Fischer, Christian, Haeussler, Simon, Marr, Carsten, Duchen, Michael, Conradt, Barbara, Rolland, Stephane, Singh, Kritarth, Besora-Casals, Laura
[
International Worm Meeting,
2021]
While the analysis of mitochondrial morphology has emerged as an important tool in the study of mitochondrial function, efficient quantification of mitochondrial microscopy images presents a difficult task and bottleneck for statistically robust conclusions. Here, we present the Mitochondrial Segmentation Network (MitoSegNet), a pretrained deep learning segmentation model that enables researchers to easily exploit the power of deep learning for the quantification of mitochondrial morphology (Fischer, Besora-Casals et al. 2020). The MitoSegNet was generated by training a modified fully convolutional neural network with fluorescent microscopy, maximum-intensity projection images, depicting mitochondria in body wall muscle cells of adult C. elegans worms. We tested the performance of MitoSegNet against three feature-based segmentation algorithms and the machine-learning segmentation tool Ilastik. MitoSegNet outperformed all other methods in both pixelwise and morphological segmentation accuracy. We successfully applied MitoSegNet to unseen fluorescence microscopy images of mitoGFP expressing mitochondria in wild-type and catp-6ATP13A2 mutant C. elegans adults. Additionally, MitoSegNet was capable of accurately segmenting mitochondria in HeLa cells treated with fragmentation inducing reagents. We provide MitoSegNet for all operating systems as an easy-to-use graphical user interface tool that combines segmentation with morphological analysis. Reference Fischer, C. A., L. Besora-Casals, S. G. Rolland, S. Haeussler, K. Singh, M. Duchen, B. Conradt and C. Marr (2020). "MitoSegNet: Easy-to-use Deep Learning Segmentation for Analyzing Mitochondrial Morphology." iScience 23(10).