Cell membrane as object?

SHORT VERSION: I have objects (cells) and want to have the outer 4 pixels as another object (membrane). I can shrink the cell to get the cytosol, but cannot multiply this with the original cells to have only the membrane left. Is there something equivalent to ImageMath?

LONGER VERSION:
Hi,
I hope this topic has not yet been here - at least I could not find it.

My goal is to measure the fluorescence along the cell membrane and compare it to the expression of a protein inside the cell. The data sets consist of images “DNA”, “MembraneProtein” and "CytosolProtein"
I started out with a method like this:

  • IdentifyPrimaryObjects (“DNA” -> “nuclei_obj”)
  • Identify SecondaryObjects (using “nuclei_obj” on “MembraneProtein” -> “cells_obj”)
    This gives me the cells and works fine. Now, I define the membrane as the outer 4 pixels:
  • ExpandOrShrinkObjects (“cells_obj” -> “intracell_obj”)

Here, I am stuck: Is there a way of subtracting intracell_obj from calls_obj, so I would end up having only the outer 4 pixels of my cells?

I tried to achieve this by converting the objects to binary images, subtract (cells_objBin - intracell_objBin = membrane_Bin) and multiply the result with the original images, but this makes my life a bit harder, since I end up having a lot of zero-pixels in the middle of my membrane data. This also makes it hard to filter out cells that have fluorescence above the background (with transient transfection, this is important for me).

Has anyone an idea how to solve this?

Hi again,
I found a solution (and hope it is OK):

  • Identify “cells” as secondary objects
  • Shrink (–> intracell.)
  • Mask objects: “cells” masked by inverted "intracell."
    Maybe it helps another beginner :smile:

That’s what would have done. :smiley:
-Mark