Comparison of Accuracy Between a Conventional and Two Digital Intraoral Impression Techniques

Dental impressions are essential for fabricating restorations and prostheses. Traditional methods rely on polyvinyl siloxane (PVS), while newer digital techniques use intraoral scanners. Prior research has shown that digital methods can offer convenience but may differ in accuracy. However, the extent to which conventional versus digital methods compare in full-arch impressions remains unclear. This gap motivated a direct comparison of accuracy metrics like precision and trueness. No prior work had resolved whether digital systems match conventional standards in full-arch settings. This paper addresses that uncertainty by analyzing deviations in scanned impressions. The study's contribution is to provide empirical data on accuracy differences between established and emerging dental impression methods.

The goal was to assess and compare the accuracy of full-arch dental impressions using conventional PVS and two digital intraoral scanners. The specific problem addressed is whether digital methods can match or exceed traditional techniques in precision and trueness. The motivation stems from the growing adoption of digital systems in clinical settings. The authors aimed to evaluate the reliability of digital impressions in capturing full-arch anatomical details. They focused on measuring deviations in both precision and trueness across three groups. The study was designed to provide data-driven insights into the performance of digital versus conventional impression techniques. The researchers sought to determine if digital systems could be considered viable alternatives to traditional methods. The study's design allowed for a direct comparison of accuracy metrics across all tested methods.

The study involved fabricating full-arch impressions of a reference model using three techniques: conventional PVS and two intraoral optical scanners (Trios and CEREC Omnicam). Each method was applied to the same reference model to ensure consistency. Surface matching software was used to align scans within each group for analysis. Precision and trueness were quantified using mean deviations in micrometers. Statistical analysis included one-way ANOVA with post hoc tests to compare groups. The Bonferroni test was used for trueness, while Games-Howell was applied to precision. Qualitative analysis involved three-dimensional maps of deviations between scans. The methodology ensured that results could be compared objectively across all tested methods.

The conventional PVS method showed a mean precision deviation of 21.7 μm with a standard deviation of 5.4. The Trios scanner had a mean precision of 49.9 μm (±18.3), and the Omnicam had 36.5 μm (±11.12). For trueness, the conventional group had 24.3 μm (±5.7), Trios showed 87.1 μm (±7.9), and Omnicam had 80.3 μm (±12.1). The conventional method had significantly better precision (P < 0.006) and trueness (P < 0.001) than both digital methods. No significant differences were found between the two digital scanners in either metric. The statistical analysis confirmed that conventional impressions outperformed digital ones in accuracy. Local deviations were observed along palatal surfaces and incisal edges, but remained below 100 μm. These findings suggest that conventional impressions may still be more accurate in full-arch settings.

The authors concluded that conventional full-arch PVS impressions showed improved mean accuracy compared to both digital methods tested. They stated that the conventional method had significantly better precision and trueness than either digital scanner. No significant differences were found between the two digital impression systems. The findings suggest that conventional techniques may still be preferable in settings where accuracy is critical. The authors propose that digital systems may still be suitable in certain clinical contexts. They acknowledge that digital methods offer advantages in workflow and patient comfort. However, the data indicate that conventional impressions remain more accurate for full-arch applications. The authors do not claim that digital systems are inherently inferior, but rather that they currently lag in accuracy metrics.