Skip to main content

Table 3 Classification performance of different slices

From: Classification of microcalcification clusters in digital breast tomosynthesis using ensemble convolutional neural network

Models

N

AUC

ACC (%)

SEN (%)

SPEC (%)

Precision (%)

Recall (%)

F1 (%)

3D-ResNet-Anisotropic

4

0.8004

76.00

84.38

61.11

79.41

84.38

81.82

6

0.8264

72.00

71.88

72.22

82.14

71.88

76.67

8

0.8455

76.00

75.00

77.78

85.71

75.00

80.00

10

0.8229

74.00

84.38

55.56

77.14

84.38

80.60

12

0.7934

70.00

75.00

61.11

77.42

75.00

76.19