Research Article
1D CNN-Based Intracranial Aneurysms Detection in 3D TOF-MRA
Table 2
The architectures of 1D CNN discrimination Model 1.
| Layers | Feature vector | Kernel size | Stride | Output size |
| Input | 1 | — | — | 576 | Convolution | 16 | 5 | 1 | 576 | Max pooling | 16 | 2 | 2 | 288 | Convolution | 32 | 5 | 1 | 288 | Max pooling | 32 | 2 | 2 | 144 | Convolution | 64 | 5 | 1 | 144 | Max pooling | 64 | 2 | 2 | 72 | Convolution | 64 | 5 | 1 | 72 | Max pooling | 64 | 2 | 2 | 36 | Fully connected | 2304 | — | — | 2304 | Fully connected | 16 | — | — | 16 | Output | 1 | — | — | 1 |
|
|