Skip to main content

Table 2 Prediction accuracy within a given margin

From: Automatic craniomaxillofacial landmarks detection in CT images of individuals with dentomaxillofacial deformities by a two-stage deep learning model

 

Error margin

Prediction accuracy

V-Net without

Region division

2 mm

22.08%

2.5 mm

59.74%

3 mm

87.01%

4 mm

96.10%

3D UX-Net without

Region division

2 mm

35.06%

2.5 mm

61.04%

3 mm

85.71%

4 mm

96.10%

V-Net with

Region division

2 mm

61.04%

2.5 mm

89.61%

3 mm

96.10%

4 mm

98.70%

3D UX-Net with

Region division

2 mm

76.62%

2.5 mm

90.91%

3 mm

93.51%

4 mm

98.70%