Table 2.
Overview of GAN-based methods.
| Ref. | Approach | Findings | Metrics |
|---|---|---|---|
| [68] | Capsule network GAN+LeNet | Prostate image classification | Accuracy 89.20% |
| [69] | GAN+AlexNet | Parkinson's disease | Accuracy 89.23% |
| [70] | GAN+DenseNet121 | Skin lesion classification | Accuracy 94.25% |
| [71] | GAN+InceptionV3 | Breast mass classification | Accuracy 90.41% |
| [72] | GAN+ResNet50 | Brain tumor classification | Accuracy 96.25% |
| [73] | 3D U-Net, VGG16 | Brain tumor segmentation | DSC 90.1% |
| [74] | U-Net, fully connected CNN | Breast tumor segmentation | DSC 88.41% |
| [75] | DeepLapV2 [76], FCN | Left ventricle segmentation | DSC 88.0% |
| [77] | U-Net, FCN | Whole heart segmentation | DSC 86.32% |
| [78] | Autoencoder, CNN | Lung lesion segmentation | DSC 62.0% |