Comparing the accuracy of full-arch implant impressions using the conventional technique and digital scans with and without prefabricated landmarks in the mandible: An in vitro study

This study evaluated the accuracy of digital implant impressions with or without prefabricated landmarks compared with the conventional method in the edentulous mandible. An edentulous mandibular stone cast with implant abutment analogs and scan bodies in FDI #46, #43, #33, and #36 served as the master model. The scans captured with intraoral scanners (IOS) were divided into four groups: IOS-NT (no landmarks + Trios 4 scanner), IOS-NA (no landmarks + Aoralscan 3 scanner), IOS-YT (landmarks + Trios 4 scanner), and IOS-YA (landmarks + Aoralscan 3 scanner) (n=10). Landmarks were attached to the scan bodies with resin to improve scanning fluency. Conventional open-trayed technique (CNV) was performed with the 3D-printed splinting frameworks (n=10). The master model and conventional castings were scanned using a laboratory scanner, and the former served as the reference model. Overall distance and angle deviations between scan bodies were measured to determine trueness and precision. The ANOVA or Kruskal–Wallis test compared CNV group to scans without landmarks, while a generalized linear model analyzed scan groups with and without landmarks. Compared to the CNV group, the IOS-NA and IOS-NT groups showed higher overall distance trueness (p=0.009), and precision (distance, p