I understand if the two sides of a streamline sample ends in the same target region A, the streamline sample will record in seed_to_targetA and this steramline sample will contribute only once to the increasing sample number in seed_to_targetA.Is it right?

    But in the fdt_paths,no matter which case(end in the same region or not), a streamline sample will contribute twice to the increasing  sampel mumber in fdt_paths. Yes?

 

    The understanding of the computation details is usful to promote the process of my experiment .  Thank you in advance!!

 

Suyu 

 

  

 

 

   

 

2012/4/17 Saad Jbabdi <[log in to unmask]>

both.

On 17 Apr 2012, at 15:11, 仲苏玉 wrote:

Dear Saad

 

  Thank you very much.

     With the option of the classification targets, if one side of a streamline sample ends in target region A,while the other side ends in target region B, which file will record this streamline sample.Seed_to_targetA.nii,or Seed_to_targetB.nii?

   

         

   

     Best Regards,

 

  Suyu

 

 

  

   

2012/4/17 Saad Jbabdi <[log in to unmask]>

Hi

Both sides of a streamline sample are considered as the same sample, i.e. only counted once.

Saad

On 16 Apr 2012, at 12:04, 仲苏玉 wrote:

Dear Saad,

 

    Thank you very much for your replay.

    

     Yes, fdt_paths and seeds_to_targets indeed contain difference information. Fdt_path containing the output connectivity distribution represents the number of samples that from all voxels of  the  seed mask to one voxel.  

 

    I want to know if this difference is due to the differences in count methods.

 

    As the fact that fsl track a fiber in 2 directions and I want to know  if one sample will be counted twice in fdt_path or only once ? And how about in the situation of seeds_to_target? For example,if two end points of the one fiber terminate in one region, the fiber will be considered as once or twice? And how about  two end points of one fiber terminated in two regions?

 

    Best Regards and Thank you in advance

Suyu

2012/4/14 Saad Jbabdi <[log in to unmask]>

Hi - fdt_paths and seeds_to_targets do not contain the same information, that is why your results are not the same.
In your case you want to use method1.
Have a look at the probtrackx doc where these files are described.

Cheers
Saad

On 11 Apr 2012, at 02:38, Zhong Suyu wrote:

> Dear FSL experts:
>
> I am freshman of FSL.
>
> I am trying to do whole brain probabilistic tractography and to calculate connection probabilities between a set of N regions.
> First, I set one of the ROIs(i=1....N) as the seed mask , set  all the rest ROIs as the stop mask(--stop),  and set a list of targets that include all ROIs. And do protrackx. As a result,I got NXN seed(m)_to_<targets>(n) (m,n=1....N) files and N fdt_path files (each fdt_path contain the connectivity distribution from one of the N seed mask ROIs to the rest ROIs ).
> And then I determine the NxN connections in two ways: the fist way is sum the value of each voxel within each seed(m)_to_<targets>(n) file as the mth row and the nth column element of the connection. The second way is to use the fdt_path and fslstats command.
>
> I want to know if the results from the two methods are same. In my experiment, I got the different results.
> I find the number of samples got from seed_to_<target> is less than the one from fdt_path. Is it right? If yes, why the numbers are not the same?
>
>
> Thank you in advance!
>
> Suyu
>

--
Saad Jbabdi
University of Oxford, FMRIB Centre

JR Hospital, Headington, OX3 9DU, UK
(+44)1865-222466  (fax 717)
www.fmrib.ox.ac.uk/~saad

--

Saad Jbabdi

University of Oxford, FMRIB Centre

JR Hospital, Headington, OX3 9DU, UK

(+44)1865-222466  (fax 717)

www.fmrib.ox.ac.uk/~saad

--

Saad Jbabdi

University of Oxford, FMRIB Centre

JR Hospital, Headington, OX3 9DU, UK

(+44)1865-222466  (fax 717)

www.fmrib.ox.ac.uk/~saad