> Hey FSL Team,
> I have did several Diffusion measurements on one participant on two days.
> I have 2 different questions
> i) Should I combine all sessions via fslmerge and do then topup/eddy?
> ii) Or should I do topup/eddy for each session separately?
whenever you done a re-shim it is better to run them separately as eddy currently cannot use different susceptibility fields for different parts of the data.
> iii) If I have to do it separately how can I compare the results later on ? Do I need to register all DWI's for example to a mean image? If yes what happens with the b-vecs file? Do I have to reorientate the b-matrix?
That would be one way, and yes, in that case you would need to apply a rotation to the bvecs. Alternatively, you can do all calculations up to a scalar image (e.g. FA, MD or a track probability map), resample the scalar maps and compare those.
> I was wondering after I checked some files which I combined via fslmerge because they differed from the original single ones. I guess due to some motion between the sessions the fslmerge command do a kind of interpolation?
I am surprised to hear that. I don’t think fslmerge does anything more than a concatenation. Could it be a matter of different scaling of the display?
> So what would you suggest?
> Thank you for any kind of help.