Analyze mito morphology: problem with internal network not being quantified with Analyze Particles

I am trying to analyze a mitochondrial network and have not had success implementing existing plugins. I have attempted to measure aspect ratio and form factor from the parameters measured using the Analyze Particles function, but it seems the inner portion is not being included in the analysis. Has anyone else overcome this hurdle?

I have included an image that documents my work flow. Any help would be greatly appreciated

It seems your workflow analyze only the boundary of the particle.
I suppose you wish the area of the whole particle, including the hole, and the white dot in the middle? In that case, adding a “fill hole” operation after the binarisation, then applying “Analyze particle” (without skeletonization) should give you the result. Maybe you need to invert the binary image, depending on your settings.

You can check the result by using the show option to display “Count mask” -> you get black image with labeled particles. If correct, you should get only one region with label 1 (use image->adjust->brightness/contrast to visualize)

Aspect ratio and form factor should be analyzed on the binary image, not the skeleton, because the skeleton has lost the original shape information.

A drawback of looking at aspect ratio is illustrated by your example – when the mito loops back on itself the software sees a ring, not a tube. You will need to analyze in 3D to capture all details of mito. shape. This is tough, but some (e.g. Susanne Rafelski) have done it.

Maybe there is another, easier measurement that will enable you to test your hypothesis. For example, if you’re interested in whether mitos are tubular vs fragmented you can use the skeleton length (= area in analyze particles).

1 Like