Skip to main content

Table 3 Comparison with 3D U-Net and classic improved 3D U-Net on the BraTS2020 test set

From: A lightweight hierarchical convolution network for brain tumor segmentation

Methods

Dice (%)

HD95 (mm)

Params (M)

FLOPs (G)

ET

WT

TC

ET

WT

TC

3D U-Net

73.50

89.42

81.92

35.68

6.85

11.54

5.89

148.17

Res 3D U-Net

73.87

89.53

82.23

33.41

6.19

10.23

6.70

187.86

3D U-Net++

73.94

89.35

82.57

32.65

7.30

9.58

6.84

508.46

Attention 3D U-Net

74.42

90.25

82.86

30.24

6.72

9.35

6.47

151.51

LHC-Net (Ours)

76.38

90.01

83.32

30.09

6.96

6.30

1.65

35.58

  1. Bold indicates the best result for each evaluation metric
Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us