http://www.fmrib.ox.ac.uk/fslfaq/#feat_customreg
All the best,
Mark
On 3 Aug 2011, at 10:01, Dar Meshi wrote:
> hello fsl people,
>
> by the way, I attended your montreal course last month...it was extremely helpful, thank you so much! however there was only instruction on how to do full analysis in feat. thus, I'm confused on how to do each step of the analysis separately...especially hoping I can do registration in the flirt gui and then use those registered images in the feat gui for pre-stats, stats and post-stats...any thoughts on the below message?
>
> thanks!
> dar
>
>
> -----------------------------
>
>
> hi,
>
> I'm rather new to fsl and I have a question regarding registration outside of feat. so I've collected imaging data for an fmri study (t1 anatomical, fieldmap, and epi's). my epi's have the top/back part of the brain cutoff due to angling of the field of view for ofc data collection. unfortunately, nothing I try in feat (DOF and full search) can give me proper registration of the epi's to their highres t1 counterpart (within-subject). I believe it's because the part of the precuneus, etc that is missing is actually quite large. basically all the epi to highres registrations are shifted vertically by several millimeters.
>
> with the above in mind:
>
> 1. could you please advise me on how to improve my registration? I was reading through the posts here in the forum and it seems that using flirt to change the cost function to mutual information, and also trying sinc interpolation might help. sound good? any other suggestions?
>
> 2. once I use flirt, how do I then integrate these registered images back into feat? sorry, I'm sure this is very easy, but I'm new here :)
>
> thanks for the help, very much appreciated!
> dar
>
|