Hi !
We're not having success using epi_reg with fieldmaps, and hope someone can take a look.
Our fieldmap images are the f output from topup (and converted to rad/s units).
Our magnitude images are the mean of the i output from topup (i.e., the corrected spin echo pair).
A) If I run epi_reg without the fieldmap inputs, it works just fine.
B) If I include the fieldmap inputs, the EPI output is terribly distorted.
Example: http://imgur.com/a/ZOCqG
Also, I notice the magnitude image (in some subjects but not all) is not properly registered to the highres (out-FM_fieldmap2str.nii.gz).
But even when the mag is well registered to the highres, the EPI output is badly distorted.
C) If I upsample all my inputs to highres resolution and use the -nofmapreg flag
(to avoid any possibility of a failing magnitude to highres registration), the EPI output still has terrible distortions.
I include here a Dropbox link with 2 subjects (Subj7 and Subj8).
Subj7 has a failed magnitude to highres registration, but Subj8 does not.
Yet regardless, both their EPI outputs look bad, even if I use the -nofmapreg flag (C above) to avoid the registration issue.
A text file with the commands used is included in the zip file.
https://dl.dropboxusercontent.com/u/3889124/EPI_REG.zip
Thanks!
Mark
PS- If I run applytopup directly to the EPI, the results look fine.
|