Automatic quantification of fluorescence on two channels

fiji

#1

I am new to ImageJ and wanted to see if I could have some help analyzing data. I have images of host cells infected with a dual-reporter strain of Chlamydia and want to quantify the amount of one fluorophore (red channel) to the other (green channel) in various samples. I know I can do this manually by outlining the ROI (which is the population of Chlamydia ) and analyzing the pixel intensity of each channel, but I want to automate this process. Any suggestions on how to do this? Right now the signals of red and green are temporal and not constitutive, so do I need to stain with something that allows for me to visualize the bacteria consistently? I am using mounting media with DAPI to visualize nuclei, but I don’t necessary need this. Let me know what everyone thinks. Thank you!


#2

Hello Young one,
I have sent a sample cleaned and Blue channel blanked, since you don’t seen to care about Blue and I’m not a Biologist and don’t know anything about stains. So I hope this helps. It can be improved in several ways if this is headed in the right direction, then reply back.
Bob
Blue%20Channel%20blank|nullxnull


#3

Yes, this looks good. Thank you, Bob!

I guess the problem I am facing is analyzing the pixel intensity of both the red and green channels simultaneously. And additionally speeding up this process if possible (since circling each object takes some time). Thanks for your help!


#4

Good day!

You won’t be able to get reasonable intensity measures with data like the posted image because the red channel is heavily over-exposed (partially saturated).

Regards

Herbie


#5

Hello again Young One,

What I did is goto Image> color> split channels then I used Edit to clear the Blue channel. Then you could use Process> image calculator and subtract the Red channel from the Green channel and then subtract the Green from the Red THEN threshold each channel and use Analyze> Analyze Particles on each channel, with the Set Measurement to list Minimum, Maximum, and/or Mean. It may sound a little redundant but it should give you the data you want.

Bob


#6

Yes this is exactly what I was wondering how to do. Thank you!!


#7

Good Lauren,( If I may call you that) I’m glad to help. You do very important work, so Thank you!

Bob