Issues with createSurface.py
Closed this issue · 7 comments
The freesurfer files that I'm trying to run Hipsta on are from Freesurfer7.2 labeled: rh.hippoAmygLabels-T1.v21.CA.mgz, lh.hippoAmygLabels-T1.v21.CA.mgz
When looking at results through freeview these files are either unable to open or blank:
- lh.labels.mgz
- lh.mask.mgz
- lh.cropped.mgz
- lh.orig.mgz
- lh.initial_mask.mgz
- lh.close_mask.mgz
- lh.gaussian_filter.mgz
but lh.image.mgz seems to be fine
This is the error I'm getting:
`Smooth surface
--> VTK format ...
[file not found or not readable]
[ERROR: hipsta.py] Error Information:
[ERROR: hipsta.py] Type: <class 'AttributeError'>
[ERROR: hipsta.py] Value: ("'NoneType' object has no attribute 'rm_free_vertices_'",)
Traceback (most recent call last):
File "/Users/joshuanguyen/.pyenv/versions/3.10.11/envs/myenv/bin/run_hipsta", line 8, in
sys.exit(main())
File "/Users/joshuanguyen/.pyenv/versions/3.10.11/envs/myenv/lib/python3.10/site-packages/hipsta/cli/init.py", line 26, in main
_run_hipsta(args)
File "/Users/joshuanguyen/.pyenv/versions/3.10.11/envs/myenv/lib/python3.10/site-packages/hipsta/hipsta.py", line 840, in _run_hipsta
run_analysis(params)
File "/Users/joshuanguyen/.pyenv/versions/3.10.11/envs/myenv/lib/python3.10/site-packages/hipsta/hipsta.py", line 743, in run_analysis
params = smoothSurface(params)
File "/Users/joshuanguyen/.pyenv/versions/3.10.11/envs/myenv/lib/python3.10/site-packages/hipsta/createSurface.py", line 272, in smoothSurface
triaMesh.rm_free_vertices()
AttributeError: 'NoneType' object has no attribute 'rm_free_vertices'`
Please advise
Hello, thanks for your message. I speculate that the issue may be due to using the lh.hippoAmygLabels-T1.v21.CA.mgz and rh.hippoAmygLabels-T1.v21.CA.mgz files, instead of lh.hippoAmygLabels-T1.v21.mgz and rh.hippoAmygLabels-T1.v21.mgz (i.e., no CA suffix). Would using the latter files help?
Thank you! That got me a bit further but now I have a new error - do you know what this might be now?
Computing cube parametrization
--> VTK format ...
--> DONE ( V: 20178 , T: 61793 )
--> VTK format ...
--> DONE ( V: 20178 , T: 16905 )
TriaMesh with anisotropic Laplace-Beltrami
Solver: spsolve (LU decomposition) ...
[ERROR: hipsta.py] Error Information:
[ERROR: hipsta.py] Type: <class 'ValueError'>
[ERROR: hipsta.py] Value: ('all the input arrays must have same number of dimensions, but the array at index 0 has 1 dimension(s) and the array at index 1 has 0 dimension(s)',)
Traceback (most recent call last):
File "/Users/joshuanguyen/.pyenv/versions/3.12.3/bin/run_hipsta", line 8, in
sys.exit(main())
^^^^^^
File "/Users/joshuanguyen/.pyenv/versions/3.12.3/lib/python3.12/site-packages/hipsta/cli/init.py", line 26, in main
_run_hipsta(args)
File "/Users/joshuanguyen/.pyenv/versions/3.12.3/lib/python3.12/site-packages/hipsta/hipsta.py", line 840, in _run_hipsta
_run_analysis(params)
File "/Users/joshuanguyen/.pyenv/versions/3.12.3/lib/python3.12/site-packages/hipsta/hipsta.py", line 783, in _run_analysis
params = computeCubeParam(params)
^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/joshuanguyen/.pyenv/versions/3.12.3/lib/python3.12/site-packages/hipsta/computeCubeParam.py", line 800, in computeCubeParam
v4c, t4c, i4c, k4c, newVtcs, newVtcsSgn, newTetra = getSeam(
^^^^^^^^
File "/Users/joshuanguyen/.pyenv/versions/3.12.3/lib/python3.12/site-packages/hipsta/computeCubeParam.py", line 519, in getSeam
v4c, t4c, i4c, k4c, e4cBndOpen, newTetra, newVtcs, newVtcsAdj, newVtcsSgn = getSeamCase1(
^^^^^^^^^^^^^
File "/Users/joshuanguyen/.pyenv/versions/3.12.3/lib/python3.12/site-packages/hipsta/computeCubeParam.py", line 88, in getSeamCase1
i4c = np.concatenate((i4c, st.mode(i4c[t4c[i, :]])[0]))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ValueError: all the input arrays must have same number of dimensions, but the array at index 0 has 1 dimension(s) and the array at index 1 has 0 dimension(s)
(hipsta_env) (base) joshuanguyen@dhcp-v066-140 Hipsta %
Hm, this may be a bit more tricky. Could you maybe post the command that you used for running the software?
