tractolearn.transformation package#

Submodules#

tractolearn.transformation.peaks_utils module#

tractolearn.transformation.peaks_utils.get_peak_count(peak_dirs)#

Get the peak count per spatial location. All locations are expected to have the same peak count. Peaks are assumed to have 3 spatial coordinates (x, y, z). :param peak_dirs: The peak directions. Q must be a multiple of 3. :type peak_dirs: ndarray (X, Y, Z, Q)

tractolearn.transformation.peaks_utils.reshape_peaks_for_computing(peak_dirs, peak_count, spatial_dims)#

Reshape the peak dirs array so that the volume extension indices remain unchanged in the first three dimensions; the peak count indices are made the last but one dimension, and the peak orientation values are made the last dimension, typically (X, Y, Z, P, 3), where P is the number of peaks per voxel. See also dipy.direction.peaks.reshape_peaks_for_visualization :param peak_dirs: Peak directions. Q must be a multiple of 3. :type peak_dirs: ndarray (X, Y, Z, Q) (4D) :param peak_count: Peak count. :type peak_count: int :param spatial_dims: Spatial dimensions of peaks. Typically 3 (x, y, z). :type spatial_dims: int

Returns:

Reshaped peak directions.

Return type:

ndarray

tractolearn.transformation.streamline_transformation module#

tractolearn.transformation.streamline_transformation.flip_random_streamlines(streamlines, ratio=0.5)#

Flips streamlines randomly by reversing the array ordering.

Parameters:

  • streamlines (nib.streamlines.ArraySequence) – Streamlines to be flipped.

  • ratio (float, optional) – Ratio of the streamlines to be flipped. Must be in the range [0..1].

Returns:

  • flipped_streamlines (nib.streamlines.ArraySequence) – Flipped streamlines.

  • streamline_flip_indices (ndarray) – Flipped streamline indices.

tractolearn.transformation.streamline_transformation.flip_streamlines(streamlines)#

Flips streamlines by reversing the array ordering. Note that both the head/tails and the streamline sorting are reversed. To reverse only the heads/tails use reverse_head_tails; to reverse only the streamline sorting use, reverse_streamline_sorting.

Parameters:

streamlines (nib.streamlines.ArraySequence) – Streamlines to be flipped.

Returns:

flipped_streamlines – Flipped streamlines.

Return type:

nib.streamlines.ArraySequence

tractolearn.transformation.streamline_transformation.resample_streamlines(streamlines, num_points, arc_length=True)#

Resamples streamlines to a number of points. If arc length parameterization is used, resampled streamlines will have equal length segments.

Parameters:

  • streamlines (nib.streamlines.ArraySequence) – Streamlines to be resampled.

  • num_points (int) – Number of points for the resampled bundles.

  • arc_length (bool, optional) – If True, use arc length parameterization resampling.

Returns:

resampled_streamlines – Resampled streamlines.

Return type:

nib.streamlines.ArraySequence

tractolearn.transformation.streamline_transformation.sft_voxel_transform(sft)#

tractolearn.transformation.volume_utils module#

tractolearn.transformation.volume_utils.compute_isocenter(img)#

Computes the iso-center of a volumetric image in RAS space (mm).

Parameters:

img (nibabel.nifti1.Nifti1Image) – NIfTI-1 Image.

Returns:

isocenter – Iso-center of the input image.

Return type:

ndarray

tractolearn.transformation.volume_utils.compute_volume(img)#

Computes the volume of an input image in mm^3.

Parameters:

img (nibabel.nifti1.Nifti1Image) – NIfTI-1 Image.

Returns:

Volume of the input image.

Return type:

float

tractolearn.transformation.volume_utils.interpolate_volume_at_coordinates(volume: ndarray, coords: ndarray, mode: str = 'nearest', order: int = 3, cval=0.0) → ndarray#

Evaluates a 3D or 4D volume data at the given coordinates by interpolation.

Parameters:

  • volume (3D array or 4D array) – Data volume.

  • coords (ndarray of shape (N, 3)) – 3D coordinates where to evaluate the volume data.

  • mode (str, optional) – Points outside the boundaries of the input are filled according to the given mode (‘constant’, ‘nearest’, ‘reflect’ or ‘wrap’). (‘constant’ uses 0.0 as a points outside the boundary)

  • order (int, optional) – The order of the spline interpolation.

  • cval (float, optional) – Value used to fill past-boundary coordinates.

Returns:

output – Values from volume.

Return type:

2D array

Module contents#