Dear Jesper,

 

  Thank you for your help on this!  I used the FSL gui but then pulled the EF_UD_shift.nii it generated, which I assume is the output of the same command. Looking both at FSL's output to firefox and the shift.nii, it does look like it transitions fairly smoothly and the maxima are where you describe.  Their absolute values are around 8-12, with higher values/larger clusters of high values around OFC.  In addition, there's a very small number of voxels with high vlaues (~13) around inferior parietal lobule that seem to line up with the oddities I'm seeing in the b0 correction.  I've attached the FSL Pre-stats output so you can see what I'm on about, and let me know if this matches what you expect from 'smoothly changing'.  Thanks!

 

  Best,

  Steven

On Thu, May 31, 2012 at 6:23 AM, Jesper Andersson <[log in to unmask]> wrote:

Dear Steven,

> Hello!  I've never done B0 corrections before but this forum has been very helpful to me.  Currently, I'm trying to apply B0 corrections to some EPI data that I have, and my end result has been coming out looking a little funny.  I've been successful in converting the phases into the right range [0-2pi] and the fieldmap generated has looked reasonable (to me at least).  However, when I apply the correction, I get a moderate case of demonbrain (a clump of voxels stretching out of the brain and past skull in a spike).  Looking at the shift map, it does seem as though there's an odd value corresponding to where the brain becomes deformed.  I attempted to use FUGUE's despiking command to potentially fix this, but when I used the output from despiking as the fieldmap there seemed to be more deformation (in addition, the thresholded signal loss weighting images did not seem to show the whole brain anymore: all show a single orientation and the left quarter or so is missing).  Any insight into these problems would be helpful!

What do you mean by "odd value"? You are referring to a shiftmap so I assume you have used the --saveshift argument to obtain this. This map should be slowly changing (smooth) with its maximum absolute values roughly around the orbitofrontal cortex and slightly smaller maxima above the ear canals. Is this what you see, and what are the maximum values in these areas?

Jesper