Skip to main content

Table 1 Classification results via cross-validation

From: Multimodal microscopy for automated histologic analysis of prostate cancer

Data set

Feature Extraction

AUC

Sensitivity (%)

Specificity (%)

Mf

  

AVG

STD

 

AVG

STD

 

Data1

IR & HE

0.982

0.0030

90

94.76

1.64

13

    

95

90.91

1.62

 
    

99

77.80

5.52

 
 

HE only

0.968

0.0052

90

91.64

2.26

11

    

95

83.90

1.91

 
    

99

53.43

13.65

 

Data2

IR & HE

0.974

0.0145

90

92.53

7.11

7

    

95

84.19

10.84

 
    

99

49.54

22.51

 
 

HE only

0.880

0.0175

90

61.34

10.31

8

    

95

22.21

10.06

 
    

99

11.21

6.01

 
  1. AVG and STD denote average and standard deviation across ten repeats of cross-valdiation. Mf is the median size of the feature set obtained by feature selection from training data. Column "Feature Extraction" indicates if features were obtained using H&E as well as IR data, or with H&E data alone. The parameter γ of a radial basis kernel for SVM is set to 1.