Unable to differentiate between different myofiber types in the image


#1

I am trying to identify the copper, green and black color myofibers from my image using cell profiler, and I tried to use identify modules. However, these modules are not able to identify the different colors correctly. I have attached my pipeline to this post. Any help would be greatly appreciated.

Test1.cpproj (765.2 KB)


#2

Hi there,

Please use module UnmixColors instead of ColorsToGray

In my test, you would need 2 separate UnmixColors,

Hope that helps.


#3

Hi,

Thank you so much for your help, is there any way where I can identify the black fibers as well?
I have attached another sample image


#4

The black fibers are tricky, if you want to directly extract them out of a color image.

I would suggest to extract first green and red fibers, then simply do a MaskImage operation to get the black fibers, follow this “equation”

original image - green fibers - red fibers = whatever left = mostly black fibers + some peri/endomysium

:smiley:


#5

Thank you for the help! I had one last question, I want to extract the green and copper and black seperately from the image and measure the area, roundness and nuclei in those fibers, I tried using identify primary, secondary and tertiary object modules but it does not identify the fibers correctly. What would you suggest me to do?


#6

Hi,

I think you may be a bit confused about what primary, secondary etc objects are- hopefully this can help. I’d think all of your objects should be identified as primary objects.