spinalcordtoolbox / spinalcordtoolbox

Comprehensive and open-source library of analysis tools for MRI of the spinal cord.

Spinal Cord Toolbox

Spinal Cord Toolbox (SCT) is a comprehensive, free and open-source set of command-line tools dedicated to the processing and analysis of spinal cord MRI data.

Key Features

For a thorough introduction to SCT's features, please refer to our 2024 Review "Reproducible Spinal Cord Quantitative MRI Analysis with the Spinal Cord Toolbox".

In summary:

• Segmentation of the spinal cord and gray matter
• Segmentation of pathologies (e.g. multiple sclerosis lesions)
• Detection of anatomical highlights (e.g. ponto-medullary junction, spinal cord centerline, vertebral levels)
• Registration to template, and deformation (e.g. straightening)
• Motion correction for diffusion and functional MRI time series
• Computation of quantitative MRI metrics (e.g. diffusion tensor imaging, magnetization transfer)
• Texture analysis (e.g. grey level co-occurrence matrix)
• Extraction of metrics within anatomical regions (e.g. white matter tracts)
• Manual labeling and segmentation via a Graphical User Interface (GUI)
• Warping field creation and application
• NIFTI volume manipulation tools for common operations

Installation

For macOS and Linux users, the simplest way to install SCT is to download the latest release, then launch the install script:

./install_sct

For more complex installation setups (Windows users, Docker, FSLeyes integration), see the Installation page.

Usage

Once installed, there are three main ways to use SCT:

1. Command-line tools

The primary way to invoke SCT is through terminal commands. For example:

$ sct_deepseg_sc -i t2.nii.gz -c t2

Cropping the image around the spinal cord...
Normalizing the intensity...
Segmenting the spinal cord using deep learning on 2D patches...
Reassembling the image...
Resampling the segmentation to the native image resolution using linear interpolation...
Binarizing the resampled segmentation...
Compute shape analysis: 100%|################| 55/55 [00:00<00:00, 106.05iter/s]

Done! To view results, type:
fsleyes t2.nii.gz -cm greyscale t2_seg.nii.gz -cm red -a 70.0 &

For a full overview of the available commands, see the Command-Line Tools page.

2. Multi-command pipelines

To facilitate multi-subject analyses, commands can be chained together to build processing pipelines. The best starting point for constructing a typical pipeline is the batch_processing.sh script, which is provided with your installation of SCT.

3. GUI (FSLeyes integration)

SCT provides a graphical user interface via a FSLeyes plugin. For more details, see the FSLeyes Integration page.

Who is using SCT?

SCT is trusted by the research labs of many highly-regarded institutions worldwide. A full list of endorsements can be found on the Testimonials page.

For a list of neuroimaging studies that depend on SCT, visit the Studies using SCT page.

Contact

If you have any questions or concerns, and would like to get in touch with the developers of Spinal Cord Toolbox, the best way to do so is via the Spinalcordmri.org forum.

If you would like to stay up to date on spinal cord workshops and SCT announcements (new releases, hands-on courses, etc.), please subscribe to the Spinalcordmri.org mailing list.

License

SCT is made available under the LGPLv3 license. For more details, see LICENSE.

Contributing

We happily welcome contributions. Please see the Contributing page of the developer Wiki for more information.

Warning

️Medical Disclaimer

All content found in the Spinal Cord Toolbox repository and spinalcordtoolbox.com website, including: text, images, audio, or other formats were created for informational purposes only. The content is not intended to be a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of your physician or other qualified health provider with any questions you may have regarding a medical condition. Never disregard professional medical advice or delay in seeking it because of something you have read on this website.

If you think you may have a medical emergency, call your doctor, go to the emergency department, or call your local emergency number immediately. Spinal Cord Toolbox does not recommend or endorse any specific tests, physicians, products, procedures, opinions, or other information that may be mentioned on spinalcordtoolbox.com. Reliance on any information provided by spinalcordtoolbox.com, Spinal Cord Toolbox contributors, contracted writers, or medical professionals presenting content for publication to spinalcordtoolbox.com is solely at your own risk.

spinalcordtoolbox's Issues

sct_nurbs: find alternative for robust smoothing

current issues are:

• non-robust (does not handle outliers)
• requires end points (need extrapolation if larger grid is asked)
• crashes sometimes
• smoothness depends on number of data points.

possible avenues:

• nonparametric fitting (https://gist.github.com/agramfort/850437)
• moving average (http://www.swharden.com/blog/2010-06-24-detrending-data-in-python-with-numpy/) --> modify it to make it a robust mean
• smooth (http://www2.geog.ucl.ac.uk/~plewis/geogg122_local/geogg122/Chapter5_Interpolation/python/smoothn.py)
• bspline: http://elbenshira.com/p/curvey/

sct_get_centerline: output should be in same space as input (currently RPI)

need smooth deformation to keep consistent internal structure

It seems like the internal structure of the T2 template might produce spurious deformation fields as shown below.

compact VM

currently 3.7GB

• home: 3.3GB --> due to trash. After deleting it it was ok.
• /lib: 127MB
• /usr: 2.2GB
  • /usr/share: 885MB --> used by fsl. Data from fsl ~800MB.
  • /usr/local: 339MB --> used by spinal cord toolbox
  • /usr/lib: 826MB --> OS files
• /var : 500MB --> OS files

So unfortunately these things do not seem compressible...

