The HCP pipeline paper talks about performing gradient nonlinearity correction using "a customized version of the gradient_nonlin_unwarp package available in FreeSurfer". The HCP Q2 reference manual directs you to a 3rd party python script containing gradient_unwarp.py ( https://github.com/ksubramz/gradunwarp/blob/master/Readme.md).
I'm trying both of these on some raw data from the HCP magnet, but they don't completely agree and both seem to have some unexpected behavior. Specifically what looks like some half-voxel offset and/or dimensions off by 1 (The outputs are the correct dimensions but the images look shifted) Which of these was actually used for the HCP data? If it is custom, is the customization easy to implement (or share?). Thanks, Keith Jamison _______________________________________________ HCP-Users mailing list [email protected] http://lists.humanconnectome.org/mailman/listinfo/hcp-users
