eelbrain.pipeline.SeededParc

class eelbrain.pipeline.SeededParc(seeds, mask=None, surface='white', views=None)

Parcellation that is grown from seed coordinates

Seeds are defined on fsaverage which is in MNI305 space (FreeSurfer wiki). For each seed entry, the source space vertex closest to the given coordinate will be used as actual seed, and a label will be created including all points with a surface distance smaller than a given extent from the seed vertex/vertices. The spatial extent is determined when setting the parc as analysis parameter as in e.set(parc="myparc-25"), which specifies a radius of 25 mm.

Parameters
  • seeds (dict) – {name: seed(s)} dictionary, where names are strings, including hemisphere tags (e.g., "mylabel-lh") and seed(s) are array-like, specifying one or more seed coordinate (shape (3,) or (n_seeds, 3)).

  • mask (str) – Name of a parcellation to use as mask (i.e., anything that is “unknown” in that parcellation is excluded from the new parcellation. For example, use {'mask': 'lobes'} to exclude the subcortical areas around the diencephalon.

Examples

Example with multiple seeds:

parcs = {
    'stg': SeededParc({
        'anteriorstg-lh': ((-54, 10, -8), (-47, 14, -28)),
        'middlestg-lh': (-66, -24, 8),
        'posteriorstg-lh': (-54, -57, 16),
    },
    mask='lobes'),
}

Methods

seeds_for_subject(subject)