# coding=utf-8
__author__ = "Dimitrios Karkalousos"
from abc import ABC
import torch
from omegaconf import DictConfig, OmegaConf
from pytorch_lightning import Trainer
from torch.nn import L1Loss
from mridc.collections.common.losses.ssim import SSIMLoss
from mridc.collections.common.parts.fft import ifft2
from mridc.collections.common.parts.utils import coil_combination
from mridc.collections.reconstruction.models.base import BaseMRIReconstructionModel
from mridc.collections.reconstruction.models.cascadenet.ccnn_block import CascadeNetBlock
from mridc.collections.reconstruction.models.conv.conv2d import Conv2d
from mridc.collections.reconstruction.parts.utils import center_crop_to_smallest
from mridc.core.classes.common import typecheck
__all__ = ["CascadeNet"]
[docs]class CascadeNet(BaseMRIReconstructionModel, ABC):
"""
Implementation of the Deep Cascade of Convolutional Neural Networks, as presented in Schlemper, J., \
Caballero, J., Hajnal, J. V., Price, A., & Rueckert, D.
References
----------
..
Schlemper, J., Caballero, J., Hajnal, J. V., Price, A., & Rueckert, D., A Deep Cascade of Convolutional \
Neural Networks for MR Image Reconstruction. Information Processing in Medical Imaging (IPMI), 2017. \
Available at: https://arxiv.org/pdf/1703.00555.pdf
"""
def __init__(self, cfg: DictConfig, trainer: Trainer = None):
# init superclass
super().__init__(cfg=cfg, trainer=trainer)
cfg_dict = OmegaConf.to_container(cfg, resolve=True)
self.coil_combination_method = cfg_dict.get("coil_combination_method")
self.fft_centered = cfg_dict.get("fft_centered")
self.fft_normalization = cfg_dict.get("fft_normalization")
self.spatial_dims = cfg_dict.get("spatial_dims")
self.coil_dim = cfg_dict.get("coil_dim")
# Cascades of CascadeCNN blocks
self.cascades = torch.nn.ModuleList(
[
CascadeNetBlock(
Conv2d(
in_channels=2,
out_channels=2,
hidden_channels=cfg_dict.get("hidden_channels"),
n_convs=cfg_dict.get("n_convs"),
batchnorm=cfg_dict.get("batchnorm"),
),
fft_centered=self.fft_centered,
fft_normalization=self.fft_normalization,
spatial_dims=self.spatial_dims,
coil_dim=self.coil_dim,
no_dc=cfg_dict.get("no_dc"),
)
for _ in range(cfg_dict.get("num_cascades"))
]
)
self.coil_combination_method = cfg_dict.get("coil_combination_method")
# initialize weights if not using pretrained ccnn
# TODO if not cfg_dict.get("pretrained", False)
self.train_loss_fn = SSIMLoss() if cfg_dict.get("train_loss_fn") == "ssim" else L1Loss()
self.eval_loss_fn = SSIMLoss() if cfg_dict.get("eval_loss_fn") == "ssim" else L1Loss()
self.accumulate_estimates = False
self.dc_weight = torch.nn.Parameter(torch.ones(1))
[docs] @typecheck()
def forward(
self,
y: torch.Tensor,
sensitivity_maps: torch.Tensor,
mask: torch.Tensor,
init_pred: torch.Tensor,
target: torch.Tensor,
) -> torch.Tensor:
"""
Forward pass of the network.
Parameters
----------
y: Subsampled k-space data.
torch.Tensor, shape [batch_size, n_coils, n_x, n_y, 2]
sensitivity_maps: Coil sensitivity maps.
torch.Tensor, shape [batch_size, n_coils, n_x, n_y, 2]
mask: Sampling mask.
torch.Tensor, shape [1, 1, n_x, n_y, 1]
init_pred: Initial prediction.
torch.Tensor, shape [batch_size, n_x, n_y, 2]
target: Target data to compute the loss.
torch.Tensor, shape [batch_size, n_x, n_y, 2]
Returns
-------
pred: list of torch.Tensor, shape [batch_size, n_x, n_y, 2], or torch.Tensor, shape [batch_size, n_x, n_y, 2]
If self.accumulate_loss is True, returns a list of all intermediate estimates.
If False, returns the final estimate.
"""
pred = y.clone()
for cascade in self.cascades:
pred = cascade(pred, y, sensitivity_maps, mask)
pred = torch.view_as_complex(
coil_combination(
ifft2(
pred,
centered=self.fft_centered,
normalization=self.fft_normalization,
spatial_dims=self.spatial_dims,
),
sensitivity_maps,
method=self.coil_combination_method,
dim=self.coil_dim,
)
)
_, pred = center_crop_to_smallest(target, pred)
return pred