文件列表:

IJ-SimITKVTK-1 (0, 2011-08-22)
IJ-SimITKVTK-1\IJ-Simulink_Libraries_for_VTK_and_ITK-1.pdf (216548, 2011-08-22)
IJ-SimITKVTK-1\SimITK (0, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Examples (0, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Examples\BrainProtonDensitySlice.png (34597, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Examples\BrainProtonDensitySliceBorder20.png (17925, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Examples\BrainProtonDensitySliceRotate.png (26961, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Examples\CannyEdgeDetectionFL2Model.mdl (16814, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Examples\ConfidenceConnectedFL2Model.mdl (17303, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Examples\FastMarchingFL3Model.mdl (19692, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Examples\ImageToImageRegistrationHelperUC2Model.mdl (18405, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Examples\ResampleFL3Model.mdl (17929, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Examples\ResampleUC2Model.mdl (17874, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Examples\cthead1.png (29353, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Examples\mlgaussian.m (1580, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Examples\subject04_t1w_p4_float.mha (158, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Examples\subject04_t1w_p4_float.raw (47448064, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source (0, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\BaseClassCopy.pm (2402, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\BlockGenerator.pl (26304, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\CMakeLists.txt (7784, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\Container.tpp.in (2025, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\ContainerGen.pl (3306, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\ContainerMat.cpp.in (9583, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\Copyright.txt (1531, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\FilterBlock.tpp.in (2038, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\FilterBlockMat.cpp.in (9830, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\FilterBlockRegHelper.tpp.in (1886, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\FilterMask.in (1155, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\FilterMaskGen.pm (10747, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\FilterMaskNoParameters.in (561, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\FilterXMLDescriptions.xml (32553, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\ImageConversion.h (1306, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\ImageConversion.tpp (3427, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\InterpolatorFilterBlock.tpp.in (1175, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\InterpolatorFilterBlockMat.cpp.in (8936, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\Library.mdl.in (7037, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\LibraryGen.pm (1594, 2011-08-22)
IJ-SimITKVTK-1\SimITK\Source\Reader.tpp.in (3137, 2011-08-22)
... ...

SIMITK Project Karen Li and Jing Xiang June 2008 ================== BUILDING SIMITK ================== If you are using Windows, then you can go directly to the Bin directory, which has copies of all the files that you need to use SimITK with MATLAB 2007a, 2007b, or 2008a if you already have Simulink installed for your version of MATLAB. If you need to build SimITK, then follow the instructions below. Before building SimITK, you must first ensure that ITK, MATLAB, Perl and CMake have been installed. We have successfully built and tested SimITK using ITK 3.6, MATLAB 2007a, Perl 5.10.0 and CMake 2.6. Building on Windows ------------------- SimITK has been tested on Windows using Microsoft Visual Studio 2005 as the compiler. Because SimITK uses Simulink, you should have a Simulink license, and you must have Simulink installed with your copy of Matlab. 1. Run CMakeSetup.exe. Specify the proper SimITK source folder and the folder where the binaries should be built in the appropriate fields at the top of the graphical window. Press Configure and ensure the Cache Values window correctly specifies all library paths. It is important to note that different versions of MATLAB may have different directory structures and thus these paths may need to be modified depending on the version number. If the paths are updated, Configure must be pressed again. 