Table 2 Comparison results of the HyborNet method with current popular methods in the 3DIRCADb dataset.

From: Hybrid gabor attention convolution and transformer interaction network with hierarchical monitoring mechanism for liver and tumor segmentation

Methods

Liver

Tumor

DIC

IOU

Recall

RAVD

ASSD

MSD

DIC

IOU

Recall

RAVD

ASSD

MSD

DeepLabv3+

90.07

89.89

91.13

0.014

2.58

12.46

55.44

36.73

45.34

0.014

10.76

35.76

U-Net

94.27

90.16

92.25

0.015

1.89

10.23

56.23

40.63

48.53

0.369

7.66

30.66

Attention U-Net

94.96

90.02

92.23

0.011

1.95

9.15

57.18

40.77

49.56

0.246

9.08

29.08

ResU-Net

94.47

89.84

92.35

0.012

1.84

9.42

57.46

41.66

50.14

0.126

7.06

35.06

U-Net++

95.09

90.43

93.42

0.016

1.62

8.25

58.02

41.28

50.25

0.242

7.69

32.69

Double UNet

95.75

90.33

93.74

0.004

1.66

7.59

56.83

41.94

51.03

0.240

11.51

41.51

nnformer

95.53

90.41

94.12

0.004

1.65

7.32

57.74

42.16

51.12

0.157

11.21

35.42

Swin-UNet

95.27

90.55

94.35

0.005

1.69

7.74

58.91

42.54

52.34

0.009

12.45

32.45

TransUNet

95.43

90.63

94.43

0.004

1.81

8.14

59.26

43.66

52.56

0.104

13.42

33.42

HiFormer

95.95

91.97

94.62

0.007

1.77

8.15

59.55

44.03

53.62

0.118

9.01

29.01

HyborNet

96.86

92.12

95.14

0.006

1.46

6.19

59.85

44.27

53.74

0.015

5.65

25.65

Back to article page