On Wed, 10 Aug 2005 14:31:02 +0100, Anusha Sritharan
<[log in to unmask]> wrote:
>Hi
>
>I am not sure whether I should send this mail to spm or fsl or to both,
>but decided on sending to spm first.
>
>I have a dti data set from which I have computed the FA maps using FSL-FDT
>tool box. I am planning to do VBM type analysis on FA maps using SPM2.
>
>Here is my problem:
>
>The B0 and all other images used for FA calculation when displayed using
>SPM2 have voxel size (-1.88 1.88 2.5). However, when I displayed FA map
>and other related images(computed using fsl-fdt) the voxel size is (1.88
>1.88 2.5) with left-right flipped (I checked this by looking at the global
>shape of the images). But, when these images are displayed using fslview
>they all looked very well matched, ie there is no L-R flipping in the FA
>maps.
I don't have a specific answer. But I do recall that when I used the FSL
skull-stripping tool on Analyze images, the resulting image was left-right
flipped when I viewed it in SPM2.
Not sure whether the interpretation of the orientation of Analyze images
are different between FSL and SPM2; I just dealt with it on an ad hoc
basis (in my case, wrote a Perl script to change the sign on the x-pixel-
size in the *.hdr file).
>I am con-puzzeled by this. I need to do spatial normalsation of Fa maps
>before perfomring the t-test comparison between groups and I am thinking
>of flipping the FA maps using reorient in SPM2 so that it will be in same
>L-R orientation as the templates.
>
>I look forward to hear from you all.
>
>Thanks in advance
>
>Anusha
>
|