I'm trying to correct EPI distortion with fieldmaps acquired immediately after the EPI run. The scans were acquired on a Siemens Verio 3T which gave me subtracted magnitude and phase maps. I used spm8 to create the vdm files with "brain mask" option on (otherwise all default options). The correction looked good in general (see attached image), especially the frontal distortion correction. However, it appears to produce an additional area at the bottom part of the brain (red arrow in the image) that does not seem to match any structure and was not there in the uncorrected image. Two out of four subjects' correction that I have done so far showed more or less similar pattern. Is this an over correction? Are there any parameters I should adjust to get rid of it? If not, will this be a problem for data analysis?
Your help is greatly appreciated.