Topup correction

[MartinPerez]

Implemented topup for an image and for all images in subject. Attending to issue #204

[MartinPerez]

@dohmatob Its ready but I have some doubs about subject_data.func, I am not sure if you have a method to get all the imgs as a list.

[mrahim]

is it okay to create this directory ?

[dohmatob]

On Fri, May 27, 2016 at 6:04 PM, mrahim [email protected] wrote:

In pypreprocess/nipype_preproc_fsl_utils.py
#229 (comment):

•    # Apply topup correction to images
  

•    fieldcoef = correction.outputs.out_fieldcoef
  

•    movpar = correction.outputs.out_movpar
  

•    applytopup = memory.cache(fsl.ApplyTOPUP)
  

•    return [applytopup(in_files=img, encoding_file=acq_param_file,
  

•                       in_index=[2], in_topup_fieldcoef=fieldcoef,
  

•                       in_topup_movpar=movpar, output_type='NIFTI',
  

•                       method='jac') for img in func_imgs]
  
• else:

•    # Merge AP and PA images
  

•    appa_merged = fsl.Merge(in_files=[ap_realig_img, pa_realig_img],
  

•                            dimension='t').run()
  

•    # Compute topup transformation
  

•    acq_param_file = os.path.join('/tmp', 'pypreprocess_topup',
  

is it okay to create this directory ?

No. We should store the parameters in the subject's output directory
(subject_data.output_dir).

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
https://github.com/neurospin/pypreprocess/pull/229/files/d9eba52dc19c47e59073e19f282f4d83ce9d3631#r64928000,
or mute the thread
https://github.com/notifications/unsubscribe/AAms1BP-syhHiAUENWCPtrxJxOImL1oiks5qFxYRgaJpZM4IooTp
.

Elvis Dohmatob,
sent from my potato...
http://dohmatob.blogspot.fr/, https://team.inria.fr/parietal/elvis/,
https://github.com/dohmatob

[bthirion]

I understand that adding a test is hard, but having at least an example to check whether the code runs is very important to me. The first difficulty is to find some public data for that. We can provide some local data (sbref images to keep the whole thing light and fast). Then, formally write an example that runs topup and displays the difference. Yes, it's a lot of work !

[MartinPerez]

of course I agree, I could lend one of my datasets which we plan to make public in openfmri with bids. On monday I will make the necessary files available in the server.

[dohmatob]

Any updates ?

[MartinPerez]

My lab datasets are still not public, I guess when they are they can serve as an example, otherwise I dont know of a public dataset that makes this correction.