I've run IHC (using EM48 as primary; ABC kit; visualized with DAB; Nissl counterstain) on mouse brain sections (striatum). How do I count mutant huntingtin inclusions in these sections using ImageJ?
I'm assuming you're looking to count the dark brown spots?
Anyhow, open the original image -- not JPG -- in ImageJ or FiJi -- and then Duplicate it, just to have the original as a reference Image > Duplicate...
Then you need to threshold the image.
One option is to work on the color composite you have via:
Image > Adjust > Color Threshold
Adjust the slides until you have just what you want to count highlighted in red. It may be worth zooming in -- or using a high mag image.
I got pretty good segmentation by just adjusting Saturation to select the region on the left and unchecking Pass--thus selecting the pixels that have high saturation values.
Once you have what you want in red, go to Analyze > Analyze Particles...
Here you can adjust the properties of what you want to count, for example filtering by area, if you want to exclude very small or very large particles. You can also filter by shape -- circularity. Next, you can get ImageJ to mark what it's counting in the Show menu. I normally select Overlay Masks. You can read more about the options here:
https://imagej.nih.gov/ij/docs/guide/146-30.html
In this case, btw, you can get really obvious segementation by splitting the HSB channels: Image > Adjust > HSB Stack.
Followed by Image > Stacks > Stack to Images
The Saturation image shows the brown inclusions crystal clear. Very easy to then use the greyscale thresholding: Image > Adjust > Threshold to select the pixels you want and perform the analysis.
If you want to see your selection on the original image, first go to Edit > Selection > Create Selection on the segmented image and then click back to your color original and go to Edit > Selection > Restore Selection. The particles you selected and counted will be highlighted in yellow.
- Badges
- Science topic
- Similar topics
- Computer Science
- Program