Search

Deriving the detailed synaptic connections of an entire nervous system is the unrealized goal of the nascent field of connectomics. For the fruit fly Drosophila, in particular, we need to dissect the brain, connectives, and ventral nerve cord as a single continuous unit, fix and stain it, and undertake automated segmentation of neuron membranes. To achieve this, we designed a protocol using progressive lowering of temperature dehydration (PLT), a technique routinely used to preserve cellular structure and antigenicity. We combined PLT with low temperature en bloc staining (LTS) and recover fixed neurons as round profiles with darkly stained synapses, suitable for machine segmentation and automatic synapse detection. Here we report three different PLT-LTS methods designed to meet the requirements for FIB-SEM imaging of the Drosophila brain. These requirements include: good preservation of ultrastructural detail, high level of en bloc staining, artifact-free microdissection, and smooth hot-knife cutting to reduce the brain to dimensions suited to FIB-SEM. In addition to PLT-LTS, we designed a jig to microdissect and pre-fix the fly’s delicate brain and central nervous system. Collectively these methods optimize morphological preservation, allow us to image the brain usually at 8 nm per voxel, and simultaneously speed the formerly slow rate of FIB-SEM imaging.

Due to advances in electron microscopy and deep learning, it is now practical to reconstruct a connectome, a description of neurons and the chemical synapses between them, for significant volumes of neural tissue. Smaller past reconstructions were primarily used by domain experts, could be handled by downloading data, and performance was not a serious problem. But new and much larger reconstructions upend these assumptions. These networks now contain tens of thousands of neurons and tens of millions of connections, with yet larger reconstructions pending, and are of interest to a large community of non-specialists. Allowing other scientists to make use of this data needs more than publication—it requires new tools that are publicly available, easy to use, and efficiently handle large data. We introduce neuPrint to address these data analysis challenges. Neuprint contains two major components—a web interface and programmer APIs. The web interface is designed to allow any scientist worldwide, using only a browser, to quickly ask and answer typical biological queries about a connectome. The neuPrint APIs allow more computer-savvy scientists to make more complex or higher volume queries. NeuPrint also provides features for assessing reconstruction quality. Internally, neuPrint organizes connectome data as a graph stored in a neo4j database. This gives high performance for typical queries, provides access though a public and well documented query language Cypher, and will extend well to future larger connectomics databases. Our experience is also an experiment in open science. We find a significant fraction of the readers of the article proceed to examine the data directly. In our case preprints worked exactly as intended, with data inquiries and PDF downloads starting immediately after pre-print publication, and little affected by formal publication later. From this we deduce that many readers are more interested in our data than in our analysis of our data, suggesting that data-only papers can be well appreciated and that public data release can speed up the propagation of scientific results by many months. We also find that providing, and keeping, the data available for online access imposes substantial additional costs to connectomics research.

Visual pathways from the compound eye of an insect relay to four neuropils, successively the lamina, medulla, lobula, and lobula plate in the underlying optic lobe. Among these neuropils, the medulla, lobula, and lobula plate are interconnected by the complex second optic chiasm, through which the anteroposterior axis undergoes an inversion between the medulla and lobula. Given their complex structure, the projection patterns through the second optic chiasm have so far lacked critical analysis. By densely reconstructing axon trajectories using a volumetric scanning electron microscopy (SEM) technique, we reveal the three-dimensional structure of the second optic chiasm of Drosophila melanogaster, which comprises interleaving bundles and sheets of axons insulated from each other by glial sheaths. These axon bundles invert their horizontal sequence in passing between the medulla and lobula. Axons connecting the medulla and lobula plate are also bundled together with them but do not decussate the sequence of their horizontal positions. They interleave with sheets of projection neuron axons between the lobula and lobula plate, which also lack decussations. We estimate that approximately 19,500 cells per hemisphere, about two thirds of the optic lobe neurons, contribute to the second chiasm, most being Tm cells, with an estimated additional 2,780 T4 and T5 cells each. The chiasm mostly comprises axons and cell body fibers, but also a few synaptic elements. Based on our anatomical findings, we propose that a chiasmal structure between the neuropils is potentially advantageous for processing complex visual information in parallel. The EM reconstruction shows not only the structure of the chiasm in the adult brain, the previously unreported main topic of our study, but also suggest that the projection patterns of the neurons comprising the chiasm may be determined by the proliferation centers from which the neurons develop. Such a complex wiring pattern could, we suggest, only have arisen in several evolutionary steps.

