Evaluation of the accuracy of digital impressions with different scanning strategies: An in vitro study

To evaluate the effects of three different scanning strategies on the trueness and precision of optical impressions obtained with four intraoral scanners (IOSs). The reference maxillary dental arch model was fabricated using Telio CAD, and the reference digital reference cast was obtained using a computer numerical control machine and an optical scanner (E4, 3Shape, Copenhagen, Denmark). Test scans were performed with four different IOSs (TRIOS3, MEDIT i700, CS 3600, and iTero Element 5D) by an experienced operator and three different scanning strategies (S1: manufacturer-recommended, S2: optimal per previous literature, and S3: experimental). The scan duration was recorded for each scan. All scans were converted to standard tessellation language format and imported into Geomagic Control X. The accuracy was measured by absolute deviation/distance between aligned surfaces. Data of trueness and precision of each IOS and scan duration were statistically compared using analysis of variance for repeated measures and Bonferroni post-hoc test (p < .05). No significant differences in trueness were found among strategies (S1: 9.98 µm, S2: 11.93 µm, S3: 8.84 µm; p = .388) in Trios 3 and iTero Element 5D (S1: 12.24 µm, S2: 11.53 µm, S3: 10.71 µm; p = p = .279). Scanning strategy S3 with MEDIT i700 achieved greater trueness (7.33 µm) than S2 (16.33 µm, p < .05), while no significant difference was noted between S1 (10.44 µm) and S3 (p = .291). S3 showed the highest trueness (16.28 µm) compared to S2 (24.05 µm) and S1 (24.78 µm, p < .001) for CS 3600, with no difference between S1 and S2 (p = .457). Trios 3 had higher precision with S2 (22.46 µm) than S3 (31.69 µm, p < .05), and no significant differences between S1 (25.67 µm) and S2/S3 (p > .05). MEDIT i700 with S3 (29.52 µm) was more precise than both S1 (39.52 µm) and S2 (46.24 µm) (p < .001) with no difference between the last two (p = .302). S2 yielded the highest precision (44.93 µm) compared to S3 (61.81 µm) and S1 (76.53 µm) (p < .001) for CS 3600, with S3 more precise than S1 (p < .001). Similarly, iTero Element 5D showed S2 as the most precise (30.19 µm) compared to S3 (42.80 µm) and S1 (44.45 µm) (p < .05), with no difference between S1 and S3 (p = .472). Scan durations were shorter for S3 and S1 compared to S2 in Trios 3 (p < .001), and S3 was faster than S1 and S2 for MEDIT i700 (p < .001). CS 3600 scans with S1 were quicker than S2 and S3 (p < .001). For iTero Element 5D, no significant differences were f