OverlayOutlines for edge identification

Hi All,

I am working on a pipeline to measure the edge intensity of podocyte cells, and was hoping to use the OverlayOutlines module to verify my pipeline was accurately determining where the ‘edge’ of cells are. I’d made a post here about a year ago where someone showed me an example output from this module:
Example
But when I use the module I am unable to get a similar output, and can only manage to get a single line (I know the object identification can use some work, I haven’t optimized it yet):


I haven’t seen any posts in this forum about this which makes me think I’m missing something obvious, but I’ve been struggling for a bit so was hoping someone could tell me how to get output more like the first image.

Thanks for any help!
Nick

On that blue/green previous image, it looks like:

  1. They found the nuclei using the Hoechst/DAPI channel,
  2. then grew the nuclei objects to create a larger area very roughly approximating the cell edge,
  3. then used the IdentifyTertiaryObjects module to create a ‘cytoplasmic’ donut (cell edge with nucleus punched out).
  4. overlaid the cytoplasmic donut on the merged RGB image.

In your image, you have the edge of the cell marked out. Do you also have nuclear objects identified? If so, was the nucleus used to help find the cell edge? Then you can use IdentifyTertiaryObjects to generate the donut. But if your interest is just in measuring the intensity along the cell perimeter, then you can skip the nuclear punchout, and shrink your whole cell object by several pixels to create a new object (ShrunkCellObj) and then use FindTertiaryObjects to create your cell edge object.

Note that the CP MeasureIntensity module can also measure the intensity of an object just along its edge. But I believe that’s only 1 pixel wide.