I have a doubt about the analysis of immunofluorescence signals on Image J. I need to analyse chromatin fibres (so lines with fluorescent signals) and then I want to make a Plot. The question is: How I can remove background from the Plot? I need to ideally quantify a black area of the image, then plot it and remove the list valued from the the list values of Plot of my fibres. But I dont know how to do it in a simply and automatic way. Let me know.
Thank you very much.
If your background is uniform, have you considered background subtracting your entire image?
If you measure the mean background intensity, you can remove the average from every pixel with the
Then, when you measure a line profile, the values will already be background subtracted.
Hope that helps!
thank you very much for you reply…and if the background is not homogeneous and changes for one image to the other?
You can do a (different) background substraction for every image.
Thank you. Ok, really helpful. Just the last doubt, if my image has different mean background, is it corret doing a mean of the different intensities mean and then substract the mean value?
The question is more why your images have different mean background : acquisition parameters are different ? sample have lot of variability ?
In the first case, no need to mean the means, in the second case, why not ?
Well, acquisition parameters change because there is a lot of variability in the same sample. I mean, in one slide i can change even 4 or more times the acquisition parameters because the signals are stronger or less intense depending on a lot of factors which I can’t control. So if I have to plot the signals, I need to substract at least background from every image so to make comparable profile . Moreover, if in the same image I have a part of field with more intense background and an other part of the same image with almost 0 background, doing the mean of these two values makes sense? Thank you.
As far as I know, you really can’t compare images that you took with different acquisition parameters unless you have some sort of control to correlate them, like a bead or something that is in focus and would have a constant signal in each image (I am not sure reading the background would count here because that frequently changes with the depth of the sample). And I’m not even confident that would work well.
If the signals are more or less intense, you generally have to let them be more or less intense, assuming you don’t go over the limits of what the detector can read.
I know that changing parameters images can be not comparable. I know the parameters and in the elaboration phase, I compare only fibres with same parameters acquisition. The problem is that in a same image there could be different backgrounds just because the signal has more or less background in one part of the slide rispect than other one. I think that its normal in immunofluorescence to have a variable background all along the slide.
I agree with both of you that changing parameters may introduce a bias and make images un-comparable. But I don’t feel that using a mean background (the same for all images) would correct this bias.
What I would recommand would be to use the same analysis method for all sample. Background substraction is an “adaptative” method that will take into account the uneven background.