Table 1 Overall results with statistical analysis on the TCGA testing sets in (a) classification of aggressive and non-aggressive EC, (b) TMB prediction of the aggressive EC, and (c) TMB prediction of the non-aggressive EC

From: Deep learning for endometrial cancer subtyping and predicting tumor mutational burden from histopathological slides

Task

Method

Acc.

Sens.

Spec.

MeanSS

AUROC (μ ± σ)

Fisher’s test p-value

(a) Classification of aggressive and non-aggressive EC

TR-MAMIL

0.89

0.93

0.84

0.89

0.88 ±0.05

<0.001***

 

ClassicMIL31

0.53

0.53

0.53

0.53

0.71 ± 0.18

0.179

 

CLAM26

0.83

0.89

0.74

0.81

0.85 ± 0.04

<0.001***

 

Wang et al.32

0.73

0.74

0.72

0.73

0.80 ± 0.05

<0.001***

 

Improved_InceptionV3_MS33

0.69

0.90

0.40

0.65

0.80 ± 0.00

<0.001***

 

TOAD27

0.80

0.76

0.86

0.81

0.81 ± 0.05

<0.001***

 

TransMIL28

0.83

0.76

0.91

0.84

0.88 ± 0.05

<0.001***

 

MRAN29

0.86

0.90

0.81

0.86

0.88 ± 0.05

<0.001***

(b) TMB prediction for the aggressive subtype

TR-MAMIL

0.73

0.72

0.73

0.73

0.82± 0.13

<0.001***

 

ClassicMIL31

0.61

0.53

0.65

0.59

0.62 ± 0.08

0.127

 

CLAM26

0.66

0.67

0.66

0.66

0.67 ± 0.03

0.002**

 

Wang et al.32

0.74

0.71

0.76

0.73

0.73 ± 0.02

0.436

 

Improved_InceptionV3_MS33

0.70

0.41

0.86

0.64

0.65 ± 0.05

0.017*

 

TOAD27

0.64

0.39

0.80

0.60

0.6 ± 0.05

0.038*

 

TransMIL28

0.62

0.78

0.52

0.65

0.6 ± 0.03

0.189

 

MRAN29

0.72

0.58

0.81

0.70

0.73 ± 0.02

<0.001***

(c) TMB prediction for the non-aggressive subtype

TR-MAMIL

0.66

0.76

0.61

0.68

0.56 ± 0.08

0.006**

 

ClassicMIL31

0.61

0.43

0.70

0.56

0.51 ± 0.01

0.235

 

CLAM26

0.64

0.41

0.76

0.59

0.52 ± 0.07

0.339

 

Wang et al.32

0.65

0.25

0.84

0.54

0.53 ± 0.04

0.144

 

Improved_InceptionV3_MS33

0.62

0.55

0.65

0.60

0.52 ± 0.03

0.567

 

TOAD27

0.64

0.48

0.72

0.60

0.54 ± 0.03

0.101

 

TransMIL28

0.58

0.41

0.67

0.54

0.52 ± 0.01

0.297

 

MRAN29

0.47

0.62

0.40

0.51

0.57 ± 0.02

0.567

  1. MeanSS: Mean of sensitivity and specificity; *p < 0.05; **p < 0.01; ***p < 0.001.