search

zaccharieramzi / fastmri-reproducible-benchmark

Try several methods for MRI reconstruction on the fastmri dataset. Home to the XPDNet, runner-up of the 2020 fastMRI challenge.
https://fastmri.org/leaderboards
MIT License
151 stars 50 forks source link

question about mask #181

Open WuJiaMian opened 1 year ago

WuJiaMian commented 1 year ago

Hello,author! I have a simple question that how to create a Gauss2d mask. Could you share the code of creating a 2dgauss mask? Thanks a lot.

zaccharieramzi commented 1 year ago

Hi @WuJiaMian ,

Thanks for the interest in this repo. So basically to create a Gaussian 2d mask, you should add a function doing exactly that in the util with the masking functions (here), and then use that in the preprocessing step here for single coil (and here for multicoil).

Unfortunately I don't have the code for the Gaussian 2d mask simply because it's not physically relevant. In this repo I tried to stick to the masks that are used in the fastMRI dataset which is really the strongest benchmark in MRI reconstruction.

I would be happy to review a PR with your implementation if you want some validation but of course you can do that in your own fork without including it in this repo.

Cheers