Fig. 3

Three different approaches used to quantify mitochondrial morphology. The approach A uses 3 types of images: binary image (BI) of segmented mitochondria, fluorescence maximum intensity projection (MIP) image and differential interference contrast (DIC) image and ImageJ/Fiji analyses run by macro 2. Output parameters can be expressed as values/mitochondrial fragment or values/cell. The output parameters in black are those used for consequent quantitative analysis of mitochondria in the mmi1Δ and wild-type strains. The parameters in gray are other parameters that can be calculated by the analysis. The approach B uses BI of segmented mitochondria and fluorescence MIP image as input images for MitoA analysis that calculates output parameters as values/image. The approach C uses BI of segmented mitochondria and fluorescence MIP image and ImageJ/Fiji analysis run by macro 3 that calculates output parameters as value/mitochondrial fragment. mtGFP-mitochondrial target GFP, MitoS_yeast (segmentation by the deep learning model MitoSegNet retrained for yeast cells), and MitoA (morphological analysis in the MitoSegNet analysis toolbox). The figure was created with BioRender.com.