Cell Image

Description

A gnomon Cell Image is a data strucure which could hold time series segmented image.
The correspoonding class in gnomon is gnomonCellImageDataTissueImage.
It gives possibility to access some information about image like:

• Cell property

• Number of cells

• Dimensions

• Voxel Size

Default reader plugin

The default reader of cell image form is cellImageReaderTimagetk. Which reads a 3D image file.
Extensions supported by this reader are: tif, inr.gz, inr.

Default writer plugin

The default writer of cell image form is cellImageWriterTissueImage

Gnomon Cell Image Example

Plugins which take Cell Image as input

Here is a non exhaustive list of some algorithms which take this form as input.

• morphoCellFilter

• seededWatershedSegmentationTimagetk

• signalQuantificationImageSignal

Plugins which produce Image as output

Here is a non exhaustive list of some algorithms which produce this form as output.

• seededWatershedSegmentationTimagetk

• seedImageDetectionTimagetk

• surfaceCellCurvature

from dtkcore import d_inliststring
from dtkcore import d_int

from gnomon.utils import algorithmPlugin
from gnomon.utils import load_plugin_group
from gnomon.utils.decorators import cellImageInput
from gnomon.utils.decorators import cellImageOutput
from gnomon.core import gnomonAbstractCellImageFilter

from timagetk.algorithms.morphology import label_filtering
from timagetk.components.tissue_image import TissueImage3D

load_plugin_group("cellImageData")

@algorithmPlugin(version="0.3.1", coreversion="0.81.1")
@cellImageInput("in_tissue", data_plugin="gnomonCellImageDataTissueImage")
@cellImageOutput("out_tissue", data_plugin="gnomonCellImageDataTissueImage")
class morphoCellFilter(gnomonAbstractCellImageFilter):

    def __init__(self):
        super(morphoCellFilter, self).__init__()
        self.in_tissue = {}
        self.out_tissue = {}

        self._parameters = {}
        self._parameters['method'] = d_inliststring("Morphological operation to apply", "erosion", ["erosion", "dilation", "opening", "closing"])
        self._parameters['radius'] = d_int("Radius of the structuring element", 1, 1, 50)
        self._parameters['iterations'] = d_int("Number of iteration of the morphological operation", 1, 1, 5)

    def run(self):
        self.out_tissue = {}

        for time in self.in_tissue.keys():
            img = self.in_tissue[time]  # This is a SpatialImage!
            filtered_img = label_filtering(img,
                                           method=self['method'],
                                           radius=self['radius'],
                                           iterations=self['iterations'])

            tissue = TissueImage3D(filtered_img, background=1, not_a_label=0)
            self.out_tissue[time] = tissue