Hi,
It is actually a little more complicated than this. There are
two things that you need to be careful about - the ordering
of the data (which determines how it looks in FSLView)
and the labeling of the axes (as shown in FSLView). It is only
the ordering that will affect registration, but you need to
make sure that both are correct.
The command avwswapdim $i x -y z $i will change the
ordering of the data *and the labeling* which can be a
problem.
Before doing any avwswapdim commands, can you
check whether the labels are correct or incorrect, by
looking at them in FSLView, and tell us what they are
like first (e.g. x-axis is LR and correct, y-axis is A-P but
incorrect, z-axis is S-I and correct)?
Once we have this information then we can give you
the right set of commands (possibly involving
avworient -swaporient) that will reorient your data as
you want while still preserving your labels.
Can you also please send us the output of avwhd
on an image.
Sorry that this is complicated, but fixing incorrectly
reconstructed images can be very problematic and
care needs to be taken.
All the best,
Mark
Tim Behrens wrote:
> for i in *FA*; do avwswapdim $i x -y z $i ;done
>
> will flip all your FA maps in AP.
>
> Cheers
>
> T
>
>
>
>
> On 17 Oct 2006, at 08:38, Martin Kavec wrote:
>
>> Hello to everybody,
>>
>> After running tbss_3_postreg, I get very bad registration of mean_FA
>> to MNI152. I guess it is because my original FA maps are flipped in
>> A-P direction with respect to MNI152. I tried to A-P rotate
>> $FSLDIR/etc/standard/avg152T1_brain, but the result was even worse.
>> Is there any quick cure?
>>
>> Thanks in advance.
>>
>> Martin
>
|