However, the ova from neurodebian is only 700MB with all the OS files. So that needs further investigations...

moco for fmri

regularization along z - elastic deformation

useful links:

• https://sourceforge.net/p/advants/discussion/840261/thread/2dc20aa2/?limit=25#c26a

clean sct_compute_mtr

sct_flatten_spinalcord.py not working for labelized images

Uncorrect flattened images with labelized images (vertebral levels).

Propseg: WARNING: In /usr/local/ITK-static/include/ITK-4.5/itkGaussianOperator.hxx

when running propseg on the MNI-Poly-AMU t2 template:

command ran:
sct_segmentation_propagation -i MNI-Poly-AMU_T2.nii.gz -t t2 -mesh

'''''
WARNING: In /usr/local/ITK-static/include/ITK-4.5/itkGaussianOperator.hxx, line 61
itkGaussianOperator (0x7fff5b50f0f8): Kernel size has exceeded the specified maximum width of 32 and has been truncated to 33 elements. You can raise the maximum width using the SetMaximumKernelWidth method.
'''''

sct_straighten_spinalcord: if user inputs non-binary segmentation image, the centerline is badly estimated

currently the centerline is estimated via two possible scenario, which is suboptimal. Moreover it does not work if user inputs non-binary centerline (lack accuracy). See image below:

Now we are estimating the center of mass using scipy function. This now works for all cases. See result below using center of mass:

sct_register_multimodal

2014-06-11, 16.38
pass on julien's laptop,
fail on evans

�[1;32mSubject: errsm_23�[0m
�[1;31mContrast: mt�[0m

==============================================================================================
sct_register_multimodal.py
            -i /home/django/jcohen2/code/spinalcordtoolbox/data/template/MNI-Poly-AMU_T2.nii.gz
            -d /home/django/jcohen2/code/spinalcordtoolbox/testing/data/errsm_23/mt/mtc1.nii.gz
            -s /home/django/jcohen2/code/spinalcordtoolbox/data/template/MNI-Poly-AMU_cord.nii.gz
            -t /home/django/jcohen2/code/spinalcordtoolbox/testing/data/errsm_23/mt/segmentation_binary.nii.gz
            -q /home/django/jcohen2/code/spinalcordtoolbox/testing/data/errsm_23/template/warp_template2anat.nii.gz
            -x 1
            -z /home/django/jcohen2/code/spinalcordtoolbox/testing/data/errsm_23/template/warp_anat2template.nii.gz
            -o template2mtc1.nii.gz
            -n 50x20
            -r 1
            -v 1
==============================================================================================
/home/django/jcohen2/code/spinalcordtoolbox

Check parameters:
.. Source:               /home/django/jcohen2/code/spinalcordtoolbox/data/template/MNI-Poly-AMU_T2.nii.gz
.. Destination:          /home/django/jcohen2/code/spinalcordtoolbox/testing/data/errsm_23/mt/mtc1.nii.gz
.. Segmentation source:  /home/django/jcohen2/code/spinalcordtoolbox/data/template/MNI-Poly-AMU_cord.nii.gz
.. Segmentation dest:    /home/django/jcohen2/code/spinalcordtoolbox/testing/data/errsm_23/mt/segmentation_binary.nii.gz
.. Init transfo:         /home/django/jcohen2/code/spinalcordtoolbox/testing/data/errsm_23/template/warp_template2anat.nii.gz
.. Output name:          template2mtc1.nii.gz
.. number of iterations: 50x20
.. Verbose:              1
.. Remove temp files:    1

Create local temp files...
>> c3d /home/django/jcohen2/code/spinalcordtoolbox/data/template/MNI-Poly-AMU_T2.nii.gz -o tmp.src.nii
>> c3d /home/django/jcohen2/code/spinalcordtoolbox/testing/data/errsm_23/mt/mtc1.nii.gz -o tmp.dest.nii
>> c3d /home/django/jcohen2/code/spinalcordtoolbox/data/template/MNI-Poly-AMU_cord.nii.gz -o tmp.src_seg.nii
>> c3d /home/django/jcohen2/code/spinalcordtoolbox/testing/data/errsm_23/mt/segmentation_binary.nii.gz -o tmp.dest_seg.nii

