Muscle fiber min Feret analysis

Hello,
I am a complete novice to Cell profiler and I downloaded it because of a recommendation from the treat NMD website. I have a very large number of fluorescent images from skeletal muscle cross sections to analyse and there will be many hundreds more to come as I move forward. Therefore I am looking for a semi automated method to assess several parameters from these images which are taken sections (see attached images) stained with Lamin (red; to show the outline of the muscle fiber), Pax7 (green: to detect satellite cells) and DAPI (blue; to show nuclei). Here are the 5 parameters I would like to measure:

  1. Muscle fiber minimum Feret (a measure of diameter) using the lamin staining
  2. The number of Pax7 nuclei out of the total number of DAPI positive nuclei
  3. The total number of DAPI postive nuclei
  4. The number of fibers with a DAPI positive nuclei in the middle of the fiber (a sign of muscular regeneration)
    Any help in figuring out pipelines to identify these different muscle feature (from single or multiple images) would be a massive help and I would really appreciate your advice.
    Thank you in advance for your time.
    Best,
    Philippe



Hi Philippe,

Based on your images, I think a couple of example pipelines from our Examples page might be a good first start:

  • Tissue Neighbors:
    This is a retinal epithelial cell assay but the cellular characteristics seem similar to yours. - Cell/particle counting, and scoring the percentage of stained objects:
    The approach describe here is a good basis for finding cells positive (or negative) for a stain of choice.
    Naturally, the pipelines listed were optimized for their particular images. Some tweaks will probably probably be needed for your particular assay. Hope this helps!
    -Mark

As an aside: I’m not familiar with the TREAT_NMD website. Would you mind posting a link to where you find the CellProfiler reference?
-Mark

Thanks for all the information. We are going to get started using your advice and let you know how it goes.

Concerning the Treat_NMD reference, they actually refer to CellIP (from Olympus) not cell profiler.

Thanks again

Mark-
Using your suggestions I tweaked the existing pipeline with my image. I have attached an image plus my pipeline. I have 1 issue and 1 question.

Issue: I would like to adjust the way that cells are picked such that the cell outlines overlap with the fiber staining better. I tried to change the thresholding methods but didn’t get much improvement.

Question: I would like to generate individual min ferets for each cell that is identified (or outlined). At the end I see a table with means, medians, etc. Is there a way to general a list of min ferets per object (a surrogate for fiber diameter).

Thanks


Fibers.cp (10.6 KB)

[quote=“mbray”]As an aside: I’m not familiar with the TREAT_NMD website. Would you mind posting a link to where you find the CellProfiler reference?
-Mark[/quote]

Mark-
Using your suggestions I tweaked the existing pipeline with my image. I have attached an image plus my pipeline. I have 1 issue and 1 question.

Issue: I would like to adjust the way that cells are picked such that the cell outlines overlap with the fiber staining better. I tried to change the thresholding methods but didn’t get much improvement.

Question: I would like to generate individual min ferets for each cell that is identified (or outlined). At the end I see a table with means, medians, etc. Is there a way to general a list of min ferets per object (a surrogate for fiber diameter).

Thanks


Fibers.cp (10.6 KB)

I would change the following in IdentifyPrimaryObjects:

  • Thresholding method: 3-class Otsu with middle class set to Background
  • Threshold correction factor: 1.0
  • Size of smoothing filter: 10
  • Maxima separation distance: 10

In general, the higher the smooth filter size/maxima separation distance, the less inappropriate objects splits you will get. If you need to refine things, you may want to try fiddling the smooth filter size first.

The feret diameter is not a measurement avilable with the current release. However, we have added it into our code and it will be available for the next release. In the meantime, see this post for details and try out the trunk build mentioned there to get the Feret measures.

Regards
-Mark