Automatic quadrant annotations around positively detected cells

Image Analysis
#1

Hi,

I’m a university student doing digital image analysis for the first time so would be very grateful for help.

I have run a positive cell detection on my sample and now want to analyse the microenvironment around these positive cells. I have been manually expanding circles around positive cells of interest and splitting them into quadrants (points tool, expand annotation by 100µm, draw 2 rectangles to split the circle into quadrants, subtract these from the area, split annotation). The idea behind the quadrants is that some positive cells are on the periphery of the tissue sample and so I would negate quadrants with area beyond the tissue border.

This is quite tedious, even with duplicating and moving annotations. I wondered if anyone knew of any scripts I could modify so that QuPath could randomly select a set number of positive cells and do this process for me?

Happy to accept any advice.

Best wishes

#2

Just curious, but would it be simpler to check all of your positive cells and get the makeup (count) of all cells within X distance?

Why would you negate empty quadrants? They should not affect the ratio of local cells, unless you are studying density, perhaps?

Alternatively, you might want to look into neighborhood analysis with CytoMAP.

If you really want to dig into the coding, and you have a tissue annotation, you can probably either find the intersection between the tissue and an annotation created around a cell, and check cells within that area, or you can try and manipulate one of the older tissue segmentation by depth scripts.