Bridging the Gap: Point Clouds for Merging Neurons in Connectomics

TL;DR

This paper introduces a novel point cloud-based method using CurveNet to merge neurons across missing image sections in connectomics, improving accuracy and scalability over existing approaches.

Contribution

It presents a new point cloud representation approach with CurveNet for neuron merging, demonstrating high efficiency and scalability in handling large gaps.

Findings

High accuracy in neuron merging across gaps

Efficient data usage with point cloud representations

Scalable performance beyond existing methods

Abstract

In the field of Connectomics, a primary problem is that of 3D neuron segmentation. Although deep learning-based methods have achieved remarkable accuracy, errors still exist, especially in regions with image defects. One common type of defect is that of consecutive missing image sections. Here, data is lost along some axis, and the resulting neuron segmentations are split across the gap. To address this problem, we propose a novel method based on point cloud representations of neurons. We formulate the problem as a classification problem and train CurveNet, a state-of-the-art point cloud classification model, to identify which neurons should be merged. We show that our method not only performs strongly but also scales reasonably to gaps well beyond what other methods have attempted to address. Additionally, our point cloud representations are highly efficient in terms of data, maintaining high performance with an amount of data that would be unfeasible for other methods. We believe that this is an indicator of the viability of using point cloud representations for other proofreading tasks.