Multiphase Chan and Vese Sparse Field Level Set Segmentation#

Note

Wish List Still needs additional work to finish proper creation of example.

Synopsis#

Multiphase Chan And Vese Sparse Field Level Set Segmentation.

Results#

Note

Help Wanted Implementation of Results for sphinx examples containing this message. Reconfiguration of CMakeList.txt may be necessary. Write An Example <https://itk.org/ITKExamples/Documentation/Contribute/WriteANewExample.html>

Code#

C++#

#include "itkScalarChanAndVeseSparseLevelSetImageFilter.h"
#include "itkScalarChanAndVeseLevelSetFunction.h"
#include "itkScalarChanAndVeseLevelSetFunctionData.h"
#include "itkConstrainedRegionBasedLevelSetFunctionSharedData.h"
#include "itkImageFileReader.h"
#include "itkImageFileWriter.h"
#include "itkImage.h"
#include "itkAtanRegularizedHeavisideStepFunction.h"

int
main(int argc, char ** argv)
{
  if (argc < 10)
  {
    std::cerr << "Missing arguments" << std::endl;
    std::cerr << "Usage: " << std::endl;
    std::cerr << argv[0];
    std::cerr << " inputLevelSetImage1 inputLevelSetImage2 inputLevelSetImage3";
    std::cerr << " inputFeatureImage outputLevelSetImage";
    std::cerr << " CurvatureWeight AreaWeight LaplacianWeight";
    std::cerr << " VolumeWeight Volume OverlapWeight" << std::endl;
    return EXIT_FAILURE;
  }

  unsigned int nb_iteration = 50;
  double       rms = 0.;
  double       epsilon = 1.5;
  double       curvature_weight = std::stod(argv[6]);
  double       area_weight = std::stod(argv[7]);
  double       volume_weight = std::stod(argv[8]);
  double       volume = std::stod(argv[9]);
  double       overlap_weight = std::stod(argv[10]);
  double       l1 = 1.;
  double       l2 = 1.;

  constexpr unsigned int Dimension = 2;
  using ScalarPixelType = float;

  using LevelSetImageType = itk::Image<ScalarPixelType, Dimension>;
  using FeatureImageType = itk::Image<unsigned char, Dimension>;
  using OutputImageType = itk::Image<unsigned char, Dimension>;

  using DataHelperType = itk::ScalarChanAndVeseLevelSetFunctionData<LevelSetImageType, FeatureImageType>;

  using SharedDataHelperType =
    itk::ConstrainedRegionBasedLevelSetFunctionSharedData<LevelSetImageType, FeatureImageType, DataHelperType>;

  using LevelSetFunctionType =
    itk::ScalarChanAndVeseLevelSetFunction<LevelSetImageType, FeatureImageType, SharedDataHelperType>;

  using MultiLevelSetType = itk::ScalarChanAndVeseSparseLevelSetImageFilter<LevelSetImageType,
                                                                            FeatureImageType,
                                                                            OutputImageType,
                                                                            LevelSetFunctionType,
                                                                            SharedDataHelperType>;

  using DomainFunctionType = itk::AtanRegularizedHeavisideStepFunction<ScalarPixelType, ScalarPixelType>;

  auto domainFunction = DomainFunctionType::New();
  domainFunction->SetEpsilon(epsilon);

  LevelSetImageType::Pointer contourImage1 = itk::ReadImage<LevelSetImageType>(argv[1]);
  LevelSetImageType::Pointer contourImage2 = itk::ReadImage<LevelSetImageType>(argv[2]);
  LevelSetImageType::Pointer contourImage3 = itk::ReadImage<LevelSetImageType>(argv[3]);
  FeatureImageType::Pointer  featureImage = itk::ReadImage<FeatureImageType>(argv[4]);

  auto levelSetFilter = MultiLevelSetType::New();
  levelSetFilter->SetFunctionCount(3);
  levelSetFilter->SetFeatureImage(featureImage);
  levelSetFilter->SetLevelSet(0, contourImage1);
  levelSetFilter->SetLevelSet(1, contourImage2);
  levelSetFilter->SetLevelSet(2, contourImage3);
  levelSetFilter->SetNumberOfIterations(nb_iteration);
  levelSetFilter->SetMaximumRMSError(rms);
  levelSetFilter->SetUseImageSpacing(1);
  levelSetFilter->SetInPlace(false);