Apply initial transformation to moving image...
>> WarpImageMultiTransform 3 tmp.src.nii tmp.src_reg.nii -R tmp.dest.nii /home/django/jcohen2/code/spinalcordtoolbox/testing/data/errsm_23/template/warp_template2anat.nii.gz --use-BSpline
>> WarpImageMultiTransform 3 tmp.src_seg.nii tmp.src_seg_reg.nii -R tmp.dest_seg.nii /home/django/jcohen2/code/spinalcordtoolbox/testing/data/errsm_23/template/warp_template2anat.nii.gz --use-BSpline

Pad source...
>> c3d tmp.src_reg -pad 0x0x3vox 0x0x3vox 0 -o tmp.src_reg_pad.nii

Pad destination...
>> c3d tmp.dest -pad 0x0x3vox 0x0x3vox 0 -o tmp.dest_pad.nii

Pad source segmentation...
>> c3d tmp.src_seg_reg -pad 0x0x3vox 0x0x3vox 0 -o tmp.src_seg_reg_pad.nii

Pad destination segmentation...
>> c3d tmp.dest_seg -pad 0x0x3vox 0x0x3vox 0 -o tmp.dest_seg_pad.nii

Step #1: Estimate transformation using spinal cord segmentations...
>> antsRegistration --dimensionality 3 --transform SyN[0.5,3,0] --metric MI[tmp.dest_seg_pad.nii,tmp.src_seg_reg_pad.nii,1,32] --convergence 50x20 --shrink-factors 4x1 --smoothing-sigmas 1x1mm --Restrict-Deformation 1x1x0 --output [tmp.regSeg,tmp.regSeg.nii]
Using double precision for computations.
  number of levels = 2
  fixed image: tmp.dest_seg_pad.nii
  moving image: tmp.src_seg_reg_pad.nii
Dimension = 3
Number of stages = 1
Use Histogram Matching false
Winsorize Image Intensities false
Lower Quantile = 0
Upper Quantile = 1
Stage 1 State
   Metric = Mattes
     Fixed Image = Image (0x7fd6e190bb80)
  RTTI typeinfo:   itk::Image<double, 3u>
  Reference Count: 2
  Modified Time: 559
  Debug: Off
  Object Name: 
  Observers: 
    none
  Source: (none)
  Source output name: (none)
  Release Data: Off
  Data Released: False
  Global Release Data: Off
  PipelineMTime: 0
  UpdateMTime: 396
  RealTimeStamp: 0 seconds 
  LargestPossibleRegion: 
    Dimension: 3
    Index: [0, 0, 0]
    Size: [192, 192, 11]
  BufferedRegion: 
    Dimension: 3
    Index: [0, 0, 0]
    Size: [192, 192, 11]
  RequestedRegion: 
    Dimension: 3
    Index: [0, 0, 0]
    Size: [192, 192, 11]
  Spacing: [0.84375, 0.84375, 16.5]
  Origin: [-81, 117.424, -37.3294]
  Direction: 
1 0 0
0 -0.992757 -0.120137
0 -0.120137 0.992757

  IndexToPointMatrix: 
0.84375 0 0
0 -0.837639 -1.98226
0 -0.101366 16.3805

  PointToIndexMatrix: 
1.18519 0 0
0 -1.1766 -0.142384
0 -0.00728103 0.0601671

  Inverse Direction: 
1 0 0
0 -0.992757 -0.120137
0 -0.120137 0.992757

  PixelContainer: 
    ImportImageContainer (0x7fd6e190baa0)
      RTTI typeinfo:   itk::ImportImageContainer<unsigned long, double>
      Reference Count: 1
      Modified Time: 393
      Debug: Off
      Object Name: 
      Observers: 
        none
      Pointer: 0x108c2d000
      Container manages memory: true
      Size: 405504
      Capacity: 405504

     Moving Image = Image (0x7fd6e190b790)
  RTTI typeinfo:   itk::Image<double, 3u>
  Reference Count: 2
  Modified Time: 560
  Debug: Off
  Object Name: 
  Observers: 
    none
  Source: (none)
  Source output name: (none)
  Release Data: Off
  Data Released: False
  Global Release Data: Off
  PipelineMTime: 0
  UpdateMTime: 557
  RealTimeStamp: 0 seconds 
  LargestPossibleRegion: 
    Dimension: 3
    Index: [0, 0, 0]
    Size: [192, 192, 11]
  BufferedRegion: 
    Dimension: 3
    Index: [0, 0, 0]
    Size: [192, 192, 11]
  RequestedRegion: 
    Dimension: 3
    Index: [0, 0, 0]
    Size: [192, 192, 11]
  Spacing: [0.84375, 0.84375, 16.5]
  Origin: [-81, 117.424, -37.3294]
  Direction: 
1 0 0
0 -0.992757 -0.120137
0 -0.120137 0.992757

