Skip to main content

Table 5 Multiple comparison of accuracy between CMG Net and the other four nets

From: Utility of a novel integrated deep convolutional neural network for the segmentation of hip joint from computed tomography images in the preoperative planning of total hip arthroplasty

Disease

Dependent variable

(I) NET

(J) NET

Mean difference (I − J)

Std. error

Sig

95% confidence interval

Lower bound

Upper bound

Multiple comparisons

Least significance difference (LSD)

ONFH

DOC

CMG net

FCN

0.14

0.02

< 0.05

0.10

0.19

2D UNET

0.04

0.02

< 0.05

0.00

0.09

2.5D UNET

0.20

0.02

< 0.05

0.16

0.24

3D UNET

0.22

0.02

< 0.05

0.18

0.26

ASD

CMG net

FCN

− 0.03

0.06

0.61

− 0.15

0.09

2D UNET

− 0.15

0.06

< 0.05

− 0.27

− 0.03

2.5D UNET

− 0.04

0.06

0.51

− 0.16

0.08

3D UNET

− 0.09

0.06

0.14

− 0.21

0.03

HD

CMG net

FCN

− 1.28

0.36

< 0.05

− 1.99

− 0.57

2D UNET

− 0.74

0.36

< 0.05

− 1.45

− 0.03

2.5D UNET

− 5.16

0.36

< 0.05

− 5.87

− 4.45

3D UNET

− 6.86

0.36

< 0.05

− 7.56

− 6.15

DDH

DOC

CMG net

FCN

0.15

0.02

< 0.05

0.12

0.18

2D UNET

0.09

0.02

< 0.05

0.06

0.12

2.5D UNET

0.24

0.02

< 0.05

0.21

0.27

3D UNET

0.25

0.02

< 0.05

0.22

0.28

ASD

CMG net

FCN

− 0.19

0.03

< 0.05

− 0.25

− 0.13

2D UNET

− 0.18

0.03

< 0.05

− 0.24

− 0.12

2.5D UNET

− 0.27

0.03

< 0.05

− 0.34

− 0.21

3D UNET

− 0.19

0.03

< 0.05

− 0.26

− 0.13

HD

CMG net

FCN

− 1.08

0.31

< 0.05

− 1.69

− 0.47

2D UNET

− 0.80

0.31

< 0.05

− 1.41

− 0.20

2.5D UNET

− 4.97

0.31

< 0.05

− 5.57

− 4.36

3D UNET

− 6.60

0.31

< 0.05

− 7.21

− 6.00

FNF

DOC

CMG net

FCN

0.11

0.01

< 0.05

0.09

0.14

2D UNET

0.04

0.01

< 0.05

0.01

0.06

2.5D UNET

0.21

0.01

< 0.05

0.18

0.23

3D UNET

0.19

0.01

< 0.05

0.17

0.22

ASD

CMG net

FCN

− 0.21

0.02

< 0.05

− 0.25

− 0.17

2D UNET

− 0.24

0.02

< 0.05

− 0.27

− 0.20

2.5D UNET

− 0.23

0.02

< 0.05

− 0.27

− 0.19

3D UNET

− 0.24

0.02

< 0.05

− 0.28

− 0.20

HD

CMG net

FCN

− 1.16

0.22

< 0.05

− 1.60

− 0.73

2D UNET

0.08

0.22

0.73

− 0.36

0.51

2.5D UNET

− 4.88

0.22

< 0.05

− 5.32

− 4.45

3D UNET

− 6.64

0.22

< 0.05

− 7.08

− 6.21

OA

DOC

CMG net

FCN

0.03

0.02

< 0.05

0.00

0.07

2D UNET

0.09

0.02

< 0.05

0.05

0.12

2.5D UNET

0.16

0.02

< 0.05

0.12

0.19

3D UNET

0.15

0.02

< 0.05

0.12

0.18

ASD

CMG net

FCN

− 0.04

0.07

0.62

− 0.18

0.11

2D UNET

0.11

0.07

0.14

− 0.04

0.25

2.5D UNET

− 0.10

0.07

0.16

− 0.25

0.04

3D UNET

− 0.04

0.07

0.57

− 0.19

0.10

HD

CMG net

FCN

− 2.63

0.22

< 0.05

− 3.07

− 2.18

2D UNET

− 3.19

0.22

< 0.05

− 3.63

− 2.74

2.5D UNET

− 4.72

0.22

< 0.05

− 5.17

− 4.28

3D UNET

− 5.74

0.22

< 0.05

− 6.18

− 5.29

NORMAL

DOC

CMG net

FCN

0.01

0.02

0.53

− 0.02

0.04

2D UNET

0.09

0.02

< 0.05

0.06

0.12

2.5D UNET

0.15

0.02

< 0.05

0.11

0.18

3D UNET

0.16

0.02

< 0.05

0.13

0.20

ASD

CMG net

FCN

− 0.05

0.07

0.53

− 0.19

0.10

2D UNET

− 0.02

0.07

0.81

− 0.16

0.13

2.5D UNET

− 0.09

0.07

0.24

− 0.23

0.06

3D UNET

0.04

0.07

0.60

− 0.11

0.19

HD

CMG net

FCN

− 2.75

0.27

< 0.05

− 3.28

− 2.22

2D UNET

− 3.44

0.27

< 0.05

− 3.98

− 2.91

2.5D UNET

− 4.61

0.27

< 0.05

− 5.14

− 4.08

3D UNET

− 6.00

0.27

< 0.05

− 6.54

− 5.47