Dear Maxim,
> Somehow tha FA maps obtained by DTI fit look distored and seem to
> have some
> extracerebral white parts, although the nodif_brain and
> nodif_brain_mask
> looks alright! I'm running DTI Fit on DTI data obtained in a Siemens
> TRIO. I
> converted the dicom data with dicomtonifti script - so i was
> confident with
> the bval and bvec output?
could it be that there are bad distortions in the diffusion weighted
scans? If there are bad eddy current-induced distortions in your dwi's
that would lead to ribbons of high FA along edges in the images (e.g.
on the outer surface of the cortex). If, in addition, the diffusion
gradients are sampled on a half-sphere (as opposed to a full sphere)
there will be a non-zero average distortion in the dwi's, leading to
distorted FA image.
I suggest you run a movie of all your dwi's and see if it jumps around
a lot.
Good luck Jesper
|