Efficient 2D neuron boundary segmentation with local topological constraints

TL;DR

This paper introduces a novel segmentation method for neuron membranes in 2D electron microscopy images that enforces topological constraints to improve accuracy and reduce errors, inspired by human gap completion strategies.

Contribution

The paper proposes an ILP-based segmentation approach that incorporates local topological constraints, enhancing neuron boundary detection over standard methods.

Findings

Improved segmentation accuracy compared to standard approaches

Successful gap completion and fewer topological errors

Effective integration of topological constraints into segmentation pipeline

Abstract

We present a method for segmenting neuron membranes in 2D electron microscopy imagery. This segmentation task has been a bottleneck to reconstruction efforts of the brain's synaptic circuits. One common problem is the misclassification of blurry membrane fragments as cell interior, which leads to merging of two adjacent neuron sections into one via the blurry membrane region. Human annotators can easily avoid such errors by implicitly performing gap completion, taking into account the continuity of membranes. Drawing inspiration from these human strategies, we formulate the segmentation task as an edge labeling problem on a graph with local topological constraints. We derive an integer linear program (ILP) that enforces membrane continuity, i.e. the absence of gaps. The cost function of the ILP is the pixel-wise deviation of the segmentation from a priori membrane probabilities derived from the data. Based on membrane probability maps obtained using random forest classifiers and convolutional neural networks, our method improves the neuron boundary segmentation accuracy compared to a variety of standard segmentation approaches. Our method successfully performs gap completion and leads to fewer topological errors. The method could potentially also be incorporated into other image segmentation pipelines with known topological constraints.