Efficient Connectivity-Preserving Instance Segmentation with Supervoxel-Based Loss Function

Anna Grim, Jayaram Chandrashekar, Uygar Sumbul
1/2/2025
cs.CVq-bio.NC

No official code found. Best implementation: AllenNeuralDynamics/supervoxel-loss (13 stars, MIT, last commit 2026-04-21).

Abstract

Reconstructing the intricate local morphology of neurons and their long-range projecting axons can address many connectivity related questions in neuroscience. The main bottleneck in connectomics pipelines is correcting topological errors, as multiple entangled neuronal arbors is a challenging instance segmentation problem. More broadly, segmentation of curvilinear, filamentous structures continues to pose significant challenges. To address this problem, we extend the notion of simple points from digital topology to connected sets of voxels (i.e. supervoxels) and propose a topology-aware neural network segmentation method with minimal computational overhead. We demonstrate its effectiveness on a new public dataset of 3-d light microscopy images of mouse brains, along with the benchmark datasets DRIVE, ISBI12, and CrackTree.

View on arXivView PDF

Code Implementations(1)

AllenNeuralDynamics/supervoxel-lossCommunity100%
MIT

Efficient connectivity-preserving loss function for training neural networks to perform instance segmentation.

132Python, MakefileJan 20, 20246 months agoMIT
digital-topologyneuron-segmentationsupervoxelstopological-loss-function

Similar Papers

Disconnect to Connect: A Data Augmentation Method for Improving Topology Accuracy in Image Segmentation

84% match1 repo52025

Learning Multimodal Volumetric Features for Large-Scale Neuron Tracing

83% match1 repo92024

Topology-Guaranteed Image Segmentation: Enforcing Connectivity, Genus, and Width Constraints

83% match2026

Boosting 3D Neuron Segmentation with 2D Vision Transformer Pre-trained on Natural Images

83% match2024

Skelite: Compact Neural Networks for Efficient Iterative Skeletonization

83% match1 repo102025

DINeuro: Distilling Knowledge from 2D Natural Images via Deformable Tubular Transferring Strategy for 3D Neuron Reconstruction

82% match2024

Cite this paper

@article{grim2025efficient,
  title  = {Efficient Connectivity-Preserving Instance Segmentation with Supervoxel-Based Loss Function},
  author = {Anna Grim and Jayaram Chandrashekar and Uygar Sumbul},
  year   = {2025},
  eprint = {2501.01022},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2501.01022v3}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2501.01022.svg)](https://codeofpaper.com/paper/2501.01022)

Discussion