  IndexToPointMatrix: 
0.84375 0 0
0 -0.837639 -1.98226
0 -0.101366 16.3805

  PointToIndexMatrix: 
1.18519 0 0
0 -1.1766 -0.142384
0 -0.00728103 0.0601671

  Inverse Direction: 
1 0 0
0 -0.992757 -0.120137
0 -0.120137 0.992757

  PixelContainer: 
    ImportImageContainer (0x7fd6e190fb90)
      RTTI typeinfo:   itk::ImportImageContainer<unsigned long, double>
      Reference Count: 1
      Modified Time: 554
      Debug: Off
      Object Name: 
      Observers: 
        none
      Pointer: 0x10925e000
      Container manages memory: true
      Size: 405504
      Capacity: 405504

     Weighting = 1
     Sampling Strategy = none
     NumberOfBins = 32
     Radius = 1
     Sampling percentage  = 1
   Transform = SyN
     Gradient Step = 0.5
     Update Field Sigma (voxel space) = 3
     Total Field Sigma (voxel space) = 0
     Update Field Time Sigma = 0
     Total Field Time Sigma  = 0
     Number of Time Indices = 0
     Number of Time Point Samples = 0
Registration using 1 total stages.

Stage 0
  iterations = 50x20
  convergence threshold = 1e-06
  convergence window size = 10
  number of levels = 2
  Shrink factors (level 1 out of 2): [4, 4, 1]
  Shrink factors (level 2 out of 2): [1, 1, 1]
  smoothing sigmas per level: [1, 1]
  using the Mattes MI metric (number of bins = 32, weight = 1)
  Using default NONE metricSamplingStrategy 

*** Running SyN registration (varianceForUpdateField = 3, varianceForTotalField = 0) ***

  Current level = 1 of 2
    number of iterations = 50
    shrink factors = [4, 4, 1]
    smoothing sigmas = 1 mm
    required fixed parameters = [48, 48, 11, -79.734375, 116.16756798783919, -37.481424485533395, 3.375, 3.375, 16.499998092651367, 1, 0, 0, 0, -0.9927573344542117, -0.12013689654985334, 0, -0.12013690060663551, 0.9927573349451364]
XXDIAGNOSTIC,Iteration,metricValue,convergenceValue,ITERATION_TIME_INDEX,SINCE_LAST
 1DIAGNOSTIC,     1, -1.279534399088e-02, 1.797693134862e+308, 1.2522e-01, 1.2518e-01, 
 1DIAGNOSTIC,     2, -1.374867870519e-02, 1.797693134862e+308, 1.9391e-01, 6.8693e-02, 
 1DIAGNOSTIC,     3, -1.495346809446e-02, 1.797693134862e+308, 2.6098e-01, 6.7064e-02, 
 1DIAGNOSTIC,     4, -1.423497312352e-02, 1.797693134862e+308, 3.2806e-01, 6.7081e-02, 
 1DIAGNOSTIC,     5, -1.362644975415e-02, 1.797693134862e+308, 3.9576e-01, 6.7702e-02, 
 1DIAGNOSTIC,     6, -1.273267177939e-02, 1.797693134862e+308, 4.6333e-01, 6.7569e-02, 
 1DIAGNOSTIC,     7, -1.244140922900e-02, 1.797693134862e+308, 5.3102e-01, 6.7692e-02, 
 1DIAGNOSTIC,     8, -1.212294379761e-02, 1.797693134862e+308, 5.9814e-01, 6.7124e-02, 
 1DIAGNOSTIC,     9, -1.192206125885e-02, 1.797693134862e+308, 6.6552e-01, 6.7379e-02, 
 1DIAGNOSTIC,    10, -1.183634862858e-02, -1.204498820754e-02, 7.3376e-01, 6.8236e-02, 
  Current level = 2 of 2
    number of iterations = 20
    shrink factors = [1, 1, 1]
    smoothing sigmas = 1.0000e+00 mm
    required fixed parameters = [192, 192, 11, -81, 117.42402648925781, -37.329376220703125, 0.84375, 0.84375, 16.499998092651367, 1, 0, 0, 0, -0.9927573344542117, -0.12013689654985334, 0, -0.12013690060663551, 0.9927573349451364]
XXDIAGNOSTIC,Iteration,metricValue,convergenceValue,ITERATION_TIME_INDEX,SINCE_LAST
 1DIAGNOSTIC,     1, -1.188294427336e-02, 1.797693134862e+308, 1.7782e+00, 1.0442e+00, 
 1DIAGNOSTIC,     2, -1.204289498118e-02, 1.797693134862e+308, 2.6767e+00, 8.9846e-01, 
 1DIAGNOSTIC,     3, -1.212765121139e-02, 1.797693134862e+308, 3.5633e+00, 8.8665e-01, 
 1DIAGNOSTIC,     4, -1.178380170855e-02, 1.797693134862e+308, 4.4626e+00, 8.9924e-01, 
 1DIAGNOSTIC,     5, -1.148477022465e-02, 1.797693134862e+308, 5.3465e+00, 8.8390e-01, 
 1DIAGNOSTIC,     6, -1.122054959059e-02, 1.797693134862e+308, 6.2338e+00, 8.8731e-01, 
 1DIAGNOSTIC,     7, -1.081726229362e-02, 1.797693134862e+308, 7.0880e+00, 8.5420e-01, 
 1DIAGNOSTIC,     8, -1.045450851249e-02, 1.797693134862e+308, 7.9820e+00, 8.9406e-01, 
 1DIAGNOSTIC,     9, -1.017071193630e-02, 1.797693134862e+308, 8.8457e+00, 8.6368e-01, 
 1DIAGNOSTIC,    10, -9.903069106736e-03, -1.342427600251e-02, 9.7246e+00, 8.7894e-01, 
  Elapsed time (stage 0): 9.8222e+00


