Hello,
I am trying to use CellProfiler to measure the size and shape of my cells (imaged on brightfield inverted microscope, stained with crystal violet). I use the "IdentifyPrimaryObjects" feature in my pipeline to identify my cells, but it seems to think that the cells are the background and vice versa. Has anyone had this problem? How can I fix it? Any advice would help.
Thanks!
Disclosure: I have never used CellProfiler, but this sounds to me like the "IdentifyPrimaryObjects" finds bright objects on a black background (i.e. fluorescent images), which would cause it to think that your objects (dark objects on a bright background) are background. If this is indeed the problem, one simple solution would be to invert the image prior to object detection.
Alternative solution: Using the "Versatile (H&E nuclei)" model of StarDist. This worked well for me on Nissl stained brain slices. I used Fiji, but there also seems to be a StarDist plugin for CellProfiler.
- Badges
- Science method