Separating colors in light microscopy images: Color Deconvolution in ImageJ/FIJI

Link to my small website: sites.google.com/view/kru-neu...
Here's a quick video on how to do color deconvolution in ImageJ/FIJI. It will separate out colors of various light microscopy stains, which you may find useful for counting cell or particles of specific colors. See my other videos for how to do cell counts and how to quantify color/stain intensity.

[@adoniscoleman4136]

Hello, Thank you, this is exactly what I have been looking for. I was curious as to whether you can then proceed to find the % area of the blue to represent the % fibrosis of pancreas. I understand that in Masson's Trichrome staining, some of the blood vessels may be also stained blue along with the collagen. I would greatly appreciate your help. Also, if you are open to communicating via email please let me know for future follow-up questions. Thanks again

[@kruneuro]

Hi there! When going back to my video just now, this one came up in my suggestions. It will likely be helpful for you: kzbin.info/www/bejne/a5Lco32OfLKFj9E So, after doing the color deconvolution, I suggest duplicating the blue photo via Image > Duplicate. Move the windows of these two blue images so they are side by side. Then, in the copy, go to Image > Adjust > Threshold. Adjust the sliders such that the staining pattern & shapes of this black and white (or red) image match those of the original blue image, then click Apply. Per the video I linked, go to Analyze > Set Measurements, and ensure Area, Area Fraction, and Display Label are checked. Then select the window with your thresholded image (otherwise it might process other images you have open), go to Analyze > Measure, and that should do it! One thing to confirm though is which part it is measuring - the stain or everything that is not stained. For me, since I selected dark background, the area % was for the white unstained part rather than the black - but this is just a simple fix with 100 minus that % value to get the stained area percentage. I'm unsure how to separate out blood vessels in your case, as I have never personally used the stain. But is the blood vessel staining intensity in the blue part lighter (or darker?) than the fibrosis? If that IS the case, then you can just remove most blood vessels from the blue part by careful adjustments to the threshold bars. Otherwise, you may have to apply a different stain alongside the masson trichrome that illuminates the blood vessels so that they can be subtracted out. I have that subtraction method described in another video I put out recently. You can also now find a link to my (small) website in this video's description - my email is there as well as a few other things.

[@adoniscoleman4136]

@@kruneuro Hello, Thank you so much for a speedy reply! This information is very helpful. Thank you!

[@gengpan]

Will try this

[@mariabelenolivares7745]

Thanks a lot for yhis useful tip! Where can I find the next steps to selectively cuantify the blue part of the image?

[@kruneuro]

I recently posted another video on my channel that should help you: kzbin.info/www/bejne/hWGvfpuJlK-nfdU (Title starts with "Comparing color intensities..." if the link doesn't work.)