I have been trying to simply quantify the number of cells in the tiff files I am getting using cell profiler. This is the closest I have gotten and the image processing seem to make the cells very clear, but I don’t know the right setting to get the identify primary objects to count the cells correctly. The colored dots are the cells/objects my Marco is counting but it is clearly it is making mistakes, not counting cells when they’re there and counting cells when they’re not there. Any suggestions? I tried using threshold to make the cells really stick out but this made the object count significantly worse even though it is obvious in the image. What do people usually use or what would you recommend for this?
I attached my macro and what it looks like right before it process the identify primary object which seems clear and the colored picture to the right shows what I got which is clearly wrong
This is the image processing I did (top left) and the cells it counted (top RIght) and image shown after threshold
With Image J, I get 101 particles.
Here is the heart of the macro:
run("Duplicate...", "title=temp"); setOption("ScaleConversions", true); run("8-bit"); run("Kuwahara Filter", "sampling=8"); wait(200); run("Subtract...", "value=240"); setAutoThreshold("Minimum dark");
In your EnhanceSupress feature select the feature type as “Dark holes” instead of “Speckles”. Then in “Indentify Primary Object module” changed the thresholding method & slightly alter the intensity just to capture all the object. Now it shows around ~130 objects. Please find the attached modified pipeline. This chane I had made it to the first part of the pipeline since I am not sure about the later part.
You earlier EnhanceSupress feature enhance the cells as a dark object, so it identifies the region outside that. Once this is solved you could chose normal thersholding method.
Hope this helps.Final_counting1.cpproj (1.0 MB)
Read more on our site.
Yokogawa CV8000 - The Ultimate in Confocal HCS
Join us at, https://www.slas2020.org