Getting error while Analysing Images in CellProfiler

Sample image and/or code

Images.zip (18.1 MB)

VER4 PIPELINE.cppipe (31.9 KB)

ErrorLog from console.txt (4.0 KB)

Images.zip file contains the actual tiff images.

VER4 PIPELINE.cppipe is the actual pipeline, which gives error.

ErrorLog from console.txt is the exported log from the concole window

The image is the final message we got after completing the whole process

Background

These images are about tumour cells and immune cells.

Analysis goals

To assess type of immune micro environment around tumour cells

Challenges

  • What stops you from proceeding?
    The analysis is completed, but we are unable to generate the CSV files in the output folder. We are almost trying this since a year. But, getting an error in the console window (attached in “ErrorLog from console.txt”). We are not sure why the CSV files are not getting generated. We need these CSV files to process it in FlowJo (another software).

  • What have you tried already?
    We tried it using 2.2.0, 3.8, 3.8, 3.9 and 4.0.2 versions.
    We tried in MAC and Windows machines.
    We tried the example pipeline provided by the CellProfiler, which is working normally and generating the CSV files as expected.

  • Have you found any related forum topics? If so, cross-link them.
    We are novice and no idea

  • What software packages and/or plugins have you tried?
    We don’t have any other plugins of packages.

Hi @Raja_Balu,

Thanks for uploading the pipeline, images and error log. It looks like your pipeline is capturing excessive numbers of objects, which combined with the 45 channels is exceeding the storage capacity of CellProfiler’s internal measurements buffer.

From a quick look at the images, I’m not sure what exactly you’re trying to detect? The IdentifyPrimaryObjects module picks up a lot of tiny objects, but they’re so packed together that the IdentifySecondaryObjects and IdentifyTertiaryObjects modules are almost irrelevant. You might want to verify that the current detection parameters are detecting what you actually desire.

If this is indeed what you want to measure (all 3 object sets), you might want to try limiting the number of wavelengths which are measured in a single analysis run. It should be possible to do 2 analysis runs with half of the wavelengths measured in each, then combine the resulting data.

Hope that helps.

1 Like

Hi @DStirling,

Thank you so much for your help! Much appreciate it :slight_smile:

I am unaware that we cannot process excessive number of images. As you said, I have processed everything separately. Then it produced the CSV files.

All good now. Once again, thank you very much for your help!

Cheers

Glad to hear it, I think the issue here is simply the large numbers of measurements rather than the actual quantity of images. One option to consider for the future might be to switch to using the ExportToDatabase module rather than ExportToSpreadsheet, since a database is written as the run progresses and therefore the measurements don’t all need to be stored within the buffer throughout the experiment.

1 Like

Sure! I will consider using ExportToDatabase.

Thank you once again!