WeightedOrientation

from mcot.gcoord.orientation import WeightedOrientation

class mcot.gcoord.orientation.WeightedOrientation(white, pial, sulcal_depth, resolution_grid, target_affine)

Can compute an averaged radial/tangential fiber orientations for every voxel.

__init__(white, pial, sulcal_depth, resolution_grid, target_affine)

Prepares computation of the radial/tangential fiber orientations.

Parameters

• white (CorticalMesh) – white/gray matter boundary

• pial (CorticalMesh) – pial surface

• sulcal_depth – sulcal depth

• resolution_grid – voxel size of the accelerator grid in mm

• target_affine – (4x4) array giving the voxel -> mm conversion for the target grid

Inheritance diagram

Methods

average_line_grid(shape[, zval, norient, …])	Computes the radial/tangential orientations on a grid.
average_line_vox(mm_index[, norient, power_dist])	Computes the radial/tangential hemisphere at the given point.
average_point(voxel[, power_dist])	Computes the radial and tangential hemisphere at given voxel.
average_point_grid(shape[, zval, …])	Computes the primary orientations at every point within the pial surface.
closest_vertex_grid(shape[, zval, outside_pial])	Computes the primary orientations at every point within the pial surface.

average_line_grid

WeightedOrientation.average_line_grid(shape, zval=None, norient=1000, power_dist=- 1.0)

Computes the radial/tangential orientations on a grid.

This uses the main FOTACS algorithm: 1. Draw straight lines through the point of interest connecting the cortical surfaces at both sides 2. Linearly interpolate the normal/sulcal depth gradient along this line Repeat these steps for norient random orientations. Average these orientations with the weighting set by the line length ** power_dist.

Parameters

• shape – (nx, ny, nz); defines the shape of the output volume

• zval – only evaluates a single horizontal slice if set

• norient – number of random orientations to try

• power_dist – power-law used to downweight longer faces (weight = dist ** power_dist)

average_line_vox

WeightedOrientation.average_line_vox(mm_index, norient=1000, power_dist=- 1.0)

Computes the radial/tangential hemisphere at the given point.

This uses the main FOTACS algorithm:

1. Draw straight lines through the point of interest connecting the cortical surfaces at both sides

2. Linearly interpolate the normal/sulcal depth gradient along this line

Repeat these steps for norient random orientations. Average these orientations with the weighting set by the line length ** power_dist.

Parameters

• mm_index – (3, ) vector of position in mm

• norient – number of random orientations to try

• power_dist – power-law used to downweight longer faces (weight = dist ** power_dist)

Returns

Tuple with 4 elements:

1. interpolated normal

2. interpolated sulcal depth gradient

3. length of shortest line hitting surface on both sides

4. number between 0 and 0.5 indicating location along shortest line (0 if at edge, 0.5 if in middle of gyrus)

average_point

WeightedOrientation.average_point(voxel, power_dist=- 1.0)

Computes the radial and tangential hemisphere at given voxel.

This algorithm averages the normal/sulcal depth gradient of every vertex weighted by its distance from the point of interest

Parameters

• voxel – (3, ) array with point of interest in voxel coordinates

• power_dist – power-law used to downweight longer faces (weight = dist ** power_dist)

average_point_grid

WeightedOrientation.average_point_grid(shape, zval=None, power_dist=- 1.0, outside_pial=False)

Computes the primary orientations at every point within the pial surface.

This algorithm averages the normal/sulcal depth gradient of every vertex weighted by its distance from the point of interest

Parameters

• shape – shape of the resulting array

• zval – only process a single horizontal slice

• power_dist – power-law used to downweight longer faces (weight = dist ** power_dist)

• outside_pial – if True also run for voxels outside of the pial surface

closest_vertex_grid

WeightedOrientation.closest_vertex_grid(shape, zval=None, outside_pial=False)

Computes the primary orientations at every point within the pial surface.

This algorithm selects the normal/sulcal depth gradient from the closest point.

Parameters

• shape – shape of the resulting array

• zval – only process a single horizontal slice

• outside_pial – if True also run for voxels outside of the pial surface