I am using non-linear registration in a first-level FEAT analysis, 10mm warp resolution.
For one of my subjects, the registration to standard space has distorted the brain, as is seen in the highres2standard image.
Outside of FEAT, I have tried running FNIRT for the problem subject and selected the local_linear mode for the intensity modulation parameter (intmod). Although this improved the registration slightly, it is still distorted. Since there seem to be many different parameters that I can play with, and I don't really know what they do, I though it best to get some advice before proceeding.
I would like to know if anyone has suggestions on what FNIRT parameters I might modify to resolve this problem. I have uploaded the native space highres, highres2standard (MNI152 space) and linear .mat file for this subject (reference number: 331729)
Thank you in advance for any suggestions,