Insect neural systems are a promising source of inspiration for new navigation algorithms, especially on low size, weight, and power platforms. There have been unprecedented recent neuroscience breakthroughs with Drosophila in behavioral and neural imaging experiments as well as the mapping of detailed connectivity of neural structures. General mechanisms for learning orientation in the central complex (CX) of Drosophila have been investigated previously; however, it is unclear how these underlying mechanisms extend to cases where there is translation through an environment (beyond only rotation), which is critical for navigation in robotic systems. Here, we develop a CX neural connectivity-constrained model that performs sensor fusion, as well as unsupervised learning of visual features for path integration; we demonstrate the viability of this circuit for use in robotic systems in simulated and physical environments. Furthermore, we propose a theoretical understanding of how distributed online unsupervised network weight modification can be leveraged for learning in a trajectory through an environment by minimizing orientation estimation error. Overall, our results may enable a new class of CX-derived low power robotic navigation algorithms and lead to testable predictions to inform future neuroscience experiments.

1.AbstractNeuVue is a software platform created for large-scale proofreading of machine segmentation and neural circuit reconstruction in high-resolution electron microscopy connectomics datasets. The NeuVue platform provides a robust web-based interface for proofreaders to collaboratively view, annotate, and edit segmentation and connectivity data. A backend queuing service organizes proofreader tasks into purpose-driven task types and increases proofreader throughput by limiting proofreader actions to simple, atomic operations. A collection of analytical dashboards, data visualization tools, and Application Program Interface (API) capabilities provide stakeholders real-time access to proofreading progress at an individual proofreader level as well as insights on task generation priorities. NeuVue is agnostic to the underlying data being proofread and improves upon the traditional proofreader experience through quality-of-life features that streamline complex editing operations such as splitting and merging objects in dense nanoscale segmentation.NeuVue heavily leverages cloud resources to enable proofreaders to simultaneously access and edit data on the platform. Production-quality features such as load-balancing, auto-scaling, and pre-deployment testing are all integrated into the platform’s cloud architecture. Additionally, NeuVue is powered by well-supported open-source connectomics tools from the community such as Neuroglancer, PyChunkedGraph, and Connectomics Annotation Versioning Engine (CAVE). The modular design of NeuVue facilitates easy integration and adoption of useful community tools to allow proofreaders to take advantage of the latest improvements in data visualization, processing, and analysis.We demonstrate our framework through proofreading of the mouse visual cortex data generated on the IARPA MICrONS Project. This effort has yielded over 40,000 proofreader edits across the 2 petavoxels of “Minnie” neuroimaging data. 44 unique proofreaders of various skill levels have logged a cumulative 3,740 proofreading hours, and we have been able to validate the improved connectivity of thousands of neurons in the volume. With sustained development on the platform, new integrated error detection and error correction capabilities, and continuous improvements to the proofreader model, we believe that the NeuVue framework can enable high-throughput proofreading for large-