How to choose the right channel for image analysis?


I have tried to understand how splitting channels works but it still remains unclear for me. I am analyzing COX and SDH activity in mouse muscle. One colleague taught me to use ImageJ but she was unaware why she had done the things she did with ImageJ. She said I should split channels before analyzing the image.

How do know what channel to use? Is it staining-dependent or what? Can I use the same calibration for all channels?

I have read the manual and several topics here in the forum but couldn’t find an answer…

Thank you for all the help!

1 Like

NSP, Welcome,
When checking your image before splitting channels, hover your mouse pointer over the parts of the image you are interested in. Then note the color values in the bar as red, green and blue. The highest value is the predominant color at that point, then decide which channel to use.

1 Like

You should definitely use the same pixel size for all channels, but much of what you do will depend on the original image, image format, etc. And when you split channels, the channel split *should always be in the same order.

If you want more reading, this helped me when I started:
I specifically selected the chapter on colors and channels, but there are several well laid out sections in terms of good practices for various types of analysis.

*as long as the images were taken in the same manner, same instrument, etc.


Hi and thank you for the answer!

Where can I see the color values? There is no bar visible. I tried using histogram for COX stained samples and there red channel mean was 204, green 172 and blue 137. Does this mean that I should use red channel? That would be different to what my colleague suggested. She had also analyzed the same COX activity where the samples looked pretty much the same and used blue channel… For SDH, the values were 117 (red), 131 (green) and 173 (blue).

In the photo you can see how my COX image looks like after splitting: red, green and blue channels. To my eye, the green one looks the best, not too saturated but still distinguishable differences between cell types.

I would be very grateful for any advice!

Actually, now I understood the color values in the bar! This would imply that I should use red channel (on the left in the picture), the one with the faintest optical density…

Many thanks for the great handbook!

What do you mean by splitting the channels in the same order? I thought they just appear at the same time/order after the command.

1 Like

They will if the original file is the same. I have run into cases where various users on the same projects have rearranged their channels at the microscope, and so some of the resulting files ended up with a different channel order.

And the handbook is all @petebankhead !

1 Like

Just to check, if this is a fluorescent image, I think you are fine, but if this is a brightfield image, you may want to look into color deconvolution, provided you have a simple staining setup.
I found a random video that may or may not be out of date.

Hello again NSP,
I apologize, got caught up in another project.
Did you achieve what you wished? I am not a Biologist therefore I do not know what you were referring to with the acronyms but just playing with your supplied Image this is what I acquired.
Please let me know how it turned out and if you need any further assistance.