Skip to main content

Table 4 Comparison of CNN-IKOA’s outcomes under various threshold levels

From: CNN-IKOA: convolutional neural network with improved Kepler optimization algorithm for image segmentation: experimental validation and numerical exploration

T

Accuracy

Specificity

Precision (M)

Precision (W)

Recall (M)

Recall (W)

F1-score (M)

F1-score (W)

SI

93.26

95.76

93.40

93.40

93.20

93.20

93.40

93.20

T-3

92.91

95.42

93.40

93.20

92.80

93.00

93.00

93.00

T-4

94.56

96.04

94.80

94.80

94.60

94.80

94.80

94.40

T-5

94.83

94.51

96.47

94.80

94.60

94.60

94.60

94.80

T-7

94.23

93.63

95.75

94.00

93.80

93.80

93.80

93.80

T-8

94.32

96.25

94.40

94.40

94.40

94.20

94.40

94.20

T-10

94.44

96.28

94.80

94.60

94.80

94.60

94.80

94.60

T-12

94.88

96.57

95.60

95.40

95.80

95.40

95.60

95.40

T-15

94.94

96.54

95.60

95.40

95.60

95.60

95.40

95.20