An Improved Fuzzy Connectedness Method for Automatic Three-Dimensional Liver Vessel Segmentation in CT Images

Table 2

Segmentation performance of the improved fuzzy connectedness (with one single seed), traditional fuzzy connectedness, region growing, and threshold level set (with multiple potential seeds).

Improved fuzzy connectedness

Fuzzy connectedness

Region growing

Threshold level set

ACC (%)

SEN (%)

SPE (%)

DICE (%)

ACC (%)

SEN (%)

SPE (%)

DICE (%)

ACC (%)

SEN (%)

SPE (%)

DICE (%)

ACC (%)

SEN (%)

SPE (%)

DICE (%)

3Dircadb ()

96.4 ± 1.1

73.7 ± 7.6

97.4 ± 1.3

67.3 ± 5.7

82.2 ± 21.9

67.1 ± 25.5

82.5 ± 24.7

40.4 ± 16.2

84.1 ± 21.1

66.1 ± 25.0

86.4 ± 23.4

42.4 ± 15.6

94.3 ± 4.5

51.5 ± 20.3

97.0 ± 5.2

54.0 ± 13.9

Sliver07 ()

96.8 ± 0.6

89.1 ± 6.8

97.6 ± 1.1

71.4 ± 7.6

82.9 ± 25.5

85.0 ± 17.2

83.1 ± 26.7

45.1 ± 18.8

90.6 ± 12.6

84.1 ± 19.2

90.7 ± 13.6

56.2 ± 19.4

95.3 ± 2.9

82.5 ± 6.9

95.3 ± 3.2

67.5 ± 6.7

ACC = accuracy; SEN = sensitivity; SPE = specificity; DICE = Dice coefficient.