Research Article
[Retracted] Segmentation Algorithm of Magnetic Resonance Imaging Glioma under Fully Convolutional Densely Connected Convolutional Networks
Table 1
Basic parameters of DenseNet.
| Layers | DenseNet () | Output size |
| Input | / | 256256 | Convolution | 77 conv, stride 2 | 128128 | Pooling | 33 max pool, stride 2 | 6464 | Dense block 1 | | 6464 | Transition layer 1 | 11 conv 22 average pool, stride 2 | 6464 3232 | Dense block 2 | | 3232 | Transition layer 2 | 11 conv 22 average pool, stride 2 | 3232 1616 | Dense block 3 | | 1616 | Transition layer 3 | 11 conv 22 average pool, stride 2 | 1616 88 | Dense block 4 | | 88 |
|
|