Table 6 Performance measures of ViT for each fold.

From: Bladder lesion detection using EfficientNet and hybrid attention transformer through attention transformation

Folds

Kappa (%)

Recall (%)

Precision (%)

F1-score (%)

Accuracy (%)

Fold1

92.17

93.90

92.60

93.25

94.67

Fold2

92.29

94.23

92.87

93.55

95.08

Fold3

92.50

94.46

93.09

93.77

95.33

Fold4

92.93

95.70

93.34

94.51

95.54

Fold5

92.96

96.35

93.64

94.98

95.62

Back to article page