MitoSeg: Mitochondria Segmentation Tool

TL;DR

The paper addresses automatic segmentation of mitochondria boundaries in electron tomography images to study cristae structure and disease associations. It introduces MitoSeg, a non-learning, phase-based C++ tool that combines preprocessing, ridge detection, energy mapping, curve fitting, balloon-snake segmentation, and validation to produce 2D boundaries and 3D meshes. It provides a configurable, multicore, Docker-enabled workflow with outputs in .ply and .mod formats suitable for external tools like MeshLab and IMOD. The approach enables rapid, automated dataset generation for mitochondrial analysis across modalities, facilitating downstream analyses of disease-related morphology without requiring training data.

Abstract

Recent studies suggest a potential link between the physical structure of mitochondria and neurodegenerative diseases. With advances in Electron Microscopy techniques, it has become possible to visualize the boundary and internal membrane structures of mitochondria in detail. It is crucial to automatically segment mitochondria from these images to investigate the relationship between mitochondria and diseases. In this paper, we present a software solution for mitochondrial segmentation, highlighting mitochondria boundaries in electron microscopy tomography images and generating corresponding 3D meshes.

Figures (3)

• Figure 1: Flowchart of the algorithm.
• Figure 2: Outputs. (a) Boundaries from slice 35, (b) boundaries from slice 55, (c) final output (.ply file), (d) final output (.mod file).
• Figure 3: Number of utilized CPU cores vs (a) time performance, (b) speed-up.