Loading one big image file containing thousands of separate multi-channel images in CellProfiler

Hi all,

I’m looking for the current best practices for loading a big file containing thousands of separate multi-channel images, and producing measurements for the separate images. The images are independent from each other, I want to do this purely for the benefit of not having to manage thousands of actual separate image files on my filesystem.

In particular, I have imaging flow cytometry data, and want to load all the single-cell images from an experiment from one file. I know of the tiling approach proposed here, but this still requires hundreds of separate files.


Kind regards,

Hi @mlippie

CellProfiler is capable of handling a single large file, provided the format is supported by Bioformats (most microscopy formats are). To properly extract the individual images you’ll need to tell the Metadata module to extract from image file headers, then create some rules in the NamesAndTypes module to match metadata parameters which label each channel, etc in order to pair the unpacked images. Without seeing the file you have it’s hard to give specifics, but this general approach should work.

Hope that helps

Thank you, that helped!

For future reference, I had to increase the available Java memory space under the CellProfiler preferences to make it work.