Count particles per cell

I have done a Transferrin Uptake Assay in BT549 breast cancer cells using Transferrin-Alexa488 and I have quantified them using ImageJ and mean fluorescent intensity per cell.

I want to now try and quantify the amount of particles per cell but I am having a difficult time. I cannot find any plugins or macros for ImageJ but have been told to try CellProfiler.

I tried CellProfiler and

I have attached one of my images (it is a dual layered .tif file). I have attempted to use this image with the “Cell/particle counting, and scoring the percentage of stained object” pipeline,
as seen on this webpage … ntPositive

But it does not seem to work on my images.

I was wondering if you could provide any guidance on the best approach to quantify the number of particles in each cell of my images.



Give the attached pipeline a try. Also, did you see this example pipeline?

The basic idea is to identify Cell regions and Speckles independently, then relate them by position (and counting is done automatically at that point).
(1) ColorToGray: Split the color channels
(2) Smooth – both your input channels are not quite smooth biologically. The Smooth module try to help (though they may be tweaked more!)
(3) IdentifyPrimaryObjects - ID nuclei
(4) Smooth again, same as above (worse “speckliness” for Green channel, for obvious reasons)
(5) IdentifySecondaryObjects: “Grow” cytoplasm region from the nuclei
(6) EnhanceOrSuppressFeatures: Use a tophat (“Speckles”) filter to highlight speckles, and minimize background noise
(7) IdentifyPrimaryObjects: ID Speckles
(8) RelateObjects relates each Speckle object to its “parent” Cell.

Look for CountSpeckles (or something like that) in the Cells output CSV which should contain the counts per cell.

Hope that helps!
DLpipe.cppipe (13 KB)

Thank you very much for your help!

This is pretty much exactly what I was looking for.

I was just wondering how would I find the area of each individual cell? I am able to find the mean and median areas of all the cells in the image but the cells are all different sizes. It can find the number of speckles per cell now but since the cells are all different sizes I want be able to divide the number of speckles by the area of the cell to get a speckles/unit count so it would be comparable between cells of different sizes (I have numerous pictures so my n would be rather large).

I was wondering if this was possible.

Thanks for your help!

Yes, you can access the area of each individual cell in the output Per_Object CSV. You will only see the aggregated (mean, etc) across objects in the image while you are processing the data in, say, Test Mode. But ExportToSpreadsheet will save the per-object data (including area) in the “Per_Cells.csv” file. You do need to add MeasureObjectSizeShape module first, somewhere between IdentifySecondaryObjects and ExportToSpreadsheet to do the actual measuring.

Hope that helps,