IADR Abstract Archives
Accuracy Assessment of Three Superimposition Techniques in Dental Model Alignment
Objectives: To compare the accuracy of 3D dental models superimposition using automated and manual point-based alignment techniques.
Methods: Twelve maxillary and twelve mandibular 3D dental models were generated by an intraoral scanner (3Shape Trios 3, Copenhagen, Denmark). Each model were then aligned with corresponding reference scan data generated by a laboratory scanner (3Shape E3, Copenhagen, Denmark) using three techniques: (i) automated alignment using Geomagic Wrap 2021 software, (ii) manual four-point based alignment and (iii) manual eight-point based alignment using Cloudcompare software. Friedman test was used to compare the absolute mean deviation of maxillary and mandibular 3D dental models respectively superimposed using three alignment techniques with a statistically significant value set at 0.05. Intra-examiner agreement for three techniques was calculated using intraclass correlation coefficient (ICC).
Results: Twelve pairs of maxillary and twelve pairs mandibular scan data each were superimposed using automated, manual four-point and manual eight-point based alignment techniques respectively. Three alignment techniques presented a significantly different outcome in maxillary dental models with higher mean deviation found in manual four-point based alignment than the automated and manual eight-point techniques (p=0.002). However, no significant difference between groups was observed when aligning mandibular dental models with all the three aforementioned techniques. In addition, intra-examiner agreement was excellent for automated alignment technique (ICC=1) and moderately good for both the four-point based alignment (ICC=0.738) and eight-point based alignment (ICC=0.604) techniques.
Conclusions: Automated and manual 8-point based alignment techniques yields higher accuracy than manual 4-point based alignment techniques in approximating dental models with significant results shown in maxillary dental models.
Methods: Twelve maxillary and twelve mandibular 3D dental models were generated by an intraoral scanner (3Shape Trios 3, Copenhagen, Denmark). Each model were then aligned with corresponding reference scan data generated by a laboratory scanner (3Shape E3, Copenhagen, Denmark) using three techniques: (i) automated alignment using Geomagic Wrap 2021 software, (ii) manual four-point based alignment and (iii) manual eight-point based alignment using Cloudcompare software. Friedman test was used to compare the absolute mean deviation of maxillary and mandibular 3D dental models respectively superimposed using three alignment techniques with a statistically significant value set at 0.05. Intra-examiner agreement for three techniques was calculated using intraclass correlation coefficient (ICC).
Results: Twelve pairs of maxillary and twelve pairs mandibular scan data each were superimposed using automated, manual four-point and manual eight-point based alignment techniques respectively. Three alignment techniques presented a significantly different outcome in maxillary dental models with higher mean deviation found in manual four-point based alignment than the automated and manual eight-point techniques (p=0.002). However, no significant difference between groups was observed when aligning mandibular dental models with all the three aforementioned techniques. In addition, intra-examiner agreement was excellent for automated alignment technique (ICC=1) and moderately good for both the four-point based alignment (ICC=0.738) and eight-point based alignment (ICC=0.604) techniques.
Conclusions: Automated and manual 8-point based alignment techniques yields higher accuracy than manual 4-point based alignment techniques in approximating dental models with significant results shown in maxillary dental models.