Edge intensity


I was hoping to get some clarification on the edge mean intensity. We’re trying to measure the amount of transferrin bound to the cell surface versus the intracellular pool. Up until now we’ve been reducing the cell boundary (stained by phalloidin) by a pixel and subtracting that from the cell boundary to measure the transferrin intensity “on the surface”.

I then suddenly realised that a simpler (and correct?) approach would be to simply take the edge mean intensity value for the transferrin measurement and compare it to the overall intensity of the cell? I’m unsure how the edge and overall mean intensity are related and if they can be compared directly?

Many thanks,


Given your description, this approach should work. The MeanIntensityEdge measurement does pretty much what you would expect it to: collects the pixel intensities along the perimeter of the object and takes the mean value. The one catch is that the mean intensity measurement includes the pixels along the perimeter as well. So if you actually want the cell intensity excluding the surface, you’ll need to shrink the cell object by a pixel, similar to what you’ve been doing.

Is using a single pixel for the surface working well for you? I would assume that if what you are calling the “surface” is not confined to a single-pixel perimeter, you might need to make the edge wider. If that’s the case, you can always shrink the cell by whatever number of pixels is appropriate, do the subtraction as you have before to create a border pixel object, and calculate the mean intensity of this object.