Skip to main content

Table 2 Testing results of models using 16 risk factors as diagnostic indicators

From: Diagnosis model of early Pneumocystis jirovecii pneumonia based on convolutional neural network: a comparison with traditional PCR diagnostic method

Models

Accuracy

Recall

Precision

Specificity

Sensitivity

\({F}_{1}\)

AUC

Resnet50

81.30%

0.93

78.30%

0.73

0.93

0.76

0.85

VGG19

83.50%

0.97

79.50%

0.79

0.97

0.81

0.81

Inception-V3

80.60%

0.95

76.40%

0.74

0.95

0.78

0.84

Densenet121

84.50%

0.97

78.00%

0.73

0.97

0.86

0.84

Back to article page

BMC Pulmonary Medicine

ISSN: 1471-2466

Contact us