Plugin similar to Imaris "Channel Arithmetics"?

Hi Everyone,

I am interested in setting up histocytometry for my project (and eventually CytoMAP). However, we are working on installing MatLab to the Imaris workstation so I currently do not have access to the “Channel Arithmetics” Xtension that the Gerner Lab uses for generating a spillover/compensation matrix. Does anyone know of a plugin/macros script for FIJI/ImageJ? (I don’t think Lumos is quite what I need).

Alternatively, is it acceptable to quantitate all channels in single fluorophore slides, export to csv, and perform compensation in FlowJo, then apply comp matrix to actual sample slides processed in the same way? Any thoughts @cstoltzfus?

Thanks!!!
Jaclyn

Hello @JMarie,

Imaris has a preinstalled channel arithmetic function. You might have to go into settings and enable the MATLAB runtime which comes with Imaris and does not require a separate license. My extension adds the ability to run multiple functions sequentially, and have Imaris save the image after completing the function so it is more for convenience and not strictly necessary. I haven’t used FIJI/Imagej in a while so I do not know what tools they have for adding/multiplying images together.

I am not 100% sure what you are asking here. If you are doing spectral devonvolution on the microscope channels, we do that at the pixel level before segmenting the images. What I usually import into FlowJo is the average fluorescence across each segmented cell object for each channel. I think compensating for channel overlap in FlowJo can get tricky since you could have overlap in your segmented cell objects which might look like spectral overlap. This would make it hard to tell if “spill” between two channels like CD8 and CD11c was an artifact of spectral overlap when imaging or an artifact from imperfect image segmentation leading to some signal from Dendritic cells “spilling” over into neighboring CD8 T cells.

1 Like

Thank you so much for your quick reply @cstoltzfus ! I apologize, my question was very naïve, and I was just about to remove this Q. I really appreciate you taking the time to help.

I did try to make a compensation matrix through FlowJo and what you described happened. I am in the screening phase of these experiments (multiple panels, and multiple mouse conditions), so I have been tiling 2D images just to get an idea of what the data says before I run huge z-stack tiled images. (Was curious if there was an easy way to do this without Imaris since I don’t have it on my personal PC, only in the core). I will just try the workflow you and M Gerner outlined in your papers and slightly modify it for 2D. By the way, the figures generated with CytoMAP look amazing, nice work! I look forward to integrating CytoMAP into our analysis.

Jaclyn

2 Likes

I am glad you asked your question! I have been asked similar things many times. I think the process of segmenting images and understanding what the resulting data really means and what we can do with it is something I am trying to understand better.

There are a ton of really nice tools for looking at these imaging datasets. You could try QuPath which is free, has a lot of walkthroughs, and works well on my laptop for these imaging datasets. There are a lot of people on this forum using it and they are very helpful. @Mike_Nelson even put together a nice tutorial for generating data in QuPath then importing it into CytoMAP: There and back again, QuPath<==>CytoMAP cluster analysis

1 Like