations from the original.

## Contribution

The study provides a comparative analysis of dimensional accuracy in 3D-printed edentulous jaw models using three different laboratory scanners.

## Key findings

- All printed models showed significant deviations from the reference model with RMS values between 109.2–139.7 µm.
- Optical 3D Scanner Vinyl showed the highest consistency, while AutoScan-DS-EX Pro(H) had maximum variability.
- Scanners demonstrated reliable performance, but additive manufacturing introduced clinically significant deviations.

## Abstract

The dimensional accuracy of 3D-printed edentulous jaw models is critical for successful prosthetic treatment outcomes. This study investigated the accuracy of 3D-printed working models of a completely edentulous jaw through comparative analysis of digital images generated by three laboratory scanners. A reference plaster model of a mandibular edentulous arch was digitized and used to produce ten resin models via digital light processing (DLP) technology. Each model was scanned using three different laboratory scanners: AutoScan-DS-EX, AutoScan-DS-EX Pro(H), and Optical 3D Scanner Vinyl. Digital comparison was performed using specialized software, evaluating the root mean square (RMS) deviation and percentage of values within an acceptable deviation range ±0.05 mm. All printed models showed significant deviations from the reference model (p < 0.05), with RMS values ranging from 109.2–139.7 µm and acceptable deviation percentages ranging from 29.34 to 32.31%. The mean precision RMS value was 66.37 µm. The mean intraclass correlation coefficient of 0.544 indicated moderate precision. Optical 3D Scanner Vinyl demonstrated the highest consistency, while AutoScan-DS-EX Pro(H) showed maximum variability. No statistically significant differences were found between scanners (p = 0.075). While the investigated scanners demonstrated reliable performance and sufficient accuracy, the additive manufacturing process introduced clinically significant deviations, highlighting the importance of verification between printed models and their digital originals before proceeding with clinical stages. Clinical practice should prioritize scanning systems with advanced software algorithms over those with superior hardware specifications alone.