hi!
 
I need a templete for shape analysis of lateral ventricle. I want to create templete with my data or use one of my datas as templete, how can I do it?
 
Thanks,
Gozde

Hi Gozde,

MeshMath has an option to compute mean shapes out of shapes that already have correspondence established. That means, using this method, you would have to run SPHARM a first time, then compute the average, and then compute a second time using your average as the template. Is this something you can use in your studies?

Regards,
BEATRIZ

Hi Beatriz..

Thanks for your response. I read MeshMath usege and ı think -avgMesh function can use for create template.


-avgMesh ...

in this function I try para.meta, surf.meta and .gipl as input file. But it didnt run.

What should be the input file ?

Thanks..

Gozde

With MeshMath you need to use the *surfSPHARM.vtk meshes, that are the ones that have correspondence. Lets say you have 20 LV in your population, you run all the steps in SPHARM, and you end up having 20 surfSPHARM.vtk meshes.

Then you use MeshMAth:
MeshMAth surfSPHARM_1.vtk -avgMesh surfSPHARM_2.vtk ... surfSPHARM_20.vtk

And you get an average mesh that you will have to scan convert to gipl in order to use in yet another spharm computation.
I want to emphasize that I was not suggesting this is the only way to proceed. Is only one more option. You can always use as template your first shape in the population.


Regards,
BEATRIZ

To use as templete my first shape in the population how can I create templete.coef for use in following step?

./ParaToSPHARMMesh 1_surf.meta 1_para.meta -subdivLevel 20 -spharmDegree 15 \-flipTemplete templete.coef -regTemplete templete.meta


thanks

Gozde

ParaToSPHARMMeshCLP arguments --filpTemplate and --regTemplate are used for all the shapes except for the template. Those two arguments are optional. That means for the first one, you can just not specify those two arguments. You want to specify --paraOut to get all the color maps writen as outputs, that can be very helpful in detecting flips in the data.

BEATRIZ

Thanks for response and I also got this error in GenParaMesh method:

itk::ExceptionObject (0xbfbdcb04)
Location: "Unknown"
File: /home/lbompard/NeuroLib/Libraries/Shape/Algorithms/BinaryMask3DEqualAreaParametricMeshSource.txx
Line: 172
Description: Warning: Euler equation not satisfied. Euler Number : 15636 - 15638 = -2

How can I fix the problem?

Thanks
Gozde

You need to revise your segmentation and see there are no holes in there. If you are working with LV you will have to fill the choroid plexus (that are not holes, but look almost like them).