Hi Saad,
That makes sense. Thanks for your help.
Roeland

On Nov 25, 2013, at 2:40 AM, Saad Jbabdi <[log in to unmask]> wrote:

Hi Roeland

I should have thought of this before, but the behaviour that you are observing is completely normal. Every tract is seeded in two directions, and therefore two different masks can be counted as connecting to the same sample streamline even if they are also stopping masks, if they connect to the separate halves of the streamline. 
You will notice that in your resulting matrix, the rows of the matrix (if you exclude the seed ROI) sum to a maximum of 10000, which is 2*5000.

I hope that the above is clear?


Cheers
Saad




On 25 Nov 2013, at 02:48, Hancock, Roeland <[log in to unmask]> wrote:

Thanks Saad. I've sent you the files.
Roeland

On Nov 23, 2013, at 2:08 AM, Saad Jbabdi <[log in to unmask]> wrote:

Not sure - I’ll need to look at all your masks to figure this out. You can use the link below to upload your (tar-zipped) data so I can try to replicate what you’ve done.


Cheers
Saad


On 22 Nov 2013, at 17:04, Hancock, Roeland <[log in to unmask]> wrote:

I am using FSL 5.0.5. The target masks do not overlap.


On Nov 22, 2013, at 5:00 AM, Saad Jbabdi <[log in to unmask]> wrote:

Hi

Are you using the latest version of probtrackx2? 

Can you check that your target masks don’t overlap.

Saad

On 22 Nov 2013, at 01:45, Hancock, Roeland <[log in to unmask]> wrote:

Hi list,
I would like to know the proportion of streamlines extending from a seed to set of a target regions (similar to https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=FSL;bbf1e232.1202). My target regions are a subset of my stop mask, so that tracking will stop once each target is reached and give the desired counts. My command is

probtrackx2 -x SEED -l -c 0.2 -S 2000 --steplength=0.5 -P 5000 --fibthresh=0.01 --distthresh=0.0 --sampvox=0.0 --forcedir --opd -s bedpostX/merged -m bedpostX/nodif_brain_mask --dir=SEED.probtrack --stop=STOPMASK --targetmasks=masks.txt --os2t --s2tastext 

I expect each row of the seed to target matrix to sum to <= Nsamples (5000), but this is not the case. Instead, I get values of up to 5000 for multiple columns and fdt_paths suggests that streamlines are extending several voxels into the stop mask.

Is there a bug related to termination masks, or I am misunderstanding some of the options?

Thanks,
Roeland Hancock, PhD
Postdoctoral Fellow
Department of Psychiatry
UC San Francisco




--
Saad Jbabdi
University of Oxford, FMRIB Centre

JR Hospital, Headington, OX3 9DU, UK
(+44)1865-222466  (fax 717)



--
Saad Jbabdi
University of Oxford, FMRIB Centre

JR Hospital, Headington, OX3 9DU, UK
(+44)1865-222466  (fax 717)



--
Saad Jbabdi
University of Oxford, FMRIB Centre

JR Hospital, Headington, OX3 9DU, UK
(+44)1865-222466  (fax 717)