Calculate cell statistics — epitools.analysis.cell_statistics#

This module contains functions for calculating region-based properties of labelled images using skimage.

epitools.analysis.cell_statistics.calculate_cell_statistics(image, labels, pixel_spacing, id_cells=None)#

Calculate the region based properties of a timeseries of segmented images.

Currently the following statistics are calculated for each frame of the timeseries:

  • area

  • perimeter

  • number of neighbours

skimage.measure.regionprops_table is used to calculated the area and perimeter.

skimage.graph.RAG is used to create a graph of neighbouring cells at each frame, from which the number of neighbours of each cell is calculated.

Parameters:

  • image (napari.types.ImageData) – napari.types.ImageData Timeseries of images (TYX or TZYX) for which to calculate the cell statistics.

  • labels (napari.types.LabelsData) – napari.types.LabelsData Labelled input image, must be the same shape as image. Labels with value 0 are ignored.

  • pixel_spacing (tuple[float, ...]) – tuple[float, …] The pixel spacing in each dimension of the image.

  • id_cells (list[int] | None) – list[int], optional The cell IDs to calculate statistics for. If not provided, statistics will be calculated for all cells.

Return type:

tuple[list[dict[str, npt.NDArray]], list[skimage.graph.RAG]]

Note

It is assumed that the first dimension of both image and labels corresponds to time.

Returns:

list[dict[str, np.NDArray]]

List of dictionaries, where each dictionary contains the cell statistics for a single frame. The dictionary keys are: area; perimeter; neighbours.

list[skimage.graph.RAG]

List of the network graphs constructed for each frame of the timeseries

Parameters:

  • image (napari.types.ImageData)

  • labels (napari.types.LabelsData)

  • pixel_spacing (tuple[float, ...])

  • id_cells (list[int] | None)

Return type:

tuple[list[dict[str, npt.NDArray]], list[skimage.graph.RAG]]