Plaque load and clearance rate ( watershed and positive detection)


i am a new user of qupath i tried to analyze my fluorescence maximum projection image obtained by slide scanner.

where Af555 channel is intended for thiazine red stained plaques and AF488 is for specific cell marker. my main question is to determine the plaques area which is outside cells and how much plaque has been engulfed by cells

i can not use dapi channel for the positive cell detection because it is ex-vivo assay where i added culture cells to my slices
the problem after several trials to adjust the sitting still the watershed detection is detecting plaque inside the positive cell detections.

i do not know how i can optimize this sitting and whether qupath works for this kind of images.
another issue when i open the original image which has 2 scenes i only see one and can not find a way to see the other scene.

image ex

A couple of things that might be useful here in addition to the image:

  1. What version of QuPath are you using?
  2. In many cases, multiple scenes only show up when you are in a project, which is the structure that allows you to select between multiple files/images. Are you in a project?

Also, while DAPI does not work on live cells, Hoechst and other options do. Though they can have negative effects on the cells over long periods of time…

It looks like you have some sort of nuclear marker here, but I also can’t see how you chose to outline your cells. *You mention a green channel, but that is not visible.

If you need a specific cytoplasmic marker to outline cells, QuPath doesn’t have that currently built in, and you may need to create something custom/through ImageJ, or perform the segmentation elsewhere and find a way to import it back into Qupath for quantification.

thanks a lot for quick reply!

yeah i did not use a project to open the original images i works very well now with 0.2.0-m8 version

i added another image to the google drive link where the positive detection is following the green channel.

my whole cells and slice is fixed afterwards before staining but that is why i am using dapi for region detection by for positive cell nucleus detection i did not use it because i stains the cells non-specifically


Ah, before I go off on any crazy ideas, I’ll give others/Pete a chance to chime in. One thought though,

Do you really need the cells at all then, or could you use the pixel classifier to quantify the area/percent area covered by “bright enough red” vs “bright enough red+green” ?

but unfortunately this did not work in my hand