CAVE: Connectome Annotation Versioning Engine

Abstract

Advances in Electron Microscopy, image segmentation and computational infrastructure have given rise to large-scale and richly annotated connectomic datasets which are increasingly shared across communities. To enable collaboration, users need to be able to concurrently create new annotations and correct errors in the automated segmentation by proofreading. In large datasets, every proofreading edit relabels cell identities of millions of voxels and thousands of annotations like synapses. For analysis, users require immediate and reproducible access to this constantly changing and expanding data landscape. Here, we present the Connectome Annotation Versioning Engine (CAVE), a computational infrastructure for immediate and reproducible connectome analysis in up-to petascale datasets (~1mm3) while proofreading and annotating is ongoing. For segmentation, CAVE provides a distributed proofreading infrastructure for continuous versioning of large reconstructions. Annotations in CAVE are defined by locations such that they can be quickly assigned to the underlying segment which enables fast analysis queries of CAVE's data for arbitrary time points. CAVE supports schematized, extensible annotations, so that researchers can readily design novel annotation types. CAVE is already used for many connectomics datasets, including the largest datasets available to date.

Tools for connectomic reconstruction and analysis of a femaleDrosophilaventral nerve cord

“…We imported synapse and cell body predictions into the Connectome Annotation Versioning Engine (CAVE), so that the associated cell segmentation objects for each annotation would be dynamically updated during proofreading (Dorkenwald et al, 2023b). This system allowed us to query the up-to-date connectivity graph and associated metadata, such as cell-type annotations.…”

Synaptic architecture of leg and wing premotor control networks inDrosophila

“…The computed level 2 cache maintains a representative central point in space, the volume, and the surface area for each level 2 chunk and these statistics are updated when new chunks are created due to proofreading edits. We used the Connectome Annotation Versioning Engine (CAVE) to annotate neurons and keep track of their identities through iterations of proofreading and materializations (Dorkenwald et al, 2023). Somatic segmentations of all motor neurons (downloaded at 68.8 x 68.8 x 45 nm resolution) were cleaned using a heuristic cleaning procedure that removed missing slices of data and incorrectly merged fragments.…”

A petavoxel fragment of human cerebral cortex reconstructed at nanoscale resolution

“…Because of the large number of neuronal structures, it is infeasible for a single laboratory to proofread the entire dataset manually. To facilitate scientific studies that require correction of agglomeration errors by proofreading, we provide a collaborative online proofreading platform for H01 (https://h01-release.storage.googleapis.com/proofreading.html), which is built upon the CAVE (Connectome Annotation and Versioning Engine) infrastructure ( 30 ). This tool is web based and tightly integrated with the Neuroglancer viewer.…”