Hi Wayne,

skull stripping has the same effect as using a threshold, but if you are using 
a mask image everything should be ok. As for the automatic weight estimation 
you have to keep in mind that the result is very data dependent. The safest way 
is to take one representative dataset, run the automatic estimation and adjust 
the resulting particle weight until you get reasonable results. Reasonable in 
the sense of looking good :)

Peter

Von: Wayne Su [mailto:[email protected]]
Gesendet: Montag, 5. Mai 2014 19:34
An: Neher, Peter; [email protected]
Betreff: Re: [mitk-users] MITK-DT hanging on Gibbs Tracking

Hi Peter,
I didn't use a threshold. As I said in the previous email, this issue was 
caused by skull-stripped brain. To avoid this issue, I have to use whole head 
and brain mask.
For the particle weight estimation, would you recommend to specify one certain 
value for whole study, or let the program calculate it for each scan?
Thanks,
Wayne


On 14-05-03 4:21 AM, "Peter Neher" <[email protected]> wrote:
 Hi Wayne,

 as I can see from the output you are not using a mask image. Did you use a 
threshold when reconstructing your tensors/Qballs? This results in voxels 
containing only 0, which again results in this issue with the particle weight 
estimation. You can either use a mask image to restrict the weight estimation 
and tracking to valid voxels or you should reconstruct the image without the 
threshold to avoid zero-value voxels. I recommend using a mask image because 
the particle weight estimation will be much better if only performed inside of 
the brain/WM.

 Cheers
 Peter



On 05/02/2014 09:05 PM, Wayne Su wrote:


 MITK-DT hanging on Gibbs Tracking Hi There,

  I found MITK-DT was hanging on Gibbs Tracking whenever I  set 
"particle_length" "particle_width" or "particle_weight" to "auto" ("0" for gtp 
file).  Here are messages I got:

 [Fri May  2 11:49:28 2014
 0.320] In AutoLoadModulesFromPath at 
/home/wsu/devel/MITK/Core/CppMicroServices/core/src/util/usUtils.cpp:160 : 
Auto-loading module /usr/local/MITK/bin/MitkCore/libMitkVtkShaders.so
 [0.330] In AutoLoadModulesFromPath at 
/home/wsu/devel/MITK/Core/CppMicroServices/core/src/util/usUtils.cpp:160 : 
Auto-loading module /usr/local/MITK/bin/MitkCore/libMitkIpPicSupportIO.so
 [0.340] In AutoLoadModulesFromPath at 
/home/wsu/devel/MITK/Core/CppMicroServices/core/src/util/usUtils.cpp:160 : 
Auto-loading module /usr/local/MITK/bin/MitkCore/libMitkIOExt.so
 [0.340] In AutoLoadModulesFromPath at 
/home/wsu/devel/MITK/Core/CppMicroServices/core/src/util/usUtils.cpp:160 : 
Auto-loading module /usr/local/MITK/bin/MitkCore/libMitkDiffusionIO.so
 [0.380] Start GibbsTracking ..
 [4.770] Loading qball image ...
 [5.290] GibbsTrackingFilter: generating default mask image
 [5.290] GibbsTrackingFilter: loading parameter file param.gtp
 [5.290] GibbsTrackingFilter: parameter file loaded successfully
 [5.290] GibbsTrackingFilter: estimating particle weight
 [5.290] SphereInterpolator: loading lookuptables
 ParticleGrid: allocated 4mb for 100k particles.
 EnergyComputer: 1130496 active voxels found

 I guess something wrong with the automatic searching these parameters.

 I also got following messages from MitkDIffusion GUI:
 [56.780] enabling shader
 [56.790] enabling shader
 [56.820] enabling shader
 #57.870# [VtkError] ERROR: ERROR: In 
/home/wsu/devel/MITK-superbuild/VTK-src/Rendering/OpenGL/vtkShader2.cxx, line 82
 vtkShader2 (0x5ffa570): failed at glDeleteShader 1 OpenGL errors detected
   0 : (1281) Invalid value
 #57.870# [VtkError] ERROR: ERROR: In 
/home/wsu/devel/MITK-superbuild/VTK-src/Rendering/OpenGL/vtkShader2.cxx, line 82
 vtkShader2 (0x5fe0fe0): failed at glDeleteShader 1 OpenGL errors detected
   0 : (1281) Invalid value
 #57.870# [VtkError] ERROR: ERROR: In 
/home/wsu/devel/MITK-superbuild/VTK-src/Rendering/OpenGL/vtkShaderProgram2.cxx, 
line 103
 vtkShaderProgram2 (0x5a80fd0): failed at glDeleteProgram 1 OpenGL errors 
detected
   0 : (1281) Invalid value
 #57.870# [VtkError] ERROR: ERROR: In 
/home/wsu/devel/MITK-superbuild/VTK-src/Rendering/OpenGL/vtkShader2.cxx, line 82
 vtkShader2 (0x57609d0): failed at glDeleteShader 1 OpenGL errors detected
   0 : (1281) Invalid value
 #57.870# [VtkError] ERROR: ERROR: In 
/home/wsu/devel/MITK-superbuild/VTK-src/Rendering/OpenGL/vtkShader2.cxx, line 82
 vtkShader2 (0x567ab10): failed at glDeleteShader 1 OpenGL errors detected
   0 : (1281) Invalid value
 #57.870# [VtkError] ERROR: ERROR: In 
/home/wsu/devel/MITK-superbuild/VTK-src/Rendering/OpenGL/vtkShaderProgram2.cxx, 
line 103
 vtkShaderProgram2 (0x657c8d0): failed at glDeleteProgram 1 OpenGL errors 
detected
   0 : (1281) Invalid value

 I tried the official release and found the same issues.

 Wayne


------------------------------------------------------------------------------
Is your legacy SCM system holding you back? Join Perforce May 7 to find out:
&#149; 3 signs your SCM is hindering your productivity
&#149; Requirements for releasing software faster
&#149; Expert tips and advice for migrating your SCM now
http://p.sf.net/sfu/perforce
_______________________________________________
mitk-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mitk-users

Reply via email to