rixez/pytorch_mri_variationalnetwork

A pytorch implementation for variational network for MRI reconstruction

pytorch_mri_variationalnetwork's Issues

Unable to use data from GLOBUS (.h5 files) as input to code

I have downloaded the data from GLOBUS as mentioned in README. This consists of all .h5 files. Currently, I am working on Single Coil Knee dataset. There seems to be a missing link between what the code expects as input and the data downloaded from GLOBUS.

1. Download single coil knee (train, test and val) data from GLOBUS
2. Change the root_dir to point to "knee folder"
3. run python run_varnet.py on terminal
4. error:

----------------- Options ---------------
activation: rbf
batch_size: 1
epoch: 100
error_scale: 1.0
features_out: 24
gpus: 0
loss_type: complex
loss_weight: 1.0
lr: 0.0001
mode: train
momentum: 0.0
name: coronal_pd_fs
num_act_weights: 31
num_stages: 10
optimizer: adam
root_dir: data/knee
sampling_pattern: cartesian_with_os
save_dir: exp/basic_varnet
----------------- End -------------------
Traceback (most recent call last):
File "run_varnet.py", line 46, in
dataset = KneeDataset(**args)
File "/home/m/Documents/mri_variationalnetwork/pytorch_mri_variationalnetwork-master/data_utils.py", line 64, in _ _ init _ _
self.mask = loadmat(str(self.mask_dir[0]))
IndexError: list index out of range

Any suggestions on how to proceed?

Artifact in the middle

Hi Rixez,

I tried to duplicate your work and I found there is some artifacts in the middle of picture, which has the same shape as sampling mask.
I don't know how to solve it. Thanks for your help.