Membrane marker segmentation

[ambeeblush]

Hi there!

First of all, thank you very much for this amazing tool for whole slide data analysis, we are loving it!

Just a little issue I cannot figure out, when performing the segmentation step, when I specify both Nuclear marker (DAPI) and Membrane marker (PanCK) I get an image output (segmentation_mask.tif or segmentation_data.npy) showing just nuclear masks, is there an output file showing both nuclear and membrane segmentations as separate channels? It would be super useful to check segmentation parameters

Thank you in advance for the help!
Ambra

[OmFrederic]

Hello ambeeblush,

Sorry for taking so long to answer this, it seems I am not good at keeping track of the repo issues.

The final segmentations files you are referring to (segmentation_mask.tif or segmentation_data.npy) should contain the combined nuclear and membrane segmentations. If they only look as nuclear segmentations it might be caused by different things:

• The membrane marker was not very good/did not work well with the watershed algorithm
• Some membrane settings were not correctly imputed

To troubleshoot this, you can always check the quality check images that you can locate in the analysis/quality_control folder. There you should see:

• The direct result from stardist segmentation (only nuclei)
• The expanded stardist result
• The membrane watershed result for each tile of the image (only membrane)
  This way you can see what has been done

Just in case is a parameter mismatch:

• If you are using the GUI, you are probably setting everything right. Just maybe check that the membrane marker image really ends with that marker name before the extension.
• If you are using the command line approach (pipex_batch_list.txt), make sure you adding at least the membrane_marker and membrane_diameter parameters