Table 1 Statistical analysis representation for precision, recall, and F1-score of Normal retina and different AMD grades for all of the training models using SGD optimizer.

From: Scale-adaptive model for detection and grading of age-related macular degeneration from color retinal fundus images

 

Normal

Intermediate

GA

Wet

SA + ResNet50

 Precision

0.97

0.94

1.00

0.94

 Recall

0.97

0.97

1.00

0.91

 F1-Score

0.97

0.95

1.00

0.92

ResNet50

 Precision

0.93

0.78

0.94

0.8

 Recall

0.84

0.9

0.94

0.75

 F1-Score

0.89

0.84

0.94

0.77

SA + VGG16

 Precision

0.93

0.89

0.79

0.97

 Recall

0.93

0.94

1.00

0.88

 F1-Score

0.93

0.91

0.98

0.92

VGG16

 Precision

0.96

0.88

0.89

0.88

 Recall

0.9

0.88

0.97

0.85

 F1-Score

0.93

0.88

0.93

0.86

SA + Inception

 Precision

0.91

0.89

1.00

0.97

 Recall

0.97

0.97

0.97

0.85

 F1-Score

0.97

0.93

0.98

0.9

Inception

 Precision

0.85

0.88

0.97

0.93

 Recall

0.93

0.91

0.97

0.82

 F1-Score

0.89

0.9

0.97

0.87

SA + ResNet101

 Precision

0.88

0.79

0.97

0.96

 Recall

0.93

0.91

0.97

0.76

 F1-Score

0.90

0.85

0.97

0.85

ResNet101

 Precision

0.93

0.87

0.97

0.77

 Recall

0.87

0.79

0.94

0.91

 F1-Score

0.90

0.83

0.95

0.83

SA + VGG19

 Precision

0.85

0.78

0.89

0.91

 Recall

0.93

0.85

1.00

0.64

 F1-Score

0.89

0.81

0.94

0.75

VGG19

 Precision

0.97

0.91

0.97

0.97

 Recall

0.93

0.94

1.00

0.94

 F1-Score

0.95

0.93

0.98

0.95

SA + ResNet18

 Precision

0.81

0.73

1.00

0.68

 Recall

0.87

0.73

0.91

0.70

 F1-Score

0.84

0.73

0.95

0.69

ResNet18

 Precision

0.74

0.66

0.93

0.59

 Recall

0.83

0.64

0.81

0.61

 F1-Score

0.78

0.65

0.87

0.60