Detecting Cell Clones
The CELL_CLONE overlay (added in v0.9.1) in the CellOverlay plugin detects the outlines of cell subpopulations (clones) based on a secondary input channel. The detection works as follows:
For every cell geometry, the mean intensity is measured within the cell geometry from the selected channel (the image can have multiple channels but the detection will only consider one currently). Given a user defined threshold (by default 100, currently limited to 0-255 <-> 8 bit image), the cells above the threshold are selected and create a subset from the cell network. By using a network property called "connected components" all the connected subgroups (clones) are identified and visualized.The threshold can be dynamically adjusted with the slider, the channel is fixed once the overlay is created.
The excel export accessible through the layer option menu (see example below) currently includes:
- clone_id (increasing integer)
- centroid coordinates of the clone geometry
- size/area (in pixel)
- perimeter (in pixel)
- cell count (# of cells within clone)
- border count (# of cells on the border of the clone)
- neighbor count (# of cells on the other side of the border, i.e. neighboring the clone)
- detection channel (in case the channel in the image has a name)
- the used detection_threshold.
download the cellClone_example.zip from the epitools_samples repository
open MontageCh1.tif in icy and overlay the skeleton with the CellGraph plugin by specifying the MontageSkeleton.tif as input file
Add the CELL_CLONES overlay from the CellOverlay plugin
From the layer menu choose the "Cell Clones" layer and select the export button from the menu at the bottom to save the excel file (an equivalent example excel file is included in the archive)
(Optional) switch off the red polygon_layer by clicking the eye symbol next to the "Polygons" layer
(Optional) adjust the detection_threshold dynamically by moving the slider above the export button
The JAVA code for this overlay can be found here: CellGraph/src/plugins/davhelle/cellgraph/overlays/CellCloneOverlay.java
Thanks to the Fisun lab @ EPFL for the idea of developing this overlay, the example data and helping testing it!