Integrated intensity in different channel

Hi guys, I seem to run into a problem. I want to measure the total intensity in a specific area around objects “A” to be able to say something about the presence or not (and intensity) of objects “B” in this area. Attached I have an image created by the program that indicates the outlines based on objects “A” in which I’m looking for objects “B”, projected on the channel that contains objects “B”. I think up to this point everything seems to work. But then when I calculate the integrated intensity within these outlines not every value seems to correlate well with what I see myself. What I get:

Object - Integrated intensity

1 - 111.772.337.183
2 - 144.054.323.691
3 - 182.229.343.941
4 - 0.0
5 - 0.125307089649
6 - 0.324574659346
7 - 121.829.558.024
8 - 0.621896702331

So far the high values actually correspond to the presence of objects “B” so I think that’s promising, but then:

9 - 1.499.977.133

Which is a factor 100 below the other high values while object 9 looks more intense than most of the other ones.

I applied a threshold before I measured intensity, I assume that’s where I’m making a mistake. What I did is apply a background threshold per object (the areas within the outlines in the attachment). Should I do this in a different way? My preference would be to calculate one background threshold value for all the images so that I can also subtract this value from all the measured intensities, but I’m not sure how I can do this. Does anyone have any ideas?

Thanks in advance!
Bas


My bad, I just realized that in excel unintended dots and comma’s were messing up the values.

I now set a threshold manually that seems to work. Unless someone has a better idea I’m ok working with this.

Hi Bas,
If a manual threshold works for you, I won’t argue with it :smiley: Just make sure the threshold still works across experimental runs; if not, you can try one of the automatic thresholding methods.
-Mark

Thanks! In the end I figured out how to easily measure background intensity so instead of thresholding I’m now substracting this, works perfect!