Hi - your setup looks fine. Try to run it with fewer samples per voxel (e.g. 100) then the total time it would take with N samples is approximately N/100*(time for 100 samples).
Cheers
Saad
On 14 Nov 2012, at 11:57, David Soto wrote:
> hi - I am trying to segment the thalamus based on its connectivity
> to 5 cortical sites (a la Behrens et al, Nat Neuro 03).
> I derived left and right thalamus masks and cortical ROIs
> using freesurfer, all look good when uploaded in the individual
> T1. Eddy_correct and Bedpostx seemed to complete fine.
> I have attached 2 screenshots for the FDT probtrack GUI set up
> which I used to derive and replicate the code below to do
> protracks across subjects/left&right thalamus
>
> The problem appears that it does not seem to run well
> I left one subjects running for almost 24 h now and can not
> see any hint of progression. I have a linux box with 32 cores and
> 128 Gb RAM.
>
> probtrackx appears to keep running and can see from the Terminal
> that is using 15% of RAM-
>
> I am not sure all this is expected and
> would really appreciate some advise
>
> Cheers
>
> /usr/share/fsl/5.0/bin/probtrackx2 -x /home/dsoto/Documents/fmri/rsdtianawmgui/02ST/label/lh_thalamus_ds.nii.gz -l --onewaycondition --pd -c 0.2 -S 2000 --steplength=0.5 -P 5000 --fibthresh=0.01 --distthresh=0.0 --sampvox=0.0 --xfm=/home/dsoto/Documents/fmri/rsdtianawmgui/02ST/diffdata/transforms/t12dti.mat --forcedir --opd -s /home/dsoto/Documents/fmri/rsdtianawmgui/02ST/diffdata.bedpostX/merged -m /home/dsoto/Documents/fmri/rsdtianawmgui/02ST/diffdata.bedpostX/nodif_brain_mask --dir=/home/dsoto/Documents/fmri/rsdtianawmgui/02ST/diffdata.bedpostX/leftthalfdt --targetmasks=/home/dsoto/Documents/fmri/rsdtianawmgui/02ST/diffdata.bedpostX/leftthalfdt/targets.txt --os2t
> <fdt1.png><fdt2.png>
--
Saad Jbabdi
University of Oxford, FMRIB Centre
JR Hospital, Headington, OX3 9DU, UK
(+44)1865-222466 (fax 717)
www.fmrib.ox.ac.uk/~saad
|