Total elapsed time: 9.8376e+00

Step #2: Improve local deformation using images (start from previous transformation)...
>> antsRegistration --dimensionality 3 --initial-moving-transform tmp.regSeg0Warp.nii.gz --transform SyN[0.1,1,0] --metric MI[tmp.dest_pad.nii,tmp.src_reg_pad.nii,1,32] --convergence 20 --shrink-factors 1 --smoothing-sigmas 0mm --Restrict-Deformation 1x1x0 --output [tmp.reg,tmp.src_reg_pad_reg.nii] --collapse-output-transforms 0 --interpolation BSpline[3]
Using double precision for computations.
=============================================================================
The composite transform is comprised of the following transforms (in order): 
  1. tmp.regSeg0Warp.nii.gz (type = DisplacementFieldTransform)
=============================================================================
Can't write transform file tmp.reg0DerivedInitialMovingTranslation.mat
Exception Object caught: 

itk::ImageFileWriterException (0x7fa83991a998)
Location: "unknown" 
File: /home/django/jcohen2/antsbin/ITKv4-install/include/ITK-4.5/itkImageFileWriter.hxx
Line: 151
Description:  Could not create IO object for file tmp.reg0DerivedInitialMovingTranslation.mat
  Tried to create one of the following:
    NiftiImageIO
    NrrdImageIO
    GiplImageIO
    HDF5ImageIO
    JPEGImageIO
    GDCMImageIO
    BMPImageIO
    LSMImageIO
    PNGImageIO
    TIFFImageIO
    VTKImageIO
    StimulateImageIO
    BioRadImageIO
    MetaImageIO
    MRCImageIO
  You probably failed to set a file suffix, or
    set the suffix to an unsupported type.



  number of levels = 1
  fixed image: tmp.dest_pad.nii
  moving image: tmp.src_reg_pad.nii
Dimension = 3
Number of stages = 1
Use Histogram Matching false
Winsorize Image Intensities false
Lower Quantile = 0
Upper Quantile = 1
Stage 1 State
   Metric = Mattes
     Fixed Image = Image (0x7fa83991eec0)
  RTTI typeinfo:   itk::Image<double, 3u>
  Reference Count: 2
  Modified Time: 1488
  Debug: Off
  Object Name: 
  Observers: 
    none
  Source: (none)
  Source output name: (none)
  Release Data: Off
  Data Released: False
  Global Release Data: Off
  PipelineMTime: 0
  UpdateMTime: 1325
  RealTimeStamp: 0 seconds 
  LargestPossibleRegion: 
    Dimension: 3
    Index: [0, 0, 0]
    Size: [192, 192, 11]
  BufferedRegion: 
    Dimension: 3
    Index: [0, 0, 0]
    Size: [192, 192, 11]
  RequestedRegion: 
    Dimension: 3
    Index: [0, 0, 0]
    Size: [192, 192, 11]
  Spacing: [0.84375, 0.84375, 16.5]
  Origin: [-81, 117.424, -37.3294]
  Direction: 
1 0 0
0 -0.992757 -0.120137
0 -0.120137 0.992757

  IndexToPointMatrix: 
0.84375 0 0
0 -0.837639 -1.98226
0 -0.101366 16.3805

  PointToIndexMatrix: 
1.18519 0 0
0 -1.1766 -0.142384
0 -0.00728103 0.0601671

  Inverse Direction: 
1 0 0
0 -0.992757 -0.120137
0 -0.120137 0.992757

  PixelContainer: 
    ImportImageContainer (0x7fa83991e140)
      RTTI typeinfo:   itk::ImportImageContainer<unsigned long, double>
      Reference Count: 1
      Modified Time: 1322
      Debug: Off
      Object Name: 
      Observers: 
        none
      Pointer: 0x104077000
      Container manages memory: true
      Size: 405504
      Capacity: 405504

