You must be logged in to post Login Register


Lost Your Password?
Search Forums:


 






Wildcard Usage:
*    matches any number of characters
%    matches exactly one character

Forum Diffusion Toolkit Usage step length

step length
Page:   1  
UserPost

1:34 am
August 25, 2011


kerstin

New Member

posts 1
Print this Post
1

Hi

I have a question about the step length with odf_tracker -rk2. By default, the step length should be 0.1 x (minimum) voxel size. So in my case for a voxel size of 2.5 mm isotropic, it should be 0.25 mm. However, when I import the trk files into matlab and calculate the distance between the imported points, I get values between 0.44 and 4.06. 

Spline filtering was turned off for tracking. The matlab-imported streamlines make sense anatomically, so I don't think there is something going wrong with the import.

What am I doing wrong?

Thanks

Kerstin

11:36 am
August 29, 2011


Ruopeng

Admin

posts 406
Print this Post
2

The step length is the step length of the propagation algorithm, not necessarily the step lengths of the tracks eventually saved. When the outgoing direction is not changed after a step, it will not be recorded.

Best,

Ruopeng

1:15 am
November 4, 2011


SusanZhong

Member

posts 8
Print this Post
3

Ruopeng said:

The step length is the step length of the propagation algorithm, not necessarily the step lengths of the tracks eventually saved. When the outgoing direction is not changed after a step, it will not be recorded.

Best,

Ruopeng

Dear  Ruopeng:

 

First thank you very much for you software. You  said some point are not recorded.I think you what to save storage space. But the distance between the point recorded should be the integer times of 0.1(minimum)the voxel size.I am not sure what I understand your meaning is correct?

 I have another trouble in finding the start point(seed) of the tracktography. I remember you said dti_tracker starts tracking from the center of each voxel. I think you mean if devided the recorded points of one track by the voxel size, there will be a point whoes coordinate is the integer times of the 0.5? Yes? But in my test, I can't find this point.

Is there something wrong?

Thanks and Best Regards
Susan

11:33 am
November 8, 2011


Ruopeng

Admin

posts 406
Print this Post
4

If there is no spline-filter applied, yes, the distance should be multiples of the minimum step-length. Again, if there is no spline-filter step applied, you should be able to find the seed point, not sure what's the point, though.

The spline-filter step does some optimization to smooth and clean up unecessary points on the track so it could happen to remove the seed point if there are close duplicates.

Ruopeng
Page:   1