UNI-EM: An Environment for Deep Neural Network-Based Automated Segmentation of Neuronal Electron Microscopic Images

Abstract

Recently, there has been a rapid expansion in the field of micro-connectomics, which targets the threedimensional (3D) reconstruction of neuronal networks from a stack of two-dimensional (2D) electron microscopic (EM) images. The spatial scale of the 3D reconstruction grows rapidly owing to deep neural networks (DNNs) that enable automated image segmentation. Several research teams have developed their own software pipelines for DNN-based segmentation. However, the complexity of such pipelines makes their use difficult even for computer experts and impossible for non-experts. In this study, we developed a new software program, called UNI-EM, that enables 2D-and 3D-DNNbased segmentation for non-computer experts. UNI-EM is a software collection for DNN-based EM image segmentation, including ground truth generation, training, inference, postprocessing, proofreading, and visualization. UNI-EM comes with a set of 2D DNNs, i.e., U-Net, ResNet, HighwayNet, and DenseNet. We further wrapped flood-filling networks (FFNs) as a representative 3D DNN-based neuron segmentation algorithm. The 2D-and 3D-DNNs are known to show state-of-theart level segmentation performance. We then provided two-example workflows: mitochondria segmentation using a 2D DNN as well as neuron segmentation using FFNs. Following these example workflows, users can benefit from DNN-based segmentation without any knowledge of Python programming or DNN frameworks.

Semantic segmentation of HeLa cells: An objective comparison between one traditional algorithm and four deep-learning architectures

“…Cireşan [59]et al applied deep learning architectures to detect membrane neuronal and mitosis detection in breast cancer [60]. Within EM studies, deep learning has been applied to analyse mitochondria [61,62], synapses [63] and proteins [64].…”

A bird’s-eye view of deep learning in bioimage analysis

“…But as long as human experts continue to be the gold standard in critical vision-based decision-making tasks, it seems there is still much to be gained from renewed interactions between the fields [40] , [248] , [249] , [250] . Bioimage analysis could play a pivotal role here, in a virtuous circle of helping to decipher BNNs at the microscopic level [272] , [273] , [274] and translating discoveries into improved ANNs for such studies [275] , [276] , [277] .…”

A Survey of Visualization and Analysis in High‐Resolution Connectomics

“…Users can access their service through a website, where they can retrain existing models with their training data or run DeepEM3D [ZWJ17] on a set of provided data set. The Uni‐EM tool [UBK * 19] helps to make automatic deep learning segmentation algorithms more accessible through a visual interface, making it usable for non‐programmers. Users can generate ground truth data with a paint tool in Uni‐EM and use it for network training.…”