Table 6 Comparison of existing and proposed models for the NIHChest X-ray-8 dataset.

From: Medical application driven content based medical image retrieval system for enhanced analysis of X-ray images

Models

Accuracy (%)

Precision (%)

Recall (%)

F1-score (%)

MAE (%)

MSE (%)

RMSE (%)

Bi-LSTM

89.45

90.40

92.46

93.33

4.671

1.6754

1.567

RNN

85.49

91.35

90.49

91.62

2.8658

2.674

1.565

GRU

90.31

93.44

94.41

88.53

1.6854

4.342

2.521

DenseNet 129

97.97

83.56

88.42

95.14

0.4531

3.981

0.5674

MobileNet

91.56

96.43

93.67

95.89

2.6781

0.675

3.5451

Proposed

99.67

97.10

98.43

97.09

0.0986

0.0871

0.0971

Back to article page