SSHMT: Semi-supervised Hierarchical Merge Tree for Electron Microscopy Image Segmentation

TL;DR

This paper introduces a semi-supervised hierarchical merge tree method for electron microscopy image segmentation that reduces the need for extensive ground truth data while maintaining high accuracy.

Contribution

It develops a differentiable unsupervised loss and a Bayesian model to combine supervised and unsupervised learning for segmentation.

Findings

Achieves near state-of-the-art performance with only 3-7% of labeled data.

Significantly outperforms fully supervised methods trained on the same limited data.

Demonstrates effectiveness across three EM datasets.

Abstract

Region-based methods have proven necessary for improving segmentation accuracy of neuronal structures in electron microscopy (EM) images. Most region-based segmentation methods use a scoring function to determine region merging. Such functions are usually learned with supervised algorithms that demand considerable ground truth data, which are costly to collect. We propose a semi-supervised approach that reduces this demand. Based on a merge tree structure, we develop a differentiable unsupervised loss term that enforces consistent predictions from the learned function. We then propose a Bayesian model that combines the supervised and the unsupervised information for probabilistic learning. The experimental results on three EM data sets demonstrate that by using a subset of only 3% to 7% of the entire ground truth data, our approach consistently performs close to the state-of-the-art supervised method with the full labeled data set, and significantly outperforms the supervised method with the same labeled subset.