Skip to main content

Table 3 Performance difference on average prediction results (over 5 runs) by using GT tumor data and ellipse tumor bounding box data for training, where the standard deviation is included in (·) after each performance value

From: Prediction of glioma-subtypes: comparison of performance on a DL classifier using bounding box areas versus annotated tumors

Case Study

Tumor Area

Av. Acc.(∣σ∣)

Av. Sen.(∣σ∣)

Av. Spec.(∣σ∣)

A

Ellipse

69.86(2.46)

74.20(4.39)

64.60(1.92)

 

GT

72.78(1.45)

76.05(1.63)

68.57(1.78)

Difference

 

2.92(1.45)

1.85(1.78)

3.97(1.63)

B

Ellipse

79.50(2.12)

86.65(3.28)

72.32(1.67)

 

GT

82.73(1.82)

89.70(2.00)

75.45(3.04)

Difference

 

3.23(0.3)

3.05(1.28)

3.13(1.37)