BUG: chan2detect intensity_ratio function using meanImg_chan2 instead of meanImg_chan2_corrected

[mariacozan]

Describe the issue:

I was trying to understand how suite2p determines which ROIs are considered as also being present in the second channel so I started going into the source code. I noticed that there is a function which corrects for bleed through from the green channel which is great. However, I also noticed that when the intensity_ratio is used to actually detect red cells, the meanImg_chan2 is used instead of meanImg_chan2_corrected to get the intensity values. So I was just wondering if perhaps the corrected image should be used instead? Otherwise, unless I missed it, I don't see the meanImg_chan2_corrected being used anywhere else except for visualisation purposes.

Reproduce the code example:

def intensity_ratio(ops, stats):
    """ compute pixels in cell and in area around cell (including overlaps)
        (exclude pixels from other cells) """
 .
 .
 .

# here in line 70 perhaps it should be:

mimg2 = ops["meanImg_chan2_corrected"]
#instead of:
    mimg2 = ops["meanImg_chan2"]
    inpix = cell_masks @ mimg2.flatten()
    extpix = neuropil_masks @ mimg2.flatten()
    inpix = np.maximum(1e-3, inpix)
    redprob = inpix / (inpix + extpix)
    redcell = redprob > ops["chan2_thres"]
    return np.stack((redcell, redprob), axis=-1)

Error message:

No response

Version information:

v0.10.4.dev9+gf93c1ca.d20220615

Context for the issue:

This could hopefully make the red cell detection more accurate. Many thanks!