Build subcortical slabs from a label bounding box

Computes evenly spaced coronal, axial and/or sagittal slabs spanning the bounding box of the requested labels, ready to pass as the slabs argument of create_subcortical_from_volume().

subcortical_views() was renamed to subcortical_slabs() to match the slabs argument of create_subcortical_from_volume() and create_tract_from_tractography().

Usage

subcortical_slabs(
  volume,
  labels,
  coronal = 0,
  axial = 0,
  sagittal = 0,
  pad = 0,
  reorient = TRUE
)

subcortical_views(
  volume,
  labels,
  coronal = 0,
  axial = 0,
  sagittal = 0,
  pad = 0,
  reorient = TRUE
)

Arguments

volume

Path to a label volume, or an integer array already in the builder's frame.

labels

Integer label ids whose combined bounding box frames the slabs.

coronal, axial, sagittal

Number of slabs to produce for each orientation (0 = none).

pad

Voxels by which to expand the bounding box before slabbing.

reorient

Passed to the internal volume reader; must match the value create_subcortical_from_volume() uses (default TRUE).

Value

A data.frame with columns name, type, start, end.

Details

The volume is read with the same axis reorientation the builder uses internally, so the returned slab indices are in the projection frame. This avoids a subtle trap: RNifti::readNifti() and the builder's reader can return different axis orders, so a bounding box computed from the RNifti array points the slabs at the wrong slices (silently producing empty views). Always derive slabs with this function rather than indexing the volume by hand.

See also

create_subcortical_from_volume(), aseg_context()

Examples

vol <- array(0L, dim = c(20, 20, 20))
vol[8:12, 6:14, 9:11] <- 17L
subcortical_slabs(vol, labels = 17, coronal = 3, axial = 2)
#>        name    type start end
#> 1 coronal_1 coronal     6   8
#> 2 coronal_2 coronal     9  10
#> 3 coronal_3 coronal    11  14
#> 4   axial_1   axial     9   9
#> 5   axial_2   axial    10  11