search

facebookresearch / fastMRI

A large-scale dataset of both raw MRI measurements and clinical MRI images.
https://fastmri.org
MIT License
1.3k stars 372 forks source link

the reconstruction result of jupyter notebook #164

Closed NayeeC closed 3 years ago

NayeeC commented 3 years ago

I have tried the fastMRI_tutorial.ipynb, using the multicoil_knee data. However, the results of the full sample and the subsample look the same. Why does this happen? 8}9(A~YZ40J}1ZKURVTAH10 )Q{JI0OJ5L0 A5EE$PH524G

mmuckley commented 3 years ago

Hello @NayeeC, I can't reproduce this. Have you tried restarting the notebook from the beginning to make sure that you've wiped any other variables you have? Also, I'm not sure that images looks like file1000167.h5.

NayeeC commented 3 years ago

Hello @mmuckley, I use the file1000082_v2.h5, and I think everything is ok. The dataset I donwload is knee_singlecoil_test_v2, and there are no file named file1000167.h5.

mmuckley commented 3 years ago

@NayeeC, in that case the result makes sense. The test data has already been subsampled. If you want to have fully-sampled data, you need to use the training or validation splits.

NayeeC commented 3 years ago

@mmuckley, that's it!! Thank you. By the way, is this code can do the same thing with the brain data? Someone told me that this jupyter notebook is for knee data especially.

mmuckley commented 3 years ago

I think it is untested on brain data, but the only issue I can think of is if there are any hardcoded lines for the coil counts. I think you can use it on brain data.

mmuckley commented 3 years ago

Closing for now as this seems resolved.