harpy.im.baysor_callable

harpy.im.baysor_callable#

harpy.im.baysor_callable(img, df, name_x, name_y, name_gene, config_path, tmp_dir=None, threads=1, diameter=40, min_size=None, use_prior_segmentation=True, prior_confidence=0.2)#

Perform cell segmentation using the Baysor algorithm.

Designed to be compatible with harpy.im.segment_points for distributed segmentation workflows.

Parameters:

  • img (ndarray[tuple[Any, ...], dtype[TypeVar(_ScalarT, bound= generic)]]) – The input image as a numpy array. Dimensions should follow the format (z, y, x, c).

  • df (DataFrame) – A pandas.DataFrame containing transcripts. Should include columns for coordinates and gene names.

  • name_x (str) – The name of the column in df that contains the X-coordinate of each molecule.

  • name_y (str) – The name of the column in df that contains the Y-coordinate of each molecule.

  • name_gene (str) – The name of the column in df that contains gene identities (e.g. gene symbols or IDs).

  • config_path (str | Path) – Path to the config.toml file used to configure Baysor segmentation parameters.

  • tmp_dir (str | Path | None (default: None)) – Optional temporary directory to store intermediate files. If None, a temporary directory will be created automatically.

  • threads (int (default: 1)) – Number of threads to use during Baysor execution. Defaults to 1.

  • diameter (int (default: 40)) – Approximate expected cell diameter (in pixels). Sets the “scale” parameter in Baysor, which influences segmentation granularity.

  • min_size (int | None (default: None)) – Minimum allowed cell size (in pixels). Cells smaller than this will be filtered out. If None, no filtering is applied.

  • use_prior_segmentation (bool (default: True)) – Whether to incorporate a prior segmentation mask when running Baysor. If True, segmentation results can be influenced by prior knowledge.

  • prior_confidence (int (default: 0.2)) – Degree of confidence in the prior segmentation, ranging from 0.0 (ignore prior entirely) to 1.0 (strictly adhere to prior). Defaults to 0.2.

Return type:

ndarray[tuple[Any, ...], dtype[TypeVar(_ScalarT, bound= generic)]]

Returns:

: A labeled image (numpy array) where each pixel contains an integer representing the segmented cell ID.