Influence of Intraoral Scanners, Operators, and Data Processing on Dimensional Accuracy of Dental Casts for Unsupervised Clinical Machine Learning: An In Vitro Comparative Study
Table 1
3-Factor ANOVA to evaluate the relationship between scanners, operators, and processing workflow and subsequent Hausdorff’s distance (mm).
List of independent factors
3D scanner
F = 0.94,
IOS1 = mean ± SD = 0.16 ± 0.22
IOS2 = mean ± SD = 0.21 ± 0.48
Augmentation
F = 0.61,
Native dataset = mean ± SD = 0.21 ± 0.47
Augmented dataset = mean ± SD = 0.16 ± 0.25
Clinical operator
F = 0.14,
Operator 1 = mean ± SD = 0.19 ± 0.41
Operator 2 = mean ± SD = 0.17 ± 0.34
Native dataset
Operator 1 (mean ± SD)
Operator 2 (mean ± SD)
IOS1
0.14 ± 0.21
0.18 ± 0.24
IOS2
0.28 ± 0.71
0.22 ± 0.52
Augmented dataset
Operator 1 (mean ± SD)
Operator 2 (mean ± SD)
IOS1
0.17 ± 0.22
0.14 ± 0.23
IOS2
0.19 ± 0.28
0.15 ± 0.28
Interaction effect
(1) 1. 3D scanner vs. augmentation: F = 0.45, (2) 2. 3D scanner vs. clinical operator: F = 0.21, (3) 3. Augmentation vs. clinical operator: F = 0.03, (4) 4. 3D scanner vs. augmentation vs. clinical operator: F = 0.18,
