Table 3 Comparison with state-of-the-art methods on the CVC-ClinicDB and Kvasir.

From: Multi-perspective dynamic consistency learning for semi-supervised medical image segmentation

Method

#Scans

Metrics

Labeled

Unlabeled

Dice (%)↑

Jaccard (%)↑

95HD (voxel)↓

ASD (voxel)↓

SupOnly

102 (10%)

913 (90%)

69.70

59.28

63.60

15.77

URPC18

 

69.55

59.75

61.23

14.88

MCNet22

70.48

61.05

61.47

13.38

SSNet50

73.27

62.97

63.07

16.42

DCNet23

78.08

69.79

40.68

7.37

MPDC(Ours)

79.60

71.92

30.54

2.91