     Moving Image = Image (0x7fa83991aa60)
  RTTI typeinfo:   itk::Image<double, 3u>
  Reference Count: 2
  Modified Time: 1489
  Debug: Off
  Object Name: 
  Observers: 
    none
  Source: (none)
  Source output name: (none)
  Release Data: Off
  Data Released: False
  Global Release Data: Off
  PipelineMTime: 0
  UpdateMTime: 1486
  RealTimeStamp: 0 seconds 
  LargestPossibleRegion: 
    Dimension: 3
    Index: [0, 0, 0]
    Size: [192, 192, 11]
  BufferedRegion: 
    Dimension: 3
    Index: [0, 0, 0]
    Size: [192, 192, 11]
  RequestedRegion: 
    Dimension: 3
    Index: [0, 0, 0]
    Size: [192, 192, 11]
  Spacing: [0.84375, 0.84375, 16.5]
  Origin: [-81, 117.424, -37.3294]
  Direction: 
1 0 0
0 -0.992757 -0.120137
0 -0.120137 0.992757

  IndexToPointMatrix: 
0.84375 0 0
0 -0.837639 -1.98226
0 -0.101366 16.3805

  PointToIndexMatrix: 
1.18519 0 0
0 -1.1766 -0.142384
0 -0.00728103 0.0601671

  Inverse Direction: 
1 0 0
0 -0.992757 -0.120137
0 -0.120137 0.992757

  PixelContainer: 
    ImportImageContainer (0x7fa83991c7e0)
      RTTI typeinfo:   itk::ImportImageContainer<unsigned long, double>
      Reference Count: 1
      Modified Time: 1483
      Debug: Off
      Object Name: 
      Observers: 
        none
      Pointer: 0x1046a7000
      Container manages memory: true
      Size: 405504
      Capacity: 405504

     Weighting = 1
     Sampling Strategy = none
     NumberOfBins = 32
     Radius = 1
     Sampling percentage  = 1
   Transform = SyN
     Gradient Step = 0.1
     Update Field Sigma (voxel space) = 1
     Total Field Sigma (voxel space) = 0
     Update Field Time Sigma = 0
     Total Field Time Sigma  = 0
     Number of Time Indices = 0
     Number of Time Point Samples = 0
Registration using 1 total stages.

Stage 0
  iterations = 20
  convergence threshold = 1e-06
  convergence window size = 10
  number of levels = 1
  Shrink factors (level 1 out of 1): [1, 1, 1]
  smoothing sigmas per level: [0]
  using the Mattes MI metric (number of bins = 32, weight = 1)
  Using default NONE metricSamplingStrategy 

*** Running SyN registration (varianceForUpdateField = 1, varianceForTotalField = 0) ***

  Current level = 1 of 1
    number of iterations = 20
    shrink factors = [1, 1, 1]
    smoothing sigmas = 0 mm
    required fixed parameters = [192, 192, 11, -81, 117.42402648925781, -37.329376220703125, 0.84375, 0.84375, 16.499998092651367, 1, 0, 0, 0, -0.9927573344542117, -0.12013688942932219, 0, -0.12013690060663551, 0.9927573358068157]
XXDIAGNOSTIC,Iteration,metricValue,convergenceValue,ITERATION_TIME_INDEX,SINCE_LAST
 1DIAGNOSTIC,     1, -7.405185984578e-02, 1.797693134862e+308, 8.4241e-01, 8.4238e-01, 
 1DIAGNOSTIC,     2, -7.416685794359e-02, 1.797693134862e+308, 1.6306e+00, 7.8823e-01, 
 1DIAGNOSTIC,     3, -7.429667691027e-02, 1.797693134862e+308, 2.4292e+00, 7.9856e-01, 
 1DIAGNOSTIC,     4, -7.443575182016e-02, 1.797693134862e+308, 3.2509e+00, 8.2173e-01, 
 1DIAGNOSTIC,     5, -7.453851632370e-02, 1.797693134862e+308, 4.0509e+00, 7.9996e-01, 
 1DIAGNOSTIC,     6, -7.462891831747e-02, 1.797693134862e+308, 4.8645e+00, 8.1358e-01, 
 1DIAGNOSTIC,     7, -7.467508336003e-02, 1.797693134862e+308, 5.6534e+00, 7.8892e-01, 
 1DIAGNOSTIC,     8, -7.470231353181e-02, 1.797693134862e+308, 6.4346e+00, 7.8122e-01, 
 1DIAGNOSTIC,     9, -7.477296301922e-02, 1.797693134862e+308, 7.2420e+00, 8.0737e-01, 
 1DIAGNOSTIC,    10, -7.481694391767e-02, 6.511445570990e-04, 8.0645e+00, 8.2253e-01, 
 1DIAGNOSTIC,    11, -7.482867220136e-02, 4.985201012338e-04, 8.8669e+00, 8.0243e-01, 
 1DIAGNOSTIC,    12, -7.490023421227e-02, 3.859223037805e-04, 9.6761e+00, 8.0917e-01, 
 1DIAGNOSTIC,    13, -7.493715113159e-02, 3.037841859762e-04, 1.0493e+01, 8.1642e-01, 
 1DIAGNOSTIC,    14, -7.500837784343e-02, 2.623704060583e-04, 1.1312e+01, 8.1950e-01, 
 1DIAGNOSTIC,    15, -7.499624945127e-02, 2.248124841403e-04, 1.2135e+01, 8.2301e-01, 
 1DIAGNOSTIC,    16, -7.506446438221e-02, 2.094209181930e-04, 1.2946e+01, 8.1082e-01, 
 1DIAGNOSTIC,    17, -7.504515794096e-02, 1.822666293318e-04, 1.3760e+01, 8.1441e-01, 
 1DIAGNOSTIC,    18, -7.510618672287e-02, 1.597023041366e-04, 1.4573e+01, 8.1294e-01, 
 1DIAGNOSTIC,    19, -7.513799412739e-02, 1.448281665660e-04, 1.5380e+01, 8.0657e-01, 
 1DIAGNOSTIC,    20, -7.515362088850e-02, 1.294023752298e-04, 1.6188e+01, 8.0837e-01, 
  Elapsed time (stage 0): 1.6289e+01


