| from __future__ import annotations |
|
|
| import torch |
| from torch import nn |
| import torch.nn.functional as F |
|
|
| from .blocks import ConvBlock3d |
|
|
|
|
| class SupportNet(nn.Module): |
| """Compact 3D U-Net style lesion support map network.""" |
|
|
| def __init__(self, in_channels: int = 1, base_channels: int = 24) -> None: |
| super().__init__() |
| self.enc1 = ConvBlock3d(in_channels, base_channels) |
| self.enc2 = ConvBlock3d(base_channels, base_channels * 2, stride=2) |
| self.enc3 = ConvBlock3d(base_channels * 2, base_channels * 4, stride=2) |
| self.mid = ConvBlock3d(base_channels * 4, base_channels * 4) |
| self.dec2 = ConvBlock3d(base_channels * 6, base_channels * 2) |
| self.dec1 = ConvBlock3d(base_channels * 3, base_channels) |
| self.out = nn.Conv3d(base_channels, 1, kernel_size=1) |
|
|
| def forward(self, x: torch.Tensor) -> torch.Tensor: |
| e1 = self.enc1(x) |
| e2 = self.enc2(e1) |
| e3 = self.enc3(e2) |
| z = self.mid(e3) |
| z = F.interpolate(z, size=e2.shape[2:], mode="trilinear", align_corners=False) |
| z = self.dec2(torch.cat([z, e2], dim=1)) |
| z = F.interpolate(z, size=e1.shape[2:], mode="trilinear", align_corners=False) |
| z = self.dec1(torch.cat([z, e1], dim=1)) |
| return self.out(z) |
|
|
|
|
