Example Pipelines: Cell size and Cardiac fibrosis

Hi all,

I’ve been working with cellprofiler for the past 2 months and spent A LOT of time working on 2 pipelines which I’d like to share with the community.

The first counts the amount of tissue and number of nuclei in H&E so if you divide total tissue area by no of nuclei you get a rough estimate of average cell area.

The second is for people who do not have access to a polarised microscope but still want to quantify the amount of collagen (fibrosis) stained red in heart sections stained with 0.1% fast green and 0.1% sirius red. Divide amount of fibrosis by total tissue size (separate pipeline because I manually remove blood cells as if you mask them it also masks the nuclei)

Enjoy.

Any questions / suggests welcome!
tissuespread.cp (8.97 KB)
fibrosisspread.cp (13.2 KB)


45 final.cp (14.4 KB)

ps the H&E pipeline is designed NOT to detect clusters of nuclei as these are most probably fibroblast clusters. If you want to detect them simply turn on the method to distinguish close objects in identifyprimaryobjects.

In addition to the afformentioned pipeplines I have also written another fibrosis pipeline which works on crysections (as the staining intensities differe from paraffin).

Attached are example images.
There are two pipelines which have to be run one after the other (1 then 2) because for some reason when two images are added together in the first pipeline when thresholds are applied all of the pixels from one of the images is removed regardless of true intensity.

Summary - Run pipeline 1 then pipeline 2 in same input/output folder.

Any Qs just PM me - I also have an optimized protocol for cardiac fibrosis staining if anyone wants it.
Pipeline 2 results.cp (9.17 KB)
add light and dark v1.cp (7.96 KB)


Hi Peter,

Thanks for this. It’s always great when users not only get something out of the software but also feel like contributing back to the community at large! :smile:

Cheers,
-Mark