A Context-aware Delayed Agglomeration Framework for Electron Microscopy Segmentation

TL;DR

This paper introduces a novel context-aware, delayed agglomeration framework for electron microscopy segmentation that improves accuracy by separately clustering regions and postponing certain merge decisions.

Contribution

The paper presents a new agglomerative segmentation method that incorporates context-awareness and delayed merging to enhance neuron boundary detection in EM images.

Findings

Significant improvement in segmentation accuracy over existing methods.

Effective handling of over-segmented regions in 2D and 3D datasets.

Enhanced confidence in boundary predictions through delayed merging.

Abstract

Electron Microscopy (EM) image (or volume) segmentation has become significantly important in recent years as an instrument for connectomics. This paper proposes a novel agglomerative framework for EM segmentation. In particular, given an over-segmented image or volume, we propose a novel framework for accurately clustering regions of the same neuron. Unlike existing agglomerative methods, the proposed context-aware algorithm divides superpixels (over-segmented regions) of different biological entities into different subsets and agglomerates them separately. In addition, this paper describes a "delayed" scheme for agglomerative clustering that postpones some of the merge decisions, pertaining to newly formed bodies, in order to generate a more confident boundary prediction. We report significant improvements attained by the proposed approach in segmentation accuracy over existing standard methods on 2D and 3D datasets.