Standardization of Identification Module Parameters & Background Correction

Hi guys!

I have a set of images, each containing three channels: a red and green channel of cytoplasmic CTb stain, and a blue channel labeling Fos nuclei. I am having a bit of a problem with being able to consistently identify and effectively segment cells within this image set. I thought my struggle might originate from background variation, so I applied both illumination function and enhance speckles operations in an attempt to remedy this but I don’t feel it has been sufficient. I can get a good count if I tweak a bunch of settings between each image, but this is time consuming and I was hoping I could streamline the process by having a consistent set of parameters that work for the entire image set.

Here are some sample images:
BO (4.5 MB)

And here is a sample pipeline I have been working on:
TestPipeline.cppipe (32.8 KB)

Does anyone have any advice for me? I have been working on this for a while and am a bit stumped. In case this is relevant, eventually I hope to determine colocalization between CTb and Fos channels.

Thank you!

Totally not cellprofiler related, but do you have subpopulations of cells within your images where you need to segment the cells to compare them to one another?

If you don’t you might be able to make the problem a bit easier:

You might be able to do this at a sample (image) level without segmenting your cells.

If you really need the cell by cell segmentation, don’t mind me, hopefully someone else will be able to help with the CP stuff :slight_smile:

My bad, I phrased that post very misleadingly–my main hope for the pipeline is that it’ll give me an accurate count for number of cells present. The colocalization between channels on the cellular level was just something I figured I could tack on easily once it identified and segmented cells properly. Otherwise, great point, I think this would be the best route. Thank you for your response! :grinning:

1 Like