Total elapsed time: 1.6304e+01

Concatenate transformations...
>> ComposeMultiTransform 3 tmp.warp_src2dest.nii.gz -R tmp.dest.nii tmp.reg1Warp.nii.gz tmp.regSeg0Warp.nii.gz /home/django/jcohen2/code/spinalcordtoolbox/testing/data/errsm_23/template/warp_template2anat.nii.gz
>> ComposeMultiTransform 3 tmp.warp_dest2src.nii.gz -R tmp.src.nii /home/django/jcohen2/code/spinalcordtoolbox/testing/data/errsm_23/template/warp_anat2template.nii.gz tmp.regSeg0InverseWarp.nii.gz tmp.reg1InverseWarp.nii.gz

Apply transfo source --> dest...
>> WarpImageMultiTransform 3 tmp.src.nii tmp.src_reg.nii -R tmp.dest.nii tmp.warp_src2dest.nii.gz --use-BSpline

Apply transfo dest --> source...
>> WarpImageMultiTransform 3 tmp.dest.nii tmp.dest_reg.nii -R tmp.src.nii tmp.warp_dest2src.nii.gz --use-BSpline

ERROR!!! 
FIELD: tmp.warp_dest2src.nii.gz
moving_image_filename: tmp.dest.nii components 1
output_image_filename: tmp.dest_reg.nii
reference_image_filename: tmp.src.nii
[0/1]: FIELD: tmp.warp_dest2src.nii.gz
 Not currently supported because of a lack of vector support 
Exception caught during WarpImageMultiTransform.

itk::ImageFileReaderException (0x7f8b04912688)
Location: "unknown" 
File: /home/django/jcohen2/antsbin/ITKv4-install/include/ITK-4.5/itkImageFileReader.hxx
Line: 143
Description:  Could not create IO object for file tmp.warp_dest2src.nii.gz
The file doesn't exist. 
Filename = tmp.warp_dest2src.nii.gz



Exit program.

sct_register_multimodal.py: create a tmp.XX folder for all temp files

sct_get_centerline: should not output in rpi but in native orientation

j_detrend.m not working

Not working. Should be removed. Eugenie created a new one (j_detrend_new_v2.m), but doesn't seem to work neither...

sct_segmentation_propagation: output segmentation with trilinear interpolation

This will be important for the registration to the template, when using the segmentation, because the binary segmentation has "step" effects when straightened (see picture below).

check if flirt is declared and THEN scan the whole HD

sct_warp_atlas_to_metric: also warp spinal levels

installation of patch 1.0.1 fails on Debian

see message Falk on SF

test

clean sct_compute_tsnr.sh

sct_crop: new function to crop an image by clicking on it

using:

• pil
• tcl/tk

sct_smooth_spinalcord.py: remove the need to have all slices in the input segmentation/centerline

currenty, if not all slices are covered, the following error appears:

ERROR: centerline/segmentation must cover all "z" slices of the input image.
To solve the problem, you need to crop the input image (you can use 'sct_crop_image') and generate onemore time the spinal cord centerline/segmentation from this cropped image.

sct_dmri_moco: for sinc interp, use ANTs or c3d instead of flirt

error in ANTSUseLandmarkImagesToGetAffineTransform for Estimate transfo: straight anat --> template (affine landmark-based)

data_mounir/spine_50/t2

sct_register_to_template.py -i t2_crop.nii.gz -l landmarks_rpi_crop.nii.gz -m segmentation_binary_crop.nii.gz -s superfast -r 0

Registration of straight spinal cord to template...
>> sct_register_straight_spinalcord_to_template.py -i data_rpi_straight.nii.gz -l landmarks_rpi_cross3x3_straight.nii.gz -t /usr/local/spinalcordtoolbox/data/template/MNI-Poly-AMU_T2.nii.gz -f template_label_cross.nii.gz -m /usr/local/spinalcordtoolbox/data/template/mask_gaussian_templatespace_sigma20.nii.gz -r 1 -n 3x1 -v 1

ERROR!!!

