Calculating capsule to cell size ratio using CellProfiler

I tried to search for similar topic but couldn’t find a relevant one that I can adapt to solve this problem. I’m sorry if similar topic has been answered! I’m also new to CellProfiler and I’m somewhat familiar with pipeline for calculating yeast patch size.

Attached is an example DIC microscopy image of india link stained Cryptococcus neoformans (fungal pathogen) cells. I’m trying to calculate the capsule (visualized by light “empty” space) to cell (inside the capsule) size ratio and to quantify this in many images.

I was somewhat able to identify the capsule using IdentifyPrimaryObjects module and calculate its size. However, I’m lost how I should be identifying the cell inside the capsule. I tried IdentifySecondaryObject I think a better way is to identify the cell inside then identify the capsule outside, but I cannot quite figure out how to get the IdentifyPrimaryObjects module to identify the cell inside the capsule.

Another line of thought was from the following link:


I’m again not sure how I can get the IdentifyPrimaryObjects module to identify both capsules and cells at the same time.

Please see the attached example image.

Thank you!

David

If the cell is brighter and the capsule segmentation is good, you could mask your image into just the capsules and then do IDPrimary for the cells- in theory, it should work with some tweaking.

All that to say, direct processing of DIC images isn’t always great in CellProfiler- running the images first through a pixel classifier like Weka or ilastik and THEN doing segmentation in CellProfiler is more typically our go-to advice.

Thanks for the reply! I will give ilastik a try on my DIC images then put it through CellProfiler. I will report back with the results!

Thanks again

David