Hi Katie,
I would avoid reorienting some of the FA images beforehand...
What you can do is to edit your tbss_2_reg script. Line #96, you can change the fsl_reg options to specify flirt ones:
echo "$FSLDIR/bin/fsl_reg $g $f ${g}_to_$f -e -flirt "-flirtoptionyouwant" -FA" >> .commands
or just run: $FSLDIR/bin/fsl_reg $g $f ${g}_to_$f -e -flirt "-flirtoptionyouwant" -FA
if there are only a couple of brains that requires this pre-processing.
Cheers,
Gwenaelle
--- En date de : Mer 18.3.09, Katie Karlsgodt <[log in to unmask]> a écrit :
> De: Katie Karlsgodt <[log in to unmask]>
> Objet: [FSL] mis-registration on tbss_2
> À: [log in to unmask]
> Date: Mercredi 18 Mars 2009, 16h52
> Hi,
> We have a couple of brains that do not register correctly
> using tbss_2, but that do register correctly in flirt if you
> use the full search (-180 to 180 degrees). The matrices that
> come out of flirt and tbss also seem very different. I see
> that flirt is called by tbss_2_reg and then fsl_reg, but am
> not sure what I could do to change the flirt parameters that
> are implemented. Is there a way to do that so at least for
> the difficult cases they can still go through tbss? (there
> isn't anything very obvious wrong with the raw images
> other than 1 or 2 slices cut off the top of a few, and quite
> large ventricles in some). Any advice would be appreciated!
>
> thanks!
> Katie
>
>
> _________________________________________
> Katie Karlsgodt, Ph.D.
> Post Doctoral Fellow, Neurogenetics Affinity Group
> Semel Institute
> University of California, Los Angeles
>
> phone: 310-794-9673
> fax: 310-794-9740
