Microglia IBA1 DAB

cellprofiler

#1

Hello,

I am using Cellprofiler to properly outline the shape of the microglia. However, as the images below show, the accuracy of the pipeline is off but I do not know where to start troubleshooting. It seems like the cell body is detected but some of the darker-stained processes are incorrectly identified as cell bodies as well. Any help in tweaking my pipeline would be appreciated!

Best,
Selena


#2

Hi,
Without seeing your pipeline (or knowing specific regions you’re displeased with) it’s hard to say, but it actually seems like CP is doing something reasonable- it doesn’t “know” you don’t want it to track down the processes.

You could try smoothing the image, since the processes are skinny and should “disappear” more quickly with some smoothing. You could also try to enahance the processes using EnhanceOrSuppressFeatures then threshold and mask those areas out.

Good luck!


#3

Thanks for your prompt reply! Here is my pipeline for you to reference; I have already tweaked it a bit to improve my results.
The main issue now is that cell bodies are detected as secondary objects as well. My goal is to have cell bodies as primary objects and the processes stemming from the cell bodies as secondary objects. I have come a long way and now I feel like I have the finish line in sight! smcg_mod2.cpproj (411.1 KB)