Transform a 3d stack image with multiple channels into a 3D plot

Hello everyone,

I have a 3d stack image with 2 channels (green and red), with each channel representing a different cell population. These 2 populations are interacting spatially together to form some type of spheroid structure. Anyway, I would like to transform the stack into a 3D plot in order to do some statistical analysis between the two populations. I have seen a lot of questions about going from a 3D plot to a 3D image, but rarely the other way around. Ideally, I would like to extract a matrix from each stack that has the x, y, z, and intensity of every pixel.

Thank you so much in advance.

[DGM]

How is the data arranged? The common conventions for an image is either [rows columns channels], [rows columns slices], or [rows columns channels slices], or [rows columns slices channels]. Note I never mentioned X, Y, or Z.

Plotting a volumetric intensity image is simply problematic at a concept level. How do you visualize the variations in density in a block of wood? Without reducing the problem, you can't. The object occludes itself. Even if it were transparent, it often becomes difficult to interpret the alpha-blended result -- especially for colored images. You might be able to reduce the problem by reorienting the data and producing slice images along whatever plane suits the visualization. Alternatively, if you can reduce the content to a single-channel volumetric intensity image, you could use the volume viewer app or isosurface to visualize it.

I would like to transform the stack into a 3D plot in order to do some statistical analysis between the two populations.

Plots are outputs -- visualizations of data. You do statistical analysis on the data itself, so I'm not sure what this implies about how you want to plot the result.

[Soukaynah El Hajj]

@DGM I do not have the data yet, I only have the images, I would like to extract the data. I just wanted to see if there is a pre-existing function or a direct way to extract the coordinates of the pixels without having to extract all the slices, put them in a repository, and then loop through them. I would like to end up with something such as [row columns slices intensity] for example. Pixel intensity can be helpful for me later on if I conduct any spatial distribution of intensity analyses or heatmaps.