2. Once configuring is done, press OK. A .sln file will be created by CMake in the binary folder. Open this file in Visual Studio and compile the project. 3. The SimITK .mdl Library files will be found in the same binary folder as the .sln file. The SimITK .dll files will be created in a subdirectory of the binary folder, either in Debug or Release depending on which mode was used for compiling. The .mdl and .dll files must be moved to the same folder. SimITK can now be used from this folder. Building on Linux ------------------- Because SimITK uses Simulink, you should have a Simulink license, and you must have Simulink installed with your copy of Matlab. In order to build SimITK on Linux, you must use a compiler that is compatible with MATLAB. For Matlab 7.4 (2007a) and 7.6 (2008a), the supported compiler is gcc-3.2, but we have also found that gcc-3.4 and gcc-4.0 work, while gcc-4.2 does not work. The copy of ITK that you use should be built with the same compiler. 1. Make a new directory called "Build". From inside this directory, execute: ccmake Press "c" to configure. The first time around, it will complain about missing files. Press "e" to hide the errors and go back to the main cmake screen. 2. Set MATLAB_ROOT to the directory where Matlab is installed, e.g. /usr/local/matlab74, and press "c" again. If it still complains about missing Matlab libraries, press "t" to go to the advanced view and manually enter the locations of the Matlab libraries and header file directories. 3. Set the ITK_DIR to your ITK binary directory and press "c" again. Also set the CMAKE_BUILD_TYPE to the same build type that you used for ITK, e.g. CMAKE_BUILD_TYPE=Release. 4. Press "g" to generate the make files for SimITK. When cmake exits, type "make" to build SimITK. 5. The binary directory will now contain SimITK .mdl Library files as well as SimITK .mexglx (Linux 32-bit) or .mexa*** (Linux ***-bit) files. SimITK can be used from this directory. ================== EXAMPLES ================== For all of the example models provided, ensure the .mdl model file is in the same directory as the SimITK .dll files and .mdl Library files, and then open the model in MATLAB. Clicking on the individual SimITK blocks in the model shows the parameter values of the blocks. The simulation settings of all models are set so that the models do not run continuously. These settings can be found by going to the 'Simulation' menu and selecting 'Configuration Parameters'. Note that both the start and stop simulation time are set to 0, the solver type is fixed-step and the solver is discrete. When the models are run, the output image is created in the current directory. CannyEdgeDetectionFL2Model.mdl ------------------------------ This model identifies the edges of an image by applying the itk::CannyEdgeDetectionImageFilter. The input image is a 2-dimensional computed tomography of the head and is taken from the ITK example data. The parameters of the blocks in this model are set as follows: itkReaderFL2 FileName 'cthead1.png' itkCannyEdgeDetectionImageFilterFL2 LowerThreshold 10 MaximumError [0.01 0.01] OutsideValue 0 UpperThreshold 30 Variance [0.1 0.1] itkWriterFL2 FileName 'outputCannyEdgeDetectionFL2Model.mhd' ConfidenceConnectedFL2Model.mdl ------------------------------- This model performs region growing segmentation using the itk::ConfidenceConnectedImageFilter. Before the ITK filter is executed, Gaussian smoothing is done using a MATLAB function. Thus, this pipeline demonstrates the combining of both MATLAB and ITK functionality. The input used is an MRI of the brain, taken from the ITK example data, and the output shows a white matter segmentation. The parameters of the blocks in this model are set as follows: itkReaderFL2 FileName 'BrainProtonDensitySlice.png' gaussian M-file name mlgaussian Parameters 3 itkConfidenceConnectedImageFilterFL2 InitialNeighborhoodRadius 3 Multiplier 2.5 NumberOfIterations 5 ReplaceValue 255 Seed [128 116] itkWriterFL2 FileName 'outputConfidenceConnectedFL2Model.mhd' ImageToImageRegistrationHelperUC2Model.mdl ------------------------------------------ This model demonstrates image registration using the ImageToImageRegistrationHelper block. An MRI of the brain was rotated and then registered back to itself using rigid registration with ITK's mean squares metric. The parameters of the blocks in this model are set as follows: itkReaderUC2 FileName 'BrainProtonDensitySliceBorder20.png' itkReaderUC1 FileName 'BrainProtonDensitySliceRotate.png' itkImageToImageRegistrationHelperUC2 