Check input arguments:
  straight anatomic:    data_rpi_straight.nii.gz
  landmarks anatomic:   landmarks_rpi_cross3x3_straight.nii.gz
  template:             /usr/local/spinalcordtoolbox/data/template/MNI-Poly-AMU_T2.nii.gz
  landmarks template:   template_label_cross.nii.gz
  number of iterations: 3x1
  mask anatomic:        /usr/local/spinalcordtoolbox/data/template/mask_gaussian_templatespace_sigma20.nii.gz
  Verbose:              1

Estimate transfo: straight anat --> template (affine landmark-based)...
>> ANTSUseLandmarkImagesToGetAffineTransform template_label_cross.nii.gz landmarks_rpi_cross3x3_straight.nii.gz affine tmp.straight2templateAffine.txt

ERROR!!!
 fix-label 21 movlabel 21
 fix-label 22 movlabel 22
 fix-label 23 movlabel 23
 fix-label 24 movlabel 24
 labels do not match -- exiting
Exit program.

sct_dmri_moco: create simulator of motion

sct_crop_image

two features to be added:

• specify if -dim flag starts at 0 or 1
• enable to mask cropped region, without having to provide a mask.

INST: testing doesn't work ...

After installation, the testing didn't worked because image were not found (sct_segmentation_propagation).

The sct_testing folder created during the installation should be created without sudo permission so the owner is not root.

sct_detect_vertebraldisks: new function to detect vertebral disks and levels

based on Eugenie work

source .bash_profile does not load binaries

to try on osx:

. .bash_profile

instead of

source .bash_profile

TEST: testing/sct_register_straight_spinalcord_to_template does not work

have to feed with appropriate data

adjust number of iterations (slow, fast...)

sct_smooth_spinalcord and sct_straighten_spinalcord

crashes when deleting tmp.* files, if there is a tmp.XX folder present.

Delete temporary files...
>> rm tmp.*

ERROR!!! 
rm: tmp.140705145131: is a directory
Exit program.

Need to make proper tmp folder for both functions.

sct_register_to_template: multiple versions of the file data_rpi_straight exist

a version .nii and .nii.gz

problem happens on:

• VM
• OSX

check if SCT in bash_profile by inspecting $PATH instead of inspecting bash_profile

Spinal cord length extraction in sct_process_segmentation

Extraction of spinal cord length between two user-specified levels. Must be coupled with vertebral/spinal level identification.
Example of use:
sct_process_segmentation.py -i segmentation_binary.nii.gz -p length -m map_levels.nii.gz -l C1-T12

Use partial volume interpolation for warping labels

use c3d -warp option, which enables that. See 2nd example of -warp here:
http://www.itksnap.org/pmwiki/pmwiki.php?n=Convert3D.Documentation

sct_dmri_moco.py -bval fails

error:

UnboundLocalError: local variable 'bvecs_new' referenced before assignment

sct_dmri_moco: general modifs

• at the beginning of the code, please copy input data into tmp folder for better debugging
• in tmp folder, please
  • call every mat folder like: mat_xxx
  • remove "tmp.param" prefix
• output: remove suffix "final"

VM: sct_register_to_template fails

output is this:

sct_check_dependencies: if one fail, indicate that there is a problem

currently, if ants integrity or propseg fails, it says "everything in order"

binaries for ANTs, c3d and propseg not compatible with OSX Lion

sourceforge issue "Intalling toolbox"

VM NeuroDebian

• cannot import appliance. Error: "Implementation of the USB 2.0 controller not found!"
  • made it working by unchecking "unable USB 2.0"
• webbrowser should have SF website in favs
• webbrowser: currently opens with google in German (???)
• current shared folder settings is: "C:\Users\benja_000\VM_folder_shared"

add sct_dmri_moco

sct_dmri_moco: add eddy correction with fitting

testing data: errsm_03, volumes 63 & 64.

PropSeg failure... ITK Singular matrix at the beginning

An image with lots of zero voxels (entire axial slices full of zeros) doesn't work on PropSeg.
I gave the following error directely at the beginning of the segmentation:

libc++abi.dylib: terminating with uncaught exception of type itk::ExceptionObject: /usr/local/ITKVTK/include/ITK-4.4/itkMatrix.h:240:
itk::ERROR: Singular matrix. Determinant is 0.

The error is probably in the symmetry detection. Need to check that.

PropSeg: error for some image in spinal cord detection module

PropSeg produces a segmentation fault sometimes when detecting the spinal cord for images with particular field of view and resolution (not isotropic).

The computation of number of slices and gap parameters used for spinal cord detection should be revised.

add coordinate system

make sure all python files have chmod 775 to be executable

sct_create_package: do not add the third number for the package name (only X.X)

rename all ANTS and c3d binaries with prefix "isct_", to avoid conflicts