Hi all,

Could someone tell me how I can install "SPHARM-PDM Toolbox" through Slicer? I can't find the extension in Extension Manager!

Thanks a lot,
-Mojdeh

Hi Mojdeh


What version of Slicer are you using?
(Help -> About Slicer)


Thank you,
Beatriz

Hi Beatriz,

I installed the newest version of Slicer (4.4.0) and successfully added the extension.

Thanks,
-Mojdeh

Hello 

I install 3D slicer 4.4 and install spharm-pdm extension, but I cann't see it on module part
please guide me

Hi Zohreh
Which platform are you using (linux, mac, which version)?
Slicer 4.4.0 external modules still seem to have some hickups on the Mac (it should work on linux though). Can you try Slicer 4.3.1? (you can download 4.3.1 here: http://slicer.kitware.com/midas3/slicerpackages/view ).

For usage information, check out the tutorial and the InsightJournal document (though both clearly need updating).

Martin

It seems that it also works with the 4.4.0 Nightly version (I have not tested it, but the module shows up after restart for the 6-22 nightly version).

Martin

Hi Matrin,
Thank you for your reply.

I install 3D slicer 4.3.1 on mac, I see spharm-pdm extension.
I search and download some document about using it, but I can't see result.
My image is, hippocampus segmentation that it's format is(.img,.hdr)
please guide me.

Thanks

Hi Zohreh,



You can find documentation and example data in the left column of this website (SPHARM-PDM toolbox - Docs), though as Martin said it is a little outdated. The mediawiki contains information as well (SPHARM-PDM toolbox column - MediaWiki). I think probably a good start is to process the example data (also hippocampus) and see if you can obtain the output data as described.


Hope that helps!
Beatriz

Hi Beatriz,

Thank you.

I do SegPostProcess at first step, output is created and I can see result, but in second step when I use output of this step on step GenParaMeshCLP, when I press the Apply button, status is: Completed with errors.

please help me.
thanks

I attach my error

Hi Zohreh,


Thanks for attaching your error. Could you show us what is the error message that is recorded in the error log above the one you attached?
(Click in the previous row, that says "Found Commandline module ...."). That one show display the standard terminal outputs from SPHARM-PDM.


Thanks,
Beatriz

Hi Beatriz.

Thank you for your prompt response.
I run again and attach all error.
explain about GenParaMeshCLP command line: error:

GenParaMeshCLP command line:

/Applications/Slicer.app/Contents/Extensions-22599/SPHARM-PDM/lib/Slicer-4.3/cli-modules/GenParaMeshCLP --outEulerName NULL --outLogName NULL --iter 500 --label 1 --initPara NULL /var/folders/79/hgpg9wds18g2n0t3g25sjzlc0000gn/T/Slicer/CFHE_vtkMRMLScalarVolumeNodeC.nrrd /var/folders/79/hgpg9wds18g2n0t3g25sjzlc0000gn/T/Slicer/CFHE_vtkMRMLModelNodeE.vtp /var/folders/79/hgpg9wds18g2n0t3g25sjzlc0000gn/T/Slicer/CFHE_vtkMRMLModelNodeF.vtp
GenParaMeshCLP standard output:
itk::ExceptionObject (0x7fff5fbfdc10)
Location: "Unknown"
File: /Users/kitware/Dashboards/Nightly/SlicerExtensions-431-build-64bits-QT4.7.4-PythonQt-With-Tcl-CLI-Release-nightly-4.3/SPHARM-PDM/Libraries/Shape/Algorithms/BinaryMask3DEqualAreaParametricMeshSource.txx
Line: 183
Description: Warning: Euler equation not satisfied. Euler Number : 0 - 0 = 0

thank you

this another error

Thank you so much

Hi Zohreh,


You are welcome!


The error you are getting in GenParaMeshCLP is related to the fact the topology of your input data is not spherical. Spherical topology is a requisite of SPHARM-PDM.
Euler number = 0 probably means your segmentation has a hole somewhere.
More on Euler Number: https://en.wikipedia.org/wiki/Euler_characteristic#Surfaces
I suggest you revise your input data.



I hope that helps.
Beatriz

Dear Beatriz,

Thanks again for your prompt response.

I change my data and I can create two output, 
1) Output Para Mesh.vtk 
2) Output Surface Mesh.vtk

I want to open them in 3D slicer, but I can't.
Please help me, how can see output file in 3D slicer ?
I can not see it on itksnap.

Please guide me.

Thanks.

Hi Zohreh,


Both of those outputs are surface models, name_of_the_data_surf.vtk is a triangular mesh representation of the input segmentation while name_of_the_data _para.vtk contains the spherical mapping of the previous file into the unit sphere (it should look like a small sphere in the 3D window). 

I suggest you do the 3D slicer data loading and visualization if you have trouble looking at the data. The 3DSlicer training compendium can be found here http://www.slicer.org/slicerWiki/index.php/Documentation/4.0/Training and it is a great resource to have.


Hope that helps,
Beatriz

Dear Beatriz, 

Thanks again,

I want do ParaToSPHARMMESH step, I give an error.
I do not know how can fill Output Directory and base filename.
I attach this step.

please guide me.
Thanks.

I solved my problem, 
I save output in file .coef or .mat

Hi,

I run shape analysis module, but I give an error. I attached.
Please guide me.

Thanks.

Hi, 

I do 3 below steps successfully, 

1. SegPostProcess
2. GenParaMeshCLP 
3. ParaToSPHARMMesh

I attach my result.
but I don't know how can visualized them and use spherical harmonic coefficient for classification in matlab.
please guide me.

Thanks.

Zoreh, can you start a new thread, as this question is unrelated to "Can't find 3D Slicer Extension". Also, can you be more specific in your postings so we can help you better.

I am not sure what you want to visualize, the vtk meshes or the coefficient files? 

Also, we do not support matlab, there is a separate matlab tool by Li Shen, but I don't know whether our coefficient files are compatible.

Martin

Hi Zohreh,


Thanks. 

1. Did you run your results using the ShapeAnalysisModule command or just running intermediate commands one-by-one? Running one by one is useful to diagnose problems, but it would be helpful if you could run the whole shape analysis pipeline via ShapeAnalysisModule and attach a zipped version of your output folder.

2. If I am guessing correctly, segpostprocess worked ok (your files Output Image1 look ok - by the way just an advice, refrain from using spaces in folder names and outputs, underscore is ok "_"), but something went wrong in ParaToSPHARM, when editing the surfSPHARM meshes you attach the vertices have NAN values. 

3. As Martin said, at the moment we do not provide support for SPHARM coefficient analysis. We do provide the coefficients in the *surfSPHARM.coef in an ASCII file that can be read with any standard i/o library, and have plenty of collaborators that have analyzed them by designing their own protocols. 

I hope that helps,

Beatriz