Open-source software development has skyrocketed in part due to community tools like github.com, which allows publication of code as well as the ability to create branches and push accepted modifications back to the original repository. As the number and size of EM-based datasets increases, the connectomics community faces similar issues when we publish snapshot data corresponding to a publication. Ideally, there would be a mechanism where remote collaborators could modify branches of the data and then flexibly reintegrate results via moderated acceptance of changes. The DVID system provides a web-based connectomics API and the first steps toward such a distributed versioning approach to EM-based connectomics datasets. Through its use as the central data resource for Janelia's FlyEM team, we have integrated the concepts of distributed versioning into reconstruction workflows, allowing support for proofreader training and segmentation experiments through branched, versioned data. DVID also supports persistence to a variety of storage systems from high-speed local SSDs to cloud-based object stores, which allows its deployment on laptops as well as large servers. The tailoring of the backend storage to each type of connectomics data leads to efficient storage and fast queries. DVID is freely available as open-source software with an increasing number of supported storage options.

Reconstructing a connectome from an EM dataset often requires a large effort of proofreading automatically generated segmentations. While many tools exist to enable tracing or proofreading, recent advances in EM imaging and segmentation quality suggest new strategies and pose unique challenges for tool design to accelerate proofreading. Namely, we now have access to very large multi-TB EM datasets where (1) many segments are largely correct, (2) segments can be very large (several GigaVoxels), and where (3) several proofreaders and scientists are expected to collaborate simultaneously. In this paper, we introduce NeuTu as a solution to efficiently proofread large, high-quality segmentation in a collaborative setting. NeuTu is a client program of our high-performance, scalable image database called DVID so that it can easily be scaled up. Besides common features of typical proofreading software, NeuTu tames unprecedentedly large data with its distinguishing functions, including: (1) low-latency 3D visualization of large mutable segmentations; (2) interactive splitting of very large false merges with highly optimized semi-automatic segmentation; (3) intuitive user operations for investigating or marking interesting points in 3D visualization; (4) visualizing proofreading history of a segmentation; and (5) real-time collaborative proofreading with lock-based concurrency control. These unique features have allowed us to manage the workflow of proofreading a large dataset smoothly without dividing them into subsets as in other segmentation-based tools. Most importantly, NeuTu has enabled some of the largest connectome reconstructions as well as interesting discoveries in the fly brain.

Automatic image segmentation is critical to scale up electron microscope (EM) connectome reconstruction. To this end, segmentation competitions, such as CREMI and SNEMI, exist to help researchers evaluate segmentation algorithms with the goal of improving them. Because generating ground truth is time-consuming, these competitions often fail to capture the challenges in segmenting larger datasets required in connectomics. More generally, the common metrics for EM image segmentation do not emphasize impact on downstream analysis and are often not very useful for isolating problem areas in the segmentation. For example, they do not capture connectivity information and often over-rate the quality of a segmentation as we demonstrate later. To address these issues, we introduce a novel strategy to enable evaluation of segmentation at large scales both in a supervised setting, where ground truth is available, or an unsupervised setting. To achieve this, we first introduce new metrics more closely aligned with the use of segmentation in downstream analysis and reconstruction. In particular, these include synapse connectivity and completeness metrics that provide both meaningful and intuitive interpretations of segmentation quality as it relates to the preservation of neuron connectivity. Also, we propose measures of segmentation correctness and completeness with respect to the percentage of “orphan” fragments and the concentrations of self-loops formed by segmentation failures, which are helpful in analysis and can be computed without ground truth. The introduction of new metrics intended to be used for practical applications involving large datasets necessitates a scalable software ecosystem, which is a critical contribution of this paper. To this end, we introduce a scalable, flexible software framework that enables integration of several different metrics and provides mechanisms to evaluate and debug differences between segmentations. We also introduce visualization software to help users to consume the various metrics collected. We evaluate our framework on two relatively large public groundtruth datasets providing novel insights on example segmentations.