  for (unsigned int i = 0; i < 3; ++i)
  {
    levelSetFilter->GetDifferenceFunction(i)->SetDomainFunction(domainFunction);
    levelSetFilter->GetDifferenceFunction(i)->SetCurvatureWeight(curvature_weight);
    levelSetFilter->GetDifferenceFunction(i)->SetAreaWeight(area_weight);
    levelSetFilter->GetDifferenceFunction(i)->SetOverlapPenaltyWeight(overlap_weight);
    levelSetFilter->GetDifferenceFunction(i)->SetVolumeMatchingWeight(volume_weight);
    levelSetFilter->GetDifferenceFunction(i)->SetVolume(volume);
    levelSetFilter->GetDifferenceFunction(i)->SetLambda1(l1);
    levelSetFilter->GetDifferenceFunction(i)->SetLambda2(l2);
  }

  levelSetFilter->Update();

  try
  {
    itk::WriteImage(levelSetFilter->GetOutput(), argv[5]);
  }
  catch (const itk::ExceptionObject & excep)
  {
    std::cerr << "Exception caught !" << std::endl;
    std::cerr << excep << std::endl;
    return -1;
  }

  return EXIT_SUCCESS;
}

Classes demonstrated#

template<typename TInputImage, typename TFeatureImage, typename TOutputImage, typename TFunction = ScalarChanAndVeseLevelSetFunction<TInputImage, TFeatureImage>, class TSharedData = typename TFunction::SharedDataType, typename TIdCell = unsigned int>
class ScalarChanAndVeseSparseLevelSetImageFilter : public itk::MultiphaseSparseFiniteDifferenceImageFilter<TInputImage, TFeatureImage, TOutputImage, TFunction, TIdCell>

Sparse implementation of the Chan and Vese multiphase level set image filter.

This code was adapted from the paper: 

   "An active contour model without edges"
    T. Chan and L. Vese.
    In Scale-Space Theories in Computer Vision, pages 141-151, 1999.

This code was taken from the Insight Journal paper: 

"Cell Tracking using Coupled Active Surfaces for Nuclei and Membranes"
http://www.insight-journal.org/browse/publication/642
https://hdl.handle.net/10380/3055
Author

Mosaliganti K., Smith B., Gelas A., Gouaillard A., Megason S.

That is based on the papers: 

"Level Set Segmentation: Active Contours without edge"
http://www.insight-journal.org/browse/publication/322
https://hdl.handle.net/1926/1532

and

"Level set segmentation using coupled active surfaces"
http://www.insight-journal.org/browse/publication/323
https://hdl.handle.net/1926/1533

ITK Sphinx Examples:

See itk::ScalarChanAndVeseSparseLevelSetImageFilter for additional documentation.
template<typename TInputImage, typename TFeatureImage, typename TSharedData = ConstrainedRegionBasedLevelSetFunctionSharedData<TInputImage, TFeatureImage, ScalarChanAndVeseLevelSetFunctionData<TInputImage, TFeatureImage>>>
class ScalarChanAndVeseLevelSetFunction : public itk::ScalarRegionBasedLevelSetFunction<TInputImage, TFeatureImage, TSharedData>

LevelSet function that computes a speed image based on regional integrals of probabilities.

This class implements a level set function that computes the speed image by integrating values on the image domain.

Based on the papers: 

   "An active contour model without edges"
    T. Chan and L. Vese.
    In Scale-Space Theories in Computer Vision, pages 141-151, 1999.

    "Segmenting and Tracking Fluorescent Cells in Dynamic 3-D
     Microscopy With Coupled Active Surfaces"
     Dufour, Shinin, Tajbakhsh, Guillen-Aghion, Olivo-Marin
     In IEEE Transactions on Image Processing, vol. 14, No 9, Sep. 2005
Author

Mosaliganti K., Smith B., Gelas A., Gouaillard A., Megason S.

This code was taken from the Insight Journal paper: 

"Cell Tracking using Coupled Active Surfaces for Nuclei and Membranes"
http://www.insight-journal.org/browse/publication/642
https://hdl.handle.net/10380/3055

That is based on the papers: 

"Level Set Segmentation: Active Contours without edge"
http://www.insight-journal.org/browse/publication/322
https://hdl.handle.net/1926/1532

and

"Level set segmentation using coupled active surfaces"
http://www.insight-journal.org/browse/publication/323
https://hdl.handle.net/1926/1533

ITK Sphinx Examples:

See itk::ScalarChanAndVeseLevelSetFunction for additional documentation.