@LeaA … and i would not use the whole cell as the ROI… thge biology you care about is probably only happening in some subcompartment of the whole cell, eg not in the nucleus… but only in those large roundish objects???
Maybe just analyse the parts of the cell where the biology you care about is happening. Draw a mask image with the parts you want and use that for roi selection. Maybe use oneof the image channels to segment and make the mask image.
It looks like the pixels are way too small… 45 nm…
if you are using a high NA oil immersion lens, NA 1.4, then pixels should be 60-100 nm separated.
Your image is noisier than it could be because the signal is spread out over too many pixels. Empty magnification.
You need to deconvolve the images before analysis to suppress noise and fix the systematic error of the point spread function killing the contrast of small features.
At the very least you need to de noise the images, best by deconvolution, or at a push perhaps a gaussian or mean filter.
There is a scanner mis- calibration on your confocal, i can see you maybe used bidirectional scanning, but the adjavent lines are a little out of position with respect to each other… i can see the combing effect of that in the bright objeects.
you need to optimise the imaging conditions first to get good image date, then deconvolve it, then do this analysis on just the biologically interesting parts of the cell.
then you will get a more meaningful answer.
Dont forget to state the spatial image and optical resolution the experiment is done at - or else the results have no spatial size context.
“Pixel intensity correlations were measured at xxx resolution”