opentps.core.processing.registration package
Submodules
opentps.core.processing.registration.midPosition module
- compute(CT4D, refIndex=0, baseResolution=2.5, nbProcesses=-1, tryGPU=True)
Compute mid-position image and corresponding deformations from a 4D image.
- Parameters:
CT4D (Dynamic3DSequence) – 4D image.
refIndex (int) – index of the reference phase in the 4D image.
baseResolution (double) – spacing of the highest registration resolution (i.e. spacing of the output deformation fields)
nbProcesses (int) – number of processes to be used in Morphons registration (-1 = maximum number of processes)
- Returns:
numpy array – MidP image.
list – List of deformations between MidP image and each phase of the 4D image
opentps.core.processing.registration.registration module
- class Registration(fixed, moving)
Bases:
object
Base class for registration.
- Variables:
fixed (image3D) – fixed image.
moving (image3D) – moving image.
deformed (image3D) – deformed moving image.
roiBox (list) – region of interest for registration.
- computeImageDifference(keepFixedShape=True, tryGPU=True)
Compute the difference between the fixed and moving image.
- Parameters:
keepFixedShape (bool) – If True, the moving image is resampled to the fixed image grid. If False, the fixed image is resampled to the moving image grid.
tryGPU (bool) – If True, try to use GPU.
- Returns:
diff – Difference between the 2 images.
- Return type:
3DImage
- computeSSD(fixed, deformed)
Compute the SSD metric between two images.
- Parameters:
fixed (numpy array) – Fixed image.
deformed (numpy array) – Deformed image.
- Returns:
ssd – SSD metric value.
- Return type:
double
- regularizeField(field, filterType='Gaussian', sigma=1.0, cert=None, tryGPU=True)
Regularize vector field using Gaussian convolution or normalized convolution.
- Parameters:
field (numpy array) – vector field to be regularized.
filterType (string) – type of filtering to be applied on the field.
sigma (double) – standard deviation of the Gaussian.
cert (numpy array) – certainty map associated to the data.
- resampleFixedImage(tryGPU=True)
Resample the fixed image to the moving image grid.
- Parameters:
tryGPU (bool) – If True, try to use GPU.
- Returns:
resampled
- Return type:
3DImage
- resampleMovingImage(keepFixedShape=True, tryGPU=True)
Resample the moving image to the fixed image grid.
- Parameters:
keepFixedShape (bool) – If True, the moving image is resampled to the fixed image grid. If False, the fixed image is resampled to the moving image grid.
tryGPU (bool) – If True, try to use GPU.
- Returns:
resampled – Resampled image.
- Return type:
3DImage
- setROI(ROI)
Set the ROI to be used for registration.
- Parameters:
ROI (roiMask) – ROI to be used for registration.
- translateAndComputeSSD(translation=None, tryGPU=True)
Translate the moving image and compute the SSD metric. :Parameters: * translation (list) – Translation vector.
tryGPU (bool) – If True, try to use GPU.
- Returns:
ssd – SSD metric value after translation.
- Return type:
double
- translateOrigin(Image, translation)
Translate the origin of an image.
- Parameters:
Image (3DImage) – Image from which the origin is translated.
translation (list) – Translation vector.
opentps.core.processing.registration.registrationDemons module
- class RegistrationDemons(fixed, moving, baseResolution=2.5, tryGPU=True)
Bases:
Registration
Perform Demons registration between fixed and moving images. inherited from Registration class.
- Variables:
- compute()
Perform Demon registration between fixed and moving images.
- Returns:
Deformation from moving to fixed images.
- Return type:
numpy array
opentps.core.processing.registration.registrationMorphons module
- class RegistrationMorphons(fixed, moving, baseResolution=2.5, nbProcesses=-1, tryGPU=True)
Bases:
Registration
Class for performing registration using morphons kernels. inherited from Registration class.
- Variables:
- compute()
Perform registration between fixed and moving images.
- Returns:
Deformation from moving to fixed images.
- Return type:
numpy array
- applyMorphonsKernels(image, k, is_fixed=1, tryGPU=True)
- morphonsComplexConvD(im, k)
- morphonsComplexConvS(im, k)
- morphonsConv(im, k)
opentps.core.processing.registration.registrationQuick module
- class RegistrationQuick(fixed, moving)
Bases:
Registration
Perform quick translation search between fixed and moving images. inherited from Registration class.
- compute(tryGPU=True)
Perform registration between fixed and moving images.
- Returns:
Translation from moving to fixed images.
- Return type:
- matchProfiles(fixed, moving)
Find shift between two profiles by minimizing the mean squared error (MSE).
- Parameters:
fixed (array) – fixed profile.
moving (array) – moving profile.
- Returns:
shift between profiles.
- Return type:
int
opentps.core.processing.registration.registrationRigid module
- class RegistrationRigid(fixed, moving, multimodal=False)
Bases:
Registration
Perform rigid registration between fixed and moving images. inherited from Registration class.
- Variables:
- compute()
Perform rigid registration between fixed and moving images.
- Returns:
Transform from moving to fixed images.
- Return type:
opentps.core.processing.registration.registrationTranslation module
- class RegistrationTranslation(fixed, moving, initialTranslation=[0.0, 0.0, 0.0])
Bases:
Registration
Perform translation registration between fixed and moving images. inherited from Registration class.
- Variables:
- compute()
Perform registration between fixed and moving images.
- Returns:
Translation from moving to fixed images.
- Return type: