brain_deer.surface.tracts¶

White-matter tract loading and tract scene building.

Loads streamlines from TRK/TCK (via nibabel, if available) or accepts in-memory lists of (P,3) arrays, and builds a TractLayer. Optional per-point scalars (tractometry, e.g. FA along the tract) drive colouring.

Streamline I/O depends on nibabel’s streamlines API; the scene builder itself is headless and works on plain arrays, so it is unit-testable without any tract file.

Functions¶

`load_streamlines`(→ list[numpy.ndarray])	Load streamlines from a TRK/TCK file as a list of (P,3) mm arrays (RAS+).
`subsample`(→ tuple[list[numpy.ndarray], ...)	Randomly keep at most `max_streamlines` (for responsive interaction).
`build_tracts_scene`(, cmap, vminmax, radius, ...)	Build a scene from streamlines, optionally coloured by per-point scalars.

Module Contents¶

brain_deer.surface.tracts.load_streamlines(path: str) → list[numpy.ndarray]¶: Load streamlines from a TRK/TCK file as a list of (P,3) mm arrays (RAS+).

brain_deer.surface.tracts.subsample(streamlines: Sequence[numpy.ndarray], max_streamlines: int, *, seed: int = 0, scalars: Sequence[numpy.ndarray] | None = None) → tuple[list[numpy.ndarray], list[numpy.ndarray] | None]¶: Randomly keep at most max_streamlines (for responsive interaction).

brain_deer.surface.tracts.build_tracts_scene(streamlines: Sequence[numpy.ndarray], *, scalars: Sequence[numpy.ndarray] | None = None, color: tuple[float, float, float] = (0.9, 0.55, 0.1), cmap: str = 'fa', vminmax: Sequence[float] | None = None, radius: float = 0.4, max_streamlines: int = 0, threshold: float | None = None, opacity: float = 1.0, views: Sequence[str] | None = None, background: tuple[float, float, float] = (1.0, 1.0, 1.0), colorbar: bool = True, colorbar_title: str = '', force_solid: bool = False, source_name: str | None = None) → brain_deer.surface.scene_model.SceneModel¶: Build a scene from streamlines, optionally coloured by per-point scalars.