Thank you!
This is the command that I gave:
(hipsta_env) (base) joshuanguyen@dhcp-v066-140 Hipsta % run_hipsta --filename /Users/joshuanguyen/Desktop/ASYM/Data/6049lh.hippoAmygLabels-T1.v21.mgz --hemi lh --lut freesurfer --outputdir /Users/joshuanguyen/Desktop/ASYM/Results
And this was the entire output
`----------------------------------------
Hippocampal shape and thickness analysis
Found output directory /Users/joshuanguyen/Desktop/ASYM/Results
[INFO: logging.py] Starting logging ...
[INFO: logging.py] Logfile: /Users/joshuanguyen/Desktop/ASYM/Results/logfile.txt
[INFO: logging.py] Version: 0.9.1
[INFO: logging.py] Date: 03/06/2024 12:47:07
[INFO: logging.py] Command: /Users/joshuanguyen/.pyenv/versions/3.12.3/bin/run_hipsta --filename /Users/joshuanguyen/Desktop/ASYM/Data/6049lh.hippoAmygLabels-T1.v21.mgz --hemi lh --lut freesurfer --outputdir /Users/joshuanguyen/Desktop/ASYM/Results
[INFO: hipsta.py] Evaluating arguments ...
[INFO: atlases.py] Found internal, modified look-up table for FreeSurfer.
[INFO: hipsta.py] Using /Users/joshuanguyen/Desktop/ASYM/Data/6049lh.hippoAmygLabels-T1.v21.mgz as input file
[INFO: hipsta.py] Using /Users/joshuanguyen/Desktop/ASYM/Results as output directory
[INFO: hipsta.py] Found /Users/joshuanguyen/Desktop/ASYM/Data/6049lh.hippoAmygLabels-T1.v21.mgz
[INFO: hipsta.py] Starting convertFormat() ...
Convert to mgz and copy to output directory
[INFO: hipsta.py] Starting cropImage() ...
Cropping
/Applications/freesurfer/7.2.0/bin/mri_mask -bb 5 /Users/joshuanguyen/Desktop/ASYM/Results/image/lh.orig.mgz /Users/joshuanguyen/Desktop/ASYM/Results/image/lh.orig.mgz /Users/joshuanguyen/Desktop/ASYM/Results/image/lh.cropped.mgz
[INFO: hipsta.py] Starting upsampleImage() ...
[INFO: hipsta.py] Starting copy_image_to_main() ...
[INFO: hipsta.py] Starting autoMask() ...
[INFO: hipsta.py] Starting createLabels() ...
Create labels
/Applications/freesurfer/7.2.0/bin/mri_binarize --i /Users/joshuanguyen/Desktop/ASYM/Results/lh.image.mgz --match 234 236 238 240 246 --o /Users/joshuanguyen/Desktop/ASYM/Results/labels/lh.initial_labels.mgz
/Applications/freesurfer/7.2.0/bin/fscalc /Users/joshuanguyen/Desktop/ASYM/Results/lh.image.mgz mul /Users/joshuanguyen/Desktop/ASYM/Results/labels/lh.initial_labels.mgz --o /Users/joshuanguyen/Desktop/ASYM/Results/labels/lh.initial_labels.mgz
[INFO: hipsta.py] Starting mergeMolecularLayer() ...
Attaching the molecular layer
[INFO: hipsta.py] Starting copy_labels_to_main() ...
[INFO: hipsta.py] Starting binarizeMask() ...
[INFO: hipsta.py] Starting gaussFilter() ...
Gaussian filtering
[INFO: hipsta.py] Starting longFilter() ...
[INFO: hipsta.py] Starting closeMask() ...
Applying closing operation to mask
[INFO: hipsta.py] Starting copy_mask_to_main() ...
[INFO: hipsta.py] Starting extractSurface() ...
Creating surface via marching cube algorithm
/Applications/freesurfer/7.2.0/bin/mri_mc /Users/joshuanguyen/Desktop/ASYM/Results/lh.mask.mgz 1 /Users/joshuanguyen/Desktop/ASYM/Results/surface/lh.initial_surf.vtk 1
[INFO: hipsta.py] Starting remeshSurface() ...
[INFO: hipsta.py] Starting smoothSurface() ...
Smooth surface
--> VTK format ...
--> DONE ( V: 10970 , T: 21936 )
[INFO: hipsta.py] Starting qcPlots() ...
--> VTK format ...
--> DONE ( V: 10970 , T: 21936 )
[INFO: hipsta.py] Starting checkSurface() ...
Check surfaces
--> VTK format ...
--> DONE ( V: 10970 , T: 21936 )
[INFO: check_surface.py] Euler number for /Users/joshuanguyen/Desktop/ASYM/Results/lh.surf.vtk is 2
[INFO: hipsta.py] Starting createTetraMesh() ...
Create tetrahedral mesh
--> VTK format ...
