From: Convolutional neural networks for detection of transthyretin amyloidosis in 2D scintigraphy images
AUC | Accuracy | Negative patients | Positive patients* | |||
---|---|---|---|---|---|---|
Precision | Recall | Precision | Recall | |||
Linear | 0.88 | 0.97 | 0.99 | 0.98 | 0.59 | 0.79 |
Residual | 0,89 | 0,97 | 0,99 | 0,98 | 0,58 | 0,81 |
VGG16 | 0.77 | 0.98 | 0.98 | 1.00 | 0.81 | 0.54 |
ResNet50 | 0.85 | 0.80 | 1.00 | 0.80 | 0.14 | 0.90 |
InceptionV3 | 0.78 | 0.95 | 0.98 | 0.97 | 0.39 | 0.60 |
MobileNet | 0.76 | 0.95 | 0.98 | 0.96 | 0.35 | 0.56 |