Large-scale EM Benchmark for Multi-Organelle Instance Segmentation in the Wild

TL;DR

This paper introduces a large-scale, diverse EM dataset for multi-organelle instance segmentation, highlighting current model limitations in generalization and global structure modeling in real-world biological data.

Contribution

It provides a new extensive benchmark dataset with expert-refined annotations and evaluates state-of-the-art models, revealing key challenges in EM image segmentation.

Findings

Models struggle with heterogeneity and global structures.

Current methods perform poorly on distributed organelles.

The dataset and tools will be publicly released.

Abstract

Accurate instance-level segmentation of organelles in electron microscopy (EM) is critical for quantitative analysis of subcellular morphology and inter-organelle interactions. However, current benchmarks, based on small, curated datasets, fail to capture the inherent heterogeneity and large spatial context of in-the-wild EM data, imposing fundamental limitations on current patch-based methods. To address these limitations, we developed a large-scale, multi-source benchmark for multi-organelle instance segmentation, comprising over 100,000 2D EM images across variety cell types and five organelle classes that capture real-world variability. Dataset annotations were generated by our designed connectivity-aware Label Propagation Algorithm (3D LPA) with expert refinement. We further benchmarked several state-of-the-art models, including U-Net, SAM variants, and Mask2Former. Our results show several limitations: current models struggle to generalize across heterogeneous EM data and perform poorly on organelles with global, distributed morphologies (e.g., Endoplasmic Reticulum). These findings underscore the fundamental mismatch between local-context models and the challenge of modeling long-range structural continuity in the presence of real-world variability. The benchmark dataset and labeling tool will be publicly released soon.