Document qMR data organization

[tanguyduval]

@alexfoias plz add @beliveauP to this repo I created this matlab function: https://github.com/neuropoly/agilent_7t/blob/master/data_conversion_auto/organizeqmr_demo.m https://github.com/neuropoly/agilent_7t/blob/master/data_conversion_auto/organizeqmr.m

The idea is to automatically sort data based on this variable:

% FORMAT: {'FOLDER',{'SUBFOLDER',{'keyword','value','param','value/rename'},'SUBFOLDER',{'keyword','value','param','value'}},{'keyword','value','param','value'},...}
copytype = {'qMT',{'SIRFSE',{'keyword','*sirfse*'   ,'param',{'ti'},'Gibbs',true},...
                     'SPGR',{'keyword','qMT_*'      ,'param',{'mtfrq','flipmt/mtflip','tr','te'}},...
                  },...
      'fieldsmap',    {'B0',{'keyword','b0_mapping*','param',{'te/te1','te2'}},...
                       'B1',{'keyword','*b1map_MFA*','param',{'flip1/flipangle'}}
                      },...
             'T1',          {'keyword','*t1map*'    ,'param',{'ti'}},...
  'ProtonDensity',          {'keyword','*MTV*'      ,'param',{'flip1/flipangle','tr','te'}},...
             'T2',          {'keyword','*MWF*'      ,'param',{'flip1/flipangle','tr','te','ne/nechos','nt/averages'}},...
           };

Why I did that:

• same folder organization and names (as opposed to data on osf that are very messy)
• It was very complex to get an acquisition parameter. Now just add to 'param' the parameter you need
• Post processing was done manually. Now we can add steps to preprocess the data. I added Gibbs correction by default.
• we have no clean folder to share our data (because many post-processing files were added by Tung). Now we can create clean qMR datasets easily.

@beliveauP, that would be great if you could try it and document the wiki for future users. Shall we put the infos on neuro.poly wiki or here https://github.com/neuropoly/agilent_7t/wiki ??

[tanguyduval]

Here is how the folder looks like after organization:

And before organization:

[tanguyduval]

We often acquire bad data that we don't need. For each folder, an interface let us preview 10 volumes from each file and we can pick the files we want to merge:

[jcohenadad]

awesome contribution! thank you tanguy

[tanguyduval]

TODO:

• Add option to export Phase or not
• Add option to perform motion correction
• Add option to run qMRLab

[alexfoias]

@tanguyduval I added @BeliveauP to this repository

[BeliveauP]

@tanguyduval Thanks for the script, it will be useful!! I added a GUI to it. For the moment it manages the folder structure and calls your script for file processing. Although not thoroughly completed, could you try it out? I also modified your script, could you try this on a valid dataset? I can complete the documentation.

[tanguyduval]

@BeliveauP I merged our modifications. Cool idea the GUI! Technical thing: I saw that you duplicated the script in the folder Conversion_files. It's easier to merge if you do not duplicate the scripts, Git it here to manage multiple versions!