H & E stain pipeline for muscle cells


#1

Hi my name is Damon Cargol and I’m an employee of the University of Washington’s department of biochemistry. Richard Frock and I came up with a pipe for quantitating the number of nuclei per area in H & E stained muscle cells (see the description below and attached pictures). We thought it might be useful to others, like your example page, and wondered if there was a specific place to upload files for the community of cellprofiler users. I’ll attach it here too.
Description-
“We were interested in attempting to quantify our Hematoxylin and Eosin stained skeletal muscle sections in a fast and unbiased way. In particular, we focused on quantitating total nuclei in an image of a given muscle section by dividing out the undesired features and retaining only the nuclei. We also wanted to make a comparison against similar tissue of different genetic backgrounds and given that each image represents an amorphous area of tissue with possible gaps or ends displayed, we were able to normalize the nuclei count based upon the pixel area that was analyzed within each image.” –R. Frock

Sample Images -------------------
[attachment=0]03_nuclei_count.jpg[/attachment]
[attachment=1]02_in_process_before_inversion.jpg[/attachment]
[attachment=2]01_sample_image.jpg[/attachment]







#2

Here is the rest of it, guess I can’t attach a pipe file!

[attachment=0]04_area_count.jpg[/attachment]



#3

Hi Damon,

It’s great to hear that you’d like to contribute to the Cell Profiler community, and we’d certainly be interested in checking out more of your images.

Do you have a site where we can download both the image set and your pipeline so our team can double-check the output ourselves?
-Mark


#4

Hi Mark, yeah absolutely.

You can download everything from here:
zoolcube.com/lab/cellprofiler/

Sorry for the lack of a description or anything, it’s just a directory. The pipeline is there and some sample images are divided up into the WT and KO directories.

Let me know if you have any questions.
-Damon


#5

Hi Damon,

Sorry it’s taken so long to review this, but your pipeline looked like it was pretty good as it stands. We’d have to discuss further to see if we’ll include it as an example image set/pipeline, but thanks for the contribution!

Attached is the (slightly revised) pipeline with some extra modules added. We took some time to see if we could improve nuclei segmentation, but it’s hard to assess without a better feel for the cell physiology. If you have any more comments, please let us know.

Regards,
-Mark
2008_06_23_AnalysisPIPE.zip (1.72 KB)