--> DONE ( V: 10970 , T: 21936 )
/Applications/Gmsh.app/Contents/MacOS/gmsh -3 -o /Users/joshuanguyen/Desktop/ASYM/Results/tetra-mesh/lh.tetra-gmsh.vtk /Users/joshuanguyen/Desktop/ASYM/Results/tetra-mesh/lh.tetra.geo
--> VTK format ...
--> DONE ( V: 17313 , T: 67910 )
[INFO: hipsta.py] Starting createTetraLabels() ...
Creating label files for tetrahedral meshes
/Applications/freesurfer/7.2.0/bin/mris_calc --output /Users/joshuanguyen/Desktop/ASYM/Results/tetra-labels/lh.labels-bnd.mgz /Users/joshuanguyen/Desktop/ASYM/Results/tetra-labels/lh.labels-bnd.mgz lowerlimit /Users/joshuanguyen/Desktop/ASYM/Results/tetra-labels/lh.labels-ca4.mgz
--> VTK format ...
--> DONE ( V: 17313 , T: 67910 )
Found 21936 triangles on boundary.
[INFO: hipsta.py] Starting removeBoundaryMask() ...
Removing boundary tetras from mesh
--> VTK format ...
--> DONE ( V: 17313 , T: 67910 )
Found 21936 triangles on boundary.
[INFO: hipsta.py] Starting cutTetra() ...
Cutting tetrahedral mesh
--> VTK format ...
--> DONE ( V: 17313 , T: 67910 )
TetMesh with regular Laplace
Matrix Format now: csc
Solver: spsolve (LU decomposition) ...
/Users/joshuanguyen/.pyenv/versions/3.12.3/lib/python3.12/site-packages/scipy/sparse/_index.py:108: SparseEfficiencyWarning:
Changing the sparsity structure of a csr_matrix is expensive. lil_matrix is more efficient.
Found 21696 triangles on boundary.
Searched mesh after 177 flooding iterations (0.05073904991149902 sec).
[INFO: hipsta.py] Starting checkSurface() ...
Check surfaces
--> VTK format ...
--> DONE ( V: 8787 , T: 16905 )
[INFO: check_surface.py] There are 2 boundary loops for /Users/joshuanguyen/Desktop/ASYM/Results/tetra-cut/lh.rm.open.bnd.cut.vtk
[INFO: hipsta.py] Starting computeCubeParam() ...
Computing cube parametrization
--> VTK format ...
--> DONE ( V: 20178 , T: 61793 )
--> VTK format ...
--> DONE ( V: 20178 , T: 16905 )
TriaMesh with anisotropic Laplace-Beltrami
Solver: spsolve (LU decomposition) ...
[ERROR: hipsta.py] Error Information:
[ERROR: hipsta.py] Type: <class 'ValueError'>
[ERROR: hipsta.py] Value: ('all the input arrays must have same number of dimensions, but the array at index 0 has 1 dimension(s) and the array at index 1 has 0 dimension(s)',)
Traceback (most recent call last):
File "/Users/joshuanguyen/.pyenv/versions/3.12.3/bin/run_hipsta", line 8, in
sys.exit(main())
^^^^^^
File "/Users/joshuanguyen/.pyenv/versions/3.12.3/lib/python3.12/site-packages/hipsta/cli/init.py", line 26, in main
_run_hipsta(args)
File "/Users/joshuanguyen/.pyenv/versions/3.12.3/lib/python3.12/site-packages/hipsta/hipsta.py", line 840, in _run_hipsta
_run_analysis(params)
File "/Users/joshuanguyen/.pyenv/versions/3.12.3/lib/python3.12/site-packages/hipsta/hipsta.py", line 783, in _run_analysis
params = computeCubeParam(params)
^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/joshuanguyen/.pyenv/versions/3.12.3/lib/python3.12/site-packages/hipsta/computeCubeParam.py", line 800, in computeCubeParam
v4c, t4c, i4c, k4c, newVtcs, newVtcsSgn, newTetra = getSeam(
^^^^^^^^
File "/Users/joshuanguyen/.pyenv/versions/3.12.3/lib/python3.12/site-packages/hipsta/computeCubeParam.py", line 519, in getSeam
v4c, t4c, i4c, k4c, e4cBndOpen, newTetra, newVtcs, newVtcsAdj, newVtcsSgn = getSeamCase1(
^^^^^^^^^^^^^
File "/Users/joshuanguyen/.pyenv/versions/3.12.3/lib/python3.12/site-packages/hipsta/computeCubeParam.py", line 88, in getSeamCase1
i4c = np.concatenate((i4c, st.mode(i4c[t4c[i, :]])[0]))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ValueError: all the input arrays must have same number of dimensions, but the array at index 0 has 1 dimension(s) and the array at index 1 has 0 dimension(s)
`
Thank you very much. I have to admit that I am a bit at a loss at the moment, and see no obvious reason for the processing failure. If you can share the particular segmentation file that produces the error I am happy to take a closer look while trying to replicate the issue.