MitoEM 2.0 - A Benchmark for Challenging 3D Mitochondria Instance Segmentation from EM Images
Identifier: S-BIAD2808
Published: 2026-01-19 Licence: CC0 Publisher: BioImage Archive
We present MitoEM 2.0, a curated benchmark resource for training and evaluating three-dimensional (3D) mitochondria instance segmentation in volume electron microscopy (vEM). The collection assembles multiscale vEM datasets (FIB-SEM, SBF-SEM, and ssSEM) spanning diverse tissues and species, with expert-verified instance labels emphasizing biologically difficult scenarios, including dense mitochondrial packing, hyperfused networks, and thin filamentous connections with ambiguous boundaries. All releases include native-resolution volumes and standardized processed versions, per-volume metadata (voxel size, modality, tissue, and data splits), and official train/validation/test partitions to enable reproducible benchmarking. Annotations follow a consistent protocol with quality checks and instance reindexing. Data are provided in NIfTI format with an nnU-Net–compatible layout, alongside machine-readable split files and checksums. MitoEM 2.0 facilitates robust model development and fair comparison across methods while supporting reuse in bioimage analysis, algorithm benchmarking, and teaching.
Imaging Methods: electron microscopy focussed ion beam scanning electron microscopy (FIB-SEM)
Organisms: Homo sapiens Mus musculus Drosophila melanogaster