Extracting a connectome from an electron microscopy (EM) data set requires identification of neurons and determination of connections (synapses) between neurons. As manual extraction of this information is very time-consuming, there has been extensive research efforts to automatically segment the neurons to help guide and eventually replace manual tracing. Until recently, there has been comparatively little research on automatic detection of the actual synapses between neurons. This discrepancy can, in part, be attributed to several factors: obtaining neuronal shapes is a prerequisite for the first step in extracting a connectome, manual tracing is much more time-consuming than annotating synapses, and neuronal contact area can be used as a proxy for synapses in determining connections. However, recent research has demonstrated that contact area alone is not a sufficient predictor of a synaptic connection. Moreover, as segmentation improved, we observed that synapse annotation consumes a more significant fraction of overall reconstruction time (upwards of 50% of total effort). This ratio will only get worse as segmentation improves, gating the overall possible speed-up. Therefore, we address this problem by developing algorithms that automatically detect presynaptic neurons and their postsynaptic partners. In particular, presynaptic structures are detected using a U-Net convolutional neural network (CNN), and postsynaptic partners are detected using a multilayer perceptron (MLP) with features conditioned on the local segmentation. This work is novel because it requires minimal amount of training, leverages advances in image segmentation directly, and provides a complete solution for polyadic synapse detection. We further introduce novel metrics to evaluate our algorithm on connectomes of meaningful size. When applied to the output of our method on EM data from Drosphila, these metrics demonstrate that a completely automatic prediction can be used to effectively characterize most of the connectivity correctly.

Synaptic circuits for identified behaviors in the Drosophila brain have typically been considered from either a developmental or functional perspective without reference to how the circuits might have been inherited from ancestral forms. For example, two candidate pathways for ON- and OFF-edge motion detection in the visual system act via circuits that use respectively either T4 or T5, two cell types of the fourth neuropil, or lobula plate, that exhibit narrow-field direction-selective responses and provide input to wide-field tangential neurons. T4 or T5 both have four subtypes that terminate one each in the four strata of the lobula plate. Representatives are reported in a wide range of Diptera, and both cell types exhibit various similarities in: 1) the morphology of their dendritic arbors; 2) their four morphological and functional subtypes; 3) their cholinergic profile in Drosophila; 4) their input from the pathways of L3 cells in the first neuropil, or lamina, and by one of a pair of lamina cells, L1 (to the T4 pathway) and L2 (to the T5 pathway); and 5) their innervation by a single, wide-field contralateral tangential neuron from the central brain. Progenitors of both also express the gene atonal early in their proliferation from the inner anlage of the developing optic lobe, being alone among many other cell type progeny to do so. Yet T4 receives input in the second neuropil, or medulla, and T5 in the third neuropil or lobula. Here we suggest that these two cell types were originally one, that their ancestral cell population duplicated and split to innervate separate medulla and lobula neuropils, and that a fiber crossing – the internal chiasma – arose between the two neuropils. The split most plausibly occurred, we suggest, with the formation of the lobula as a new neuropil that formed when it separated from its ancestral neuropil to leave the medulla, suggesting additionally that medulla input neurons to T4 and T5 may also have had a common origin.

The hemibrain connectome provides large-scale connectivity and morphology information for the majority of the central brain of Drosophila melanogaster. Using this data set, we provide a complete description of the Drosophila olfactory system, covering all first, second and lateral horn-associated third-order neurons. We develop a generally applicable strategy to extract information flow and layered organisation from connectome graphs, mapping olfactory input to descending interneurons. This identifies a range of motifs including highly lateralised circuits in the antennal lobe and patterns of convergence downstream of the mushroom body and lateral horn. Leveraging a second data set we provide a first quantitative assessment of inter- versus intra-individual stereotypy. Comparing neurons across two brains (three hemispheres) reveals striking similarity in neuronal morphology across brains. Connectivity correlates with morphology and neurons of the same morphological type show similar connection variability within the same brain as across two brains.

