Stitched Image would be more accurate. But it is unfortunate that the background flourescence is a gradient in each tile. It could affect your IntDen value even if you convert to grayscale reduce it to black. But it would defnitely be an approximate value. You could capture mutiple different fields and see if you are a getting similar results. But try to find atleast a single field with uniform background flourescence to get a reliable IntDen value for this run. This will confirm if your approximate values are on point.

Hi there,

There are a couple of plugins to overcome the unevenness in tile scans in ImageJ:

MosaicJ: http://bigwww.epfl.ch/thevenaz/mosaicj/

Collection stitching: https://imagej.net/Grid/Collection_Stitching_Plugin

And an interesting paper providing an algorithm to work out this effect and explaining why this happens:

https://pubmed.ncbi.nlm.nih.gov/25787148/

Anyway, I would avoid taking tile scans to do measurements in order to prevent these artefacts (since particles in the edges of any of the single images would be compromised for measurements)

However, If I have understood right you are interesting in measuring the bright particles, aren't you?

If this is the case, by performing a Rolling Ball background subtraction (Sternberg, 1983, https://imagej.nih.gov/ij/download/docs/RollingBall.pdf) you can get a pretty clear image and solve most of the unevenness in the background (see the image attached).

For background subtraction I'd work in the stitched Image, and for the measurements, that depends on what you want to measure, but make sure you apply the same settings of background subtraction to all your images.

This is what I did in ImageJ:

-Image>Type>16-bits

-Image>Type>8-bits

-Process>Substract Background...

(It is up to you to choose the value of the rolling ball, I used 100pxls)

-I gave it a green LUT (Image>LookupTables>Green)

and the adjusted min and Max for easier visualization...

(Question: Why is your image in RGB?)

Hope this helps,

Cheers

J

Thank you for your quick answers! I tried the rolling ball background subtractions as well, but I wasn't sure if this would give me the outcome I want. I'll try again with the suggestions you mentioned and see if it will give me similar measurements compared to a single tile with an uniform background.

I'm not familiar with the MosaicJ plugin, so I'll look into that as well.