Counting non-cellular features

I am analysing IHC (Hematoxylin/DAB) images of ovaries. The circular white spaces are oocytes and if the oocyte is viable, there will be a brown object inside (viable follicle). Just not sure how do I go about using Qupath to count these objects.


The easiest way would be to use manual counting:

Anything else would require devising a custom analysis pipeline in QuPath. This might be possible using a custom ImageJ macro and/or QuPath’s thresholding or pixel classification (e.g. creating filled objects from the whitespace, then calculating intensity features within these objects and classifying them) – I haven’t worked on a project like this myself so don’t really know how hard it would be, but that’s what I’d try first if manual counting wasn’t viable.

1 Like