The neural circuits responsible for animal behavior remain largely unknown. We summarize new methods and present the circuitry of a large fraction of the brain of the fruit fly Drosophila melanogaster. Improved methods include new procedures to prepare, image, align, segment, find synapses in, and proofread such large data sets. We define cell types, refine computational compartments, and provide an exhaustive atlas of cell examples and types, many of them novel. We provide detailed circuits consisting of neurons and their chemical synapses for most of the central brain. We make the data public and simplify access, reducing the effort needed to answer circuit questions, and provide procedures linking the neurons defined by our analysis with genetic reagents. Biologically, we examine distributions of connection strengths, neural motifs on different scales, electrical consequences of compartmentalization, and evidence that maximizing packing density is an important criterion in the evolution of the fly’s brain.

Understanding the circuit mechanisms behind motion detection is a long-standing question in visual neuroscience. In Drosophila melanogaster, recently discovered synapse-level connectomes in the optic lobe, particularly in ON-pathway (T4) receptive-field circuits, in concert with physiological studies, suggest a motion model that is increasingly intricate when compared with the ubiquitous Hassenstein-Reichardt model. By contrast, our knowledge of OFF-pathway (T5) has been incomplete. Here, we present a conclusive and comprehensive connectome that, for the first time, integrates detailed connectivity information for inputs to both the T4 and T5 pathways in a single EM dataset covering the entire optic lobe. With novel reconstruction methods using automated synapse prediction suited to such a large connectome, we successfully corroborate previous findings in the T4 pathway and comprehensively identify inputs and receptive fields for T5. Although the two pathways are probably evolutionarily linked and exhibit many similarities, we uncover interesting differences and interactions that may underlie their distinct functional properties.

Using FIB-SEM we report the entire synaptic connectome of glomerulus VA1v of the right antennal lobe in Drosophila melanogaster. Within the glomerulus we densely reconstructed all neurons, including hitherto elusive local interneurons. The fruitless-positive, sexually dimorphic VA1v included >11,140 presynaptic sites with ~38,050 postsynaptic dendrites. These connected input olfactory receptor neurons (ORNs, 51 ipsilateral, 56 contralateral), output projection neurons (18 PNs), and local interneurons (56 of >150 previously reported LNs). ORNs are predominantly presynaptic and PNs predominantly postsynaptic; newly reported LN circuits are largely an equal mixture and confer extensive synaptic reciprocity, except the newly reported LN2V with input from ORNs and outputs mostly to monoglomerular PNs, however. PNs were more numerous than previously reported from genetic screens, suggesting that the latter failed to reach saturation. We report a matrix of 192 bodies each having >50 connections; these form 88% of the glomerulus’ pre/postsynaptic sites.

The aim in high-resolution connectomics is to reconstruct complete neuronal connectivity in a tissue. Currently, the only technology capable of resolving the smallest neuronal processes is electron microscopy (EM). Thus, a common approach is to perform automatic segmentation of EM images, followed by manual proofreading by experts to fix errors. We developed an algorithm and software library to not only improve the accuracy of the initial automatic segmentation, but also point out the image coordinates where it is likely to have made errors. Our software, called gala (graph-based active learning of agglomeration), improves the state of the art in agglomerative image segmentation. It is implemented in Python and makes extensive use of the scientific Python stack (numpy, scipy, networkx, scikit-learn, scikit-image, and others). We present here the software architecture of the gala library, and discuss several designs that we consider would be generally useful for other segmentation packages. We also discuss the limitations of the gala library and how we intend to address them.

The ability to automatically segment an image into distinct regions is a critical aspect in many visual processing applications. Because inaccuracies often exist in automatic segmentation, manual segmentation is necessary in some application domains to correct mistakes, such as required in the reconstruction of neuronal processes from microscopic images. The goal of the automated segmentation tool is traditionally to produce the highest-quality segmentation, where quality is measured by the similarity to actual ground truth, so as to minimize the volume of manual correction necessary. Manual correction is generally orders-of-magnitude more time consuming than automated segmentation, often making handling large images intractable. Therefore, we propose a more relevant goal: minimizing the turn-around time of automated/manual segmentation while attaining a level of similarity with ground truth. It is not always necessary to inspect every aspect of an image to generate a useful segmentation. As such, we propose a strategy to guide manual segmentation to the most uncertain parts of segmentation. Our contributions include 1) a probabilistic measure that evaluates segmentation without ground truth and 2) a methodology that leverages these probabilistic measures to significantly reduce manual correction while maintaining segmentation quality.