Evaluation of scan abutment and model trueness and precision in 3D printed dental models with different resin viscosities and conventional dental stone models

The aim of this in vitro study was to evaluate the trueness and precision of additively manufactured (AM) and stone models, as well as the scan bodies (SBs) within these models, produced using model resins with different viscosities. A reference model was created by placing four parallel implants in a bilaterally edentulous prefabricated dental model. Based on the aim of the study, three model groups were established: stone models (ST), models produced using a low-viscosity resin (DentaMODEL, Asiga, 400 mPa·s) (AM-A), and models produced using a high-viscosity resin (Keyprint, Keystone, 600–1200 mPa·s) (AM-K). Stone models were obtained by taking impressions of the reference model using splinted open-tray impression copings and a polyether impression material, followed by casting with Type IV dental stone. The 3D-printed models were fabricated according to the manufacturers' instructions using the two resins. These procedures were repeated until eight working models were obtained for each group (n=8). The reference model and the 24 working models were scanned using a high-resolution desktop scanner for analysis. The study measured the model RMS value, RMS values of the SB, 3D and angular deviations of SB, as well as interimplant linear and angular deviations. All analyses were performed to assess trueness and precision. For model alignment and deviation analysis, metrology software (Geomagic Control X, Geomagic Inc., 3D Systems, NC, USA) was used. Statistical analysis of the data was performed using Jamovi software (Version 2.6) through the Kolmogorov-Smirnov (for normality), Levene's test (for homogeneity of variances), one-way ANOVA, and the Games-Howell post-hoc test (α=0.05). According to the RMS of the models, a statistically significant difference was found among all groups (p<0.043), and the trueness ranking was AM-A>AM-K>ST. Regarding precision, the AM-K group showed significantly lower values compared to the AM-A group (p=0.015). Based on the one-way ANOVA analysis of RMS trueness of SB, statistically significant differences were found for SBs 14 and 26 (p<0.013), while significant differences were observed for SBs 14 and 16 in terms of precision (p<0.016). According to the 3D deviation trueness, the AM-K group showed significantly higher values compared to the ST group for SB 14 (p=0.01), and significantly higher results than the other groups for SB 16 (p<0.025). For SB 26, a significant difference was found among all groups (p<0.014), with the trueness ranking being ST>AM-A>AM-K. In terms of precision, the AM-A group showed significantly higher results than the ST group for SB 14 (p=0.021). When angular deviation trueness were evaluated, the ST group showed significantly higher values than the other groups for SB 24 (p<0.007). For SB 26, the AM-A group showed significantly lower values than the AM-K group (p=0.007). Regarding angular deviation precision, only SB 24 showed significantly higher values for the ST group compared to the AM-K group (p=0.040). No significant difference was observed in the trueness for interimplant linear deviation (p>0.147). However, in terms of precision, the AM-K group showed significantly higher values than the ST group for the 14–16 interimplant linear deviation (p=0.033). Regarding the trueness of interimplant angular deviation, the ST group demonstrated significantly lower results than the other groups for the 24–26 inter-implant angular deviation (p<0.002). In terms of precision, only the AM-A group showed significantly lower values than the ST group for the 24–26 inter-implant angular deviation (p=0.020). When the mean deviation results of SBs at the implant and multi-unit abutment levels were compared within each model, a statistically significant difference was found between SBs at the implant and multi-unit abutment levels across all model groups (p<0.018). When comparing the trueness of dental implant models, AM models were found to be superior to stone models. However, in terms of precision, AM models can be considered as an alternative to stone models. In most cases, the positional trueness of implant analogs in the AM-A group was comparable to that of the ST group. However, although not all differences were statistically significant, the consistently highest values observed in the AM-K group across analyses indicate that the viscosity of the resin used is a key factor influencing implant positioning. While models showed similarity in SB precision, significantly higher values were recorded in the AM-K group in certain cases. The use of high-viscosity resin should be considered as a potential factor that increases the possibility of error in repeated model production.