I am using the cell counter plugin in Fiji (version 2.2.1., I always run the Updater before using Fiji) to manually count two types of immunofluorescence stained cells in two regions of interest. I need to save a trace of the counting with my image and therefore use the “Export image” function supplied in the plugin. However, the markers are TINY, nearly invisible. In “Options”, I found how to change the color of the markers, but not the size. How can I do that?
Welcome to the forum, @cschreiweis!
Please note that the Cell Counter is maintained for backwards compatibility, but much of its functionality is available via ImageJ’s own multi-point selection tool (I just added a notice to the wiki page).
In particluar, its settings dialog provides an option to choose between different marker sizes:
Does that help in your case?
Yes, that helps indeed!
When playing with the multi-pointer and following the help indications, I still cannot figure out how to get my counts for two different cell types for two different regions of interest. I found how to choose my ROIs and my two different pointers, but the color displayed is never the one I selected and the counts in the two different ROIs do not correspond with the table when pressing ‘m’. Why is that?
Also, can I export a “state-of-art” file and continue later with the counting? Do I just do “save as”? When re-opeing my file, the two ROIs are now displayed, but indicated as a single ROI.
Thank you very much!
As the Help button explains, you can File Save As Tiff… to save the image + points, or File Save As Selection… to save just the points without the image (much smaller file typically).
It seems like the color is only used when you have a single counter type. As soon as you start using multiple counter types, the colors revert to a hardcoded order: 0=yellow, 1=purple, 2=cyan, 3=orange, etc. I presume this is to ensure good contrast. If you really need the ability to customize the colors, you could ask @Wayne for this feature.
In my quick test, the Results table does indeed have a number of rows equal to the total number of points. But it does not include a column indicating the group to which each point belongs. The order appears to be simply the order you clicked the points, meaning that all points in a particular group may not be adjacent in the table. I agree that improving this would be useful.
Thank you very much for agreeing that an improvement of the tool would be useful. Talking to our data analysis manager of our imaging facility, I get the impression that many users would be happy to have the option to count two different types of entities in more than one ROI and get the exact count of each subpopulation. Would you be the one who could take care of this? I would be really really grateful. If yes, how fast would that be possible? I have a huge dataset to go through and would love to start asap…
I think you can already do that… just press T to add the multi-point to the ROI manager. Then start a new multi-point for any subsets of points which need it. The downside is that the colors will overlap. But you can then measure only the multi-point ROI(s) which are relevant, which should make the Results table more clear.
No, the Multi-Point Tool is part of ImageJ 1.x; @Wayne is the developer. See the ImageJ1 wiki page for details. The project is open source, however—if you have a developer at your facility able to work on these improvements, you could submit a patch to Wayne.