CONFIRMS: A Toolkit for Scalable, Black Box Connectome Assessment and Investigation

Abstract

The nanoscale connectomics community has recently generated automated and semi-automated "wiring diagrams" of brain subregions from terabytes and petabytes of dense 3D neuroimagery. This process involves many challenging and imperfect technical steps, including dense 3D image segmentation, anisotropic nonrigid image alignment and coregistration, and pixel classification of each neuron and their individual synaptic connections. As data volumes continue to grow in size, and connectome generation becomes increasingly commonplace, it is important that the scientific community is able to rapidly assess the quality and accuracy of a connectome product to promote dataset analysis and reuse. In this work, we share our scalable toolkit for assessing the quality of a connectome reconstruction via targeted inquiry and large-scale graph analysis, and to provide insights into how such connectome proofreading processes may be improved and optimized in the future. We illustrate the applications and ecosystem on a recent reference dataset.Clinical relevance-Large-scale electron microscopy (EM) data offers a novel opportunity to characterize etiologies and neurological diseases and conditions at an unprecedented scale. EM is useful for low-level analyses such as biopsies; this increased scale offers new possibilities for research into areas such as neural networks if certain bottlenecks and problems are overcome.

NeuVue: A Framework and Workflows for High-Throughput Electron Microscopy Connectomics Proofreading

“…NeuVue-Queue's backend consists of a MongoDB database coupled with an Express application server which is hosted on a scalable fleet of Docker containers. NeuVue-Queue and Neuvue-Client were developed as an extension of Colocard, a tool used for previous MICrONS evaluation efforts as a part of the CONFIRMS framework (Bishop et al 2021).…”

Scalable graph analysis tools for the connectomics community

“…We developed MossDB to serve as an associative datastore for relating graph data, imagery data, and other annotations. MossDB is a simple schemaless wrapper API around industry standards like MongoDB or DynamoDB (or even flat JSON files on disk) that provides (1) immutable references to data, without storing the data itself, (2) authentication, and (3) dataset discovery for secondary analysis.…”

“…Such a graph would exceed the memory- and time-budgets of most institutions. Thus, there are two main limitations to consider: (1) The size of the raw graph data may exceed what can be stored on a single machine; (2) The RAM requirements of the queue may exceed the RAM of the machine.…”

“…While the total size of modern datasets continues to grow, the subset of immediately relevant data is often much smaller than the full dataset (1, 3). This is both a technical challenge as well as logistical challenge for many researchers, who must first query the full dataset to find the relevant subset, and then analyze the subgraph.…”

“…1), scalable and distributed computing approaches are needed to efficiently query and analyze these networks. The connectomics community has begun to develop tools to make this process easier and more efficient, either by subsampling networks to process subsets independently (1), or by adopting existing industry standards for graph storage and processing (2).…”