Hello,

I have read the above thread with interest because I am also working through steps to incorporate B0 distortion correction. I am using FSL for the correction and then include the corrected files in AFNI. I wanted to show you my steps to make sure they make sense. I also had a couple questions along the way.

1) Convert Siemens DICOM files to fsl NIFTI using MRIConvert

2) First, prepare the map images:

a. brain extract (bet_gui) and erode (fslmaths [mag_brain] –ero [mag_brain_ero]) the magnitude image and
b. use fsl_prepare_fieldmap to create the fieldmap_radians image (command-line or fsl_prepare_fieldmap_gui).
c. More details can be found here: [fsl.fmrib.ox.ac.uk]

3) Instead of using FSL FEAT for B0 unwarping, use the command-line tool fugue (dwell is the effective EPI echo spacing)
fugue –i [M001_r01.nii] --dwell=[0.00047] --loadfmap=[field_rad_ero.nii.gz] –unwarpdir=y- -u [M001_r01_fmap]

4) Note specific to my epi files: Before running fugue, I first converted the DICOM files to AFNI using to3d because I used a sparse sampling study design and needed to convert the EPI raw files with the specific volume acquisition timing. Then, I converted the file using 3dAFNItoNIFTI. I double checked the headers to make sure the right information was displayed as far as the TR and volume acquisition timing and orientation.

a. The AFNI epi file had the correct the header info.
b. I used fslhd to look at the header of the converted .nii file. The TR was correct and orientation was also RAI. I did not see a way though to check that the sparse sampling volume acquisition timing was included. I can assume it though, right?

5) Then, I used the fieldmap corrected epi files directly in uber_subject.py without any further conversion.

6) uber_subject.py: I wanted to check on a string of warnings that I received during processing (I used non-linear alignment for the MNI template). I am not quite sure how to act on these warnings.

++ 3drefit: AFNI version=AFNI_2011_12_21_1014 (Apr 22 2015) [64-bit]
++ Authored by: RW Cox
++ Processing AFNI dataset __ats_tmp___upad40_pre.anat.un_XYZ_OYmeqfZ3JqF7n9LNC7gDAA.aff+orig
*+ WARNING: Changing the space of an ORIG view dataset may cause confusion!
*+ WARNING: NIFTI copies will be interpreted as TLRC view (not TLRC space).
*+ WARNING: Consider changing the view of the dataset to TLRC view also
++ 3drefit processed 1 datasets

Thanks for having a look at my steps.
Maria