Hi sorry to be a bother,
but I've compared both raw and processed fieldmaps of subjects that DID NOT run with eddy_cuda and fieldmaps with subjects that DID run with the eddy_cuda and everything looks comparable, including the values inside the brain for the HZ converted field-maps. This also does not appear to be an issue with a handful of subjects but with about 25% of my sample (n=1380). I also compared the json files and again everything looks OK.
If it is helpful, for some failed subjects (but not all) it will output the following files only:
*_eddy_corr.eddy_post_eddy_shell_alignment_parameters
*_eddy_corr.eddy_post_eddy_shell_PE_translation_parameters
The values in the PE translation files look comparable to successful subjects (actually it's often less).
Thanks again,
Amanda
|