Skip to main content

Table 1 Characteristics and data collection parameters for the study population

From: Deep semi-supervised learning for automatic segmentation of inferior alveolar nerve using a convolutional neural network

Characteristic

Training and tuning

(A)

(N = 83)

Internal validation

(A)

(N = 15)

External validation

(B)

(N = 20)

External validation

(C)

(N = 20)

Age (in years)

59.9 ± 17.2

63.1 ± 16.9

40.0 ± 19.7

43 ± 18.6

Male

44

8

10

10

Female

39

7

10

10

Tube voltage (kV)

120

120

90

90

Tube current (mA)

5

5

4

4

Scan time (s)

16.8

16.8

14.3

15

Voxel size (mm)

0.3

0.3

0.3

0.08 ~ 0.25

FOV (mm2)

230 × 170

230 × 170

170 × 135

150 × 150

Focal spot (mm)

0.58

0.58

0.70

0.50

  1. Internal dataset: Korea University Anam Hospital (A); External dataset: Korea University Ansan Hospital (B) and Korea University Guro Hospital (C)
  2. FOV, field-of-view