the fixing step itself causes varying degrees of stress and may distort the mitochondrial morphology in nature/living conditions. I would say, if possible, it's good to complement your EM results with living cell imaging at high resolution. If you stain the organelles with appropriate trackers (e.g. MitoTracker Deep Red) and use optimized parameters, you can reconstruct dynamics in 3D, and use softwares to measure and quantify things in 3D or 2D. example: Imaris.

Dear Nai-Kei Wong

Thank you so much for your wise answer and suggestions.

My situation, however, is that I don't have anything but my TEM acquisitions from leukemia cells.

To what extent , do you think, the TEM can distort the native morphology of the mitochondria?

And in my case what segmentation approach do yo recommend? I started doing it manually saving ROI set for each picture, and measure the particles..

This is of course only for area , circularity, elongation , etc.

Avigail Rein and Yuntian Guan

I can suggest Dragonfly (www.theobjects.com) from Object Research Systems, it is free for non commercial use.

Dragonfly has many segmentation tools from active shape to deep neural network. The following video, (https://www.youtube.com/watch?v=1WVlskyuw94) is a deep learning tutorial.

You can then use the Object Analysis plugin in Dragonfly to get the measurements you need.

Do not hesitate to contact me if you want more information.