FYI, this is what I finally did to create the skeleton (it could be of help to another newbie like me :), being all_FA the 4D volume with all the FA images already registered, and 300 my FA threshold based on visual inspection (FA ranging from 0 to 1000 in my particular case):


***************************************************************************************************************************************************************************

# create mean FA
echo "creating valid mask and mean FA"
$FSLDIR/bin/fslmaths all_FA -max 0 -Tmin -bin mean_FA_mask -odt char
$FSLDIR/bin/fslmaths all_FA -mas mean_FA_mask all_FA
$FSLDIR/bin/fslmaths all_FA -Tmean mean_FA

# create skeleton of mean FA
$FSLDIR/bin/tbss_skeleton -i mean_FA -o mean_FA_skeleton

# create skeleton mask using threshold
${FSLDIR}/bin/fslmaths mean_FA_skeleton -thr 300 -bin mean_FA_skeleton_mask

# create skeleton distancemap (for use in projection search)
${FSLDIR}/bin/fslmaths mean_FA_mask -mul -1 -add 1 -add mean_FA_skeleton_mask mean_FA_skeleton_mask_dst
${FSLDIR}/bin/distancemap -i mean_FA_skeleton_mask_dst -o mean_FA_skeleton_mask_dst

# project all FA data onto skeleton
${FSLDIR}/bin/tbss_skeleton -i mean_FA -p 300 mean_FA_skeleton_mask_dst ${FSLDIR}/data/standard/LowerCingulum_1mm all_FA all_FA_skeletonised

***************************************************************************************************************************************************************************

Best,
Miguel


 

On Tue, Feb 16, 2010 at 7:42 PM, Miguel Burgaleta <[log in to unmask]> wrote:
Thanks David. I will play around with the script and see what happens. 

Best,
Miguel


On Tue, Feb 16, 2010 at 5:47 PM, David Flitney <[log in to unmask]> wrote:
I suspect not. Despite it's name, the tbss_3_postreg script, is actually doing registration stages. Instead, if you take a look at the script then I think you need to manually run stages after/around that with the line:

echo "merging all upsampled FA images into single 4D image"

Note the call to tbss_skeleton. 

Hope that helps.

On 16 Feb 2010, at 16:22, Miguel Burgaleta wrote:

Hi FSL experts,

I need to combine an in-house image registration protocol with TBSS. I was wondering if it is possible to feed tbss_3_postreg with a 4D file made of images that are already nonlinearly registered. I would just need to create the mean of those FAs and make the skeleton from it, right? I believe the former step can be easily given with fslmaths, but I don't see a straightforward way to make the skeleton with FSL's tools. Any help will be highly appreciated.

Thanks in advance,
Miguel


-- 
Dave Flitney, IT Manager
Oxford Centre for Functional MRI of the Brain
E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717