InitialMethodEnum 2 EnableAffineRegistration 0 EnableBSplineRegistration 0 EnableInitialRegistration 1 EnableRigidRegistration 1 RigidMaxIterations 500 RigidSamplingRatio 0.05 RigidTargetError 0.1 RigidInterpolationMethodEnum 0 RigidMetricMethodEnum 2 RigidOptimizationMethodEnum 1 [all remaining Affine and BSpline parameters can be set to 0] itkWriterUC2 FileName 'outputImageToImageRegistrationHelperUC2Model.png' ResampleUC2Model.mdl -------------------- This model applies the itk::ResampleImageFilter to a 2-dimensional image. In addition to the image, this filter expects a transform and an interpolator as inputs. The transform is used to map the input image to an output image, and the interpolator is used to generate image intensities at non-grid positions. Interpolators are represented in SimITK as special blocks with no parameters. In this case, the LinearInterpolateImageFunctionUC2 block is used and connected to the special third input of the ResampleImageFilter block. Transforms are also represented in SimITK as special blocks. Unlike the other filters and special blocks, transforms are not specific to pixeltype. Therefore, they are automatically generated separately from the other filters into the libraries SimITKTransformLibrary2D.mdl and SimITKTransformLibrary3D.mdl, distinguished only by dimensionality. All transforms take two arrays as parameters, Parameters and FixedParameters. The size and content of these arrays will vary according to transform. In this sample model, the TranslationTransform2D block is used to specify a translation of 25 along both the X and Y axes. The transform block is connected to the special fourth input of the ResampleImageFilter block. The parameters of the blocks in this model are set as follows: itkReaderUC2 FileName 'cthead1.png' itkTranslationTransform2D Parameters [25 25] FixedParameters [] itkResampleImageFilterUC2 DefaultPixelValue 0 OutputOrigin [0 0] OutputSpacing [1 1] Size [256 256] itWriterUC2 FileName 'outputResampleUC2Model.png' FastMarchingFL3Model.mdl ------------------------ The FastMarchingFL3Model links several SimITK blocks together into an extended pipeline which follows the Fast Marching Segmentation example given in the ITK Software Guide. The model also demonstrates the use of the NodeContainer special block. A 3-dimensional MRI of the brain is passed as input to the CurvatureAnisotropicDiffusionImageFilter. The output is a smoothed image which is then passed to the GradientMagnitudeRecursiveGaussianImageFilter and then to the SigmoidImageFilter before finally being used as input to the FastMarchingImageFilter. The FastMarchingImageFilter requires a set of one or more seed points to begin segmentation. In the model, these seeds are set using the SimITK NodeContainer special block. This block takes two parameters, an array of N seed values and an array of respective indices which will have a size of N*ImageDimensionality. Once the parameters have been set and the output of the NodeContainer block has been connected to the special third input of the FastMarchingImageFilter block, the filter can be used. This example model shows a segmentation of white matter. The parameters of the blocks in this model are set as follows: itkReaderFL3 FileName 'subject04_t1w_p4_float.mha' itkCurvatureAnisotropicDiffusionImageFilterFL3 ConductanceParameter 9.0 NumberOfIterations 5 TimeStep 0.0625 itkGradientMagnitudeRecursiveGaussianImageFilterFL3 Sigma 1.0 itkSigmoidImageFilterFL3 Alpha -15 Beta 2.0 OutputMaximum 1 OutputMinimum 0 itkNodeContainerFL3 Value [0.5] Index [101 118 ***] itkFastMarchingImageFilterFL3 StoppingValue 100 itWriterFL3 FileName 'outputFastMarchingFL3Model.mhd' ResampleFL3Model.mdl -------------------- This model is similar to the ResampleUC2Model but instead uses 3-dimensional data. The transform must now be taken from the 3D Transform library and here we choose the CenteredEuler3DTransform3D block. This block takes an array of 9 elements which specify a rotation, the center of rotation and a translation. We specify a small rotation of 20 around the Z-axis and connect the transform to the ResampleImageFilter block. The parameters of the blocks in this model are set as follows: itkReaderFL3 FileName 'subject04_t1w_p4_float.mha' itkCenteredEuler3DTransform3D Parameters [0 0 20 127.5 127.5 90 0 0 0] FixedParameters [] itkResampleImageFilterFL3 DefaultPixelValue 0 OutputOrigin [0 0 0] OutputSpacing [1 1 1] Size [256 256 181] itWriterFL3 FileName 'outputResampleFL3Model.mhd'

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：