Table 4 The evaluation metrics of the Deep learning system on the internal and external datasets.

From: A deep learning-based method for assessing tricuspid regurgitation using continuous wave Doppler spectra

TR severity

Internal dataset

External dataset

Sensitivity (%)

Precision (%)

F1 Score (%)

AUC

Sensitivity (%)

Precision (%)

F1 score (%)

AUC

Mild

87.80

97.12

92.23

0.88

88.00

91.67

89.80

0.86

Moderate

86.80

84.77

85.77

0.84

85.54

86.94

86.23

0.79

Severe

92.80

86.57

89.58

0.89

91.46

87.50

89.44

0.87

Macro average

89.13

89.49

89.19

0.87

88.33

88.70

88.49

0.84

Back to article page