Dear Tal,
unfortunately, there is no easy way (so far) to
convert an areg .dof into a flirt .mat... So the best
way to continue the tbss preprocessings is to register
your diffusion images in two steps.
You can do so by editing tbss_3_postreg script and by:
1) replacing the areg transformation from the best
target to the MNI152 by the flirt one in the
tbss_3_postreg script:
- erase line 68
- replace line 69: ${D}/transformation target.hdr
target_to_MNI152.hdr -dofin target_to_MNI152.dof
-linear -target $M
by -> $FSLDIR/bin/flirt -in target.hdr -ref $M -out
target_to_MNI152.hdr -init flirtaffinetransf.mat
-applyxfm
2) separating the two transformations applied to your
diffusion images:
- replace line 76: ${D}/transformation2 ${f}.hdr
${f}_to_${best}_nonlinear_hr.hdr -dofin2
${f}_to_${best}_nonlinear.dof -dofin1
../stats/target_to_MNI152.dof -linear -target
../stats/$M
by -> ${D}/transformation ${f}.hdr
${f}_to_${best}_tmp.hdr -dofin
${f}_to_${best}_nonlinear.dof -linear -target
${best}.hdr
followed by -> $FSLDIR/bin/flirt -in
${f}_to_${best}_tmp.hdr -ref $M -out
${f}_to_${best}_nonlinear_hr.hdr -init
flirtaffinetransf.mat -applyxfm
Hope this will help,
Gwenaelle
--- Tal Herbsman <[log in to unmask]> a écrit :
> Hi Gwenaelle,
> Thanks for the tip. I just got around to giving it a
> try and the FLIRT 12
> deg affine transform behaved much better than areg
> as you suggested. One
> question though, how do I go about converting the 12
> dof .mat affine
> transform matrix to the 15 dof areg .dof so I can
> continue using
> tbss_3_postreg script?
>
> Thanks,
> Tal.
>
> On Wed, 10 Jan 2007 10:52:10 +0100, Gwenaëlle DOUAUD
> <[log in to unmask]> wrote:
>
> >Hi Tal,
> >
> >not sure what happened, but I would say that this
> is
> >only the affine registration between your best
> target
> >and the T1 MNI152 that didn't work correctly. Could
> >you check so? If it is correct, then run flirt
> which
> >is much more robust (instead of areg) to generate
> the
> >affine transformation...
> >
> >Hope this helps,
> >Gwenaelle
> >
> >
> >--- Tal Herbsman <[log in to unmask]> a écrit :
> >
> >> I redid a tbss run and got a skeleton which is
> >> poorly alligned w/ the MNI
> >> template (see attached .png image --> corpus
> >> collusum is well off MNI
> >> corpus, etc...) What do you folks suggest? Can I
> >> refine the registration to
> >> MNI space? Should I try to select another target
> >> instead of the best fit
> >> target generated in tbss step 2 (I ran the
> hungry,
> >> no target version of the
> >> algorithm) in an attempt to get a better
> >> registration?
> >>
> >> Is this strictly an aesthetic issue--i.e. the
> >> figures generated from the
> >> tbss run will look poorly alligned but results
> are
> >> still valid--or should I
> >> be looking for some deeper problem?
> >>
> >> Thanks as always,
> >>
> >> Tal.
> >>
> >>
> >> -------------------------
> >>
> >> Tal Herbsman, MD
> >>
> >> Department of Psychiatry
> >> Medical University of South Carolina
> >> 67 President Street, 5 South
> >> Charleston, SC 29425
> >>
> >> Pager: (843) 792-2123 #15493
> >>
> >
> >
> >__________________________________________________
> >Do You Yahoo!?
> >En finir avec le spam? Yahoo! Mail vous offre la
> meilleure protection
> possible contre les messages non sollicités
> >http://mail.yahoo.fr Yahoo! Mail
>
>========================================================================
>
___________________________________________________________________________
Découvrez une nouvelle façon d'obtenir des réponses à toutes vos questions !
Profitez des connaissances, des opinions et des expériences des internautes sur Yahoo! Questions/Réponses
http://fr.answers.yahoo.com
|