Table 3 Comparison of experimental results on ISIC17 and ISIC18 datasets.

From: VMKLA-UNet: vision Mamba with KAN linear attention U-Net

Dataset

Model

mIoU (%)

DSC (%)

Spe (%)

Sen (%)

ISIC17

UNet1

76.98

86.99

97.43

86.82

UNet++30

75.44

86.00

97.34

85.40

UTNetV25

77.35

87.23

98.05

84.85

TransFuse27

79.21

88.40

97.98

87.14

MALUNet28

78.78

88.13

98.47

84.78

UNetV229

82.18

90.22

98.40

88.71

VM-UNet5

80.23

89.03

97.58

89.90

Ours

84.51 ± 0.132

91.60 ± 0.092

98.13 ± 0.039

93.24 ± 0.114

ISIC18

UNet1

77.86

87.55

96.69

85.86

UNet++30

78.31

87.83

95.75

88.65

Attn-UNet9

78.43

87.91

96.23

87.60

UTNetV25

78.97

88.25

96.48

87.60

SANet31

79.52

88.59

95.97

89.46

TransFuse27

80.63

89.27

95.74

91.28

MALUNet28

80.25

89.04

96.19

89.74

UNetV229

80.71

89.32

96.94

88.34

VM-UNet5

81.35

89.71

96.13

91.12

Ours

84.16 ± 0.121

91.40 ± 0.072

97.56 ± 0.027

91.26 ± 0.113

  1. Significant values are in bold.