PyNutil.interpolate_volume#

PyNutil.interpolate_volume(*, segmentation_folder, alignment_json, colour, atlas, scale=1.0, missing_fill=nan, do_interpolation=True, k=5, batch_size=200000, use_atlas_mask=True, value_mode='pixel_count', segmentation_format='binary', segmentation_mode=True, intensity_channel='grayscale', min_intensity=None, max_intensity=None, return_orientation='asr')[source]#

Project section data into atlas-space volumes.

Parameters:

  • segmentation_folder (str) – Path to the folder containing segmentation images or source images.

  • alignment_json (str) – Path to the registration JSON passed to PyNutil.read_alignment().

  • colour – Segmentation color or class identifier to extract. Use None or "auto" to defer selection to the segmentation adapter.

  • atlas (object) – Atlas definition used to determine the target volume shape. This may be a BrainGlobe atlas object or AtlasData.

  • scale (float) – Isotropic scaling factor applied to the atlas output shape.

  • missing_fill (float) – Fill value assigned to voxels with no sampled data when interpolation is disabled or when uncovered voxels remain after processing.

  • do_interpolation (bool) – If True, fill uncovered voxels using k-nearest-neighbor interpolation.

  • k (int) – Number of neighbors to use during interpolation.

  • batch_size (int) – Number of query voxels processed per interpolation batch.

  • use_atlas_mask (bool) – If True, restrict interpolation to voxels inside the atlas mask.

  • value_mode (str) – Output volume mode. Supported values are "pixel_count", "mean", and "object_count".

  • segmentation_format (str) – Name of the segmentation adapter to use when segmentation_mode is enabled.

  • segmentation_mode (bool) – If True, treat input files as segmentation outputs. If False, treat them as source images and derive intensities from intensity_channel.

  • intensity_channel (str) – Image channel to convert to intensity values when segmentation_mode=False.

  • min_intensity (Optional[int]) – Optional lower threshold for intensity-mode inputs.

  • max_intensity (Optional[int]) – Optional upper threshold for intensity-mode inputs.

Returns:

A tuple (interpolated_volume, frequency_volume, damage_volume). The first element stores the requested value volume, the second stores per-voxel sampling frequency, and the third is a binary damage mask.

Return type:

tuple[numpy.ndarray, numpy.ndarray, numpy.ndarray]

Examples

Build atlas-space volumes from segmentation images:

>>> gv, fv, dv = interpolate_volume(
...     segmentation_folder="path/to/segmentations/",
...     alignment_json="path/to/alignment.json",
...     colour=[0, 0, 0],
...     atlas=atlas,
... )