pyXenium.contour.build_contour_feature_table#

build_contour_feature_table(sdata, *, contour_key, he_image_key=_DEFAULT_IMAGE_KEY, inner_rim_um=20.0, outer_rim_um=30.0, include_pathomics=True, embedding_backend=None, pathology_backends=None, precomputed_edge_gradients=None)#

Build a contour-centric multimodal feature table.

The returned payload keeps a wide contour-level table in contour_features and exposes supporting matrices for pseudobulk RNA/protein, pathway activity, ligand- receptor summaries, zone-level composition, and signed-distance gradients.

Parameters:

  • sdata (XeniumSlide)

  • contour_key (str)

  • he_image_key (str)

  • inner_rim_um (float)

  • outer_rim_um (float)

  • include_pathomics (bool)

  • embedding_backend (Any)

  • pathology_backends (Any)

  • precomputed_edge_gradients (DataFrame | None)

Return type:

dict[str, Any]