Cell Counter - ImageJ and Image Subset

Image Analysis
#1

Has anyone figured out a way to count your image using cell counter, and then query subsections of the image to get “local” counts?

My lab analyzes atherosclerotic lesions and we will count and classify cells in the entire lesion, but then we are interested in subregion counts. I would like to establish a count and then be able to query different areas/ROIs.

I’m thinking I could write something that overlaid the counts on a blank space and ID the color code of each classification and use the particle counter to identify the dots created by the counter to generate new counts, but hoping someone has already solved this problem…

#2

Not 100% sure, but if your objects are fairly easily detectable, you may want to look into something like QuPath that supports generating regions and then performing cell detection within them. How effective the cell detection itself is may vary depending on the type of cell or staining, though.

If you can detect the objects with ImageJ (basic), you should be able to use similar code in QuPath.

You could, for example, use a pixel classifier to define the annotations like the box he draws at that time point, then perform cell detection within those regions. Counts are automatically generated for annotations based on what is “in/under” them.

#3

Use the built in multi-point tool to count and classify the cells, type “b” (Image>Overlay>Add Selection) to move the resulting point selection to an overlay, outline the area of interest in the image and then use the new Image>Overlay>Measure Overlay command. You will need to upgrade to the ImageJ 1.52h12 daily build, which adds the “Measure Overlay” command.

2 Likes