Comparison of Three-Dimensional Accuracy Between Digital and Conventional Definitive Impressions of Edentulous Arches: A Clinical Study.

Background and Aims: This clinical study evaluated the three-dimensional discrepancies between digital and conventional definitive impressions of edentulous arches for complete denture fabrication. Methods: Ten edentulous patients (n = 16 arches: 8 maxillary and 8 mandibular) were enrolled. For each case, three digital data sets were acquired: (1) a direct intraoral scan (IOS), (2) a scan of a conventional definitive impression obtained using a 3D-printed custom tray (CI3D), and (3) a scan of a conventional definitive impression made with an acrylic special tray (CI). All STL files were first reversed in Geomagic Design X and then aligned using best-fit algorithms in Geomagic Control X, with IOS considered the reference data set. Statistical analyses were performed using SPSS and Python for supporting data visualization. Dimensional discrepancies were evaluated using absolute 3D distance and root mean square (RMS) deviation. Paired-sample t-tests were used to compare CI3D and CI deviations from IOS. Subgroup analyses were conducted for maxillary and mandibular arches, and correlation analysis was performed between deviation metrics. Results: No statistically significant differences were found between CI3D and CI in their deviations from the IOS reference. The overall mean difference between groups was -0.11 ± 1.62 mm, and the RMS difference was -0.10 ± 0.68 mm (p = 0.75 and 93, respectively). Subgroup analysis revealed no significant differences in maxillary or mandibular arches (all p > 0.5). A moderate positive but not significant correlation (r = 0.31, p = 0.21) was observed between mean deviation in CI and RMS deviation in CI3D. Conclusion: Within the limitations of this clinical study, which included edentulous ridges with minimal to moderate bone loss (Cawood and Howell Class III), 3D-printed and acrylic tray-based conventional definitive impressions demonstrated comparable dimensional accuracy to intraoral scanning, showing no statistically significant or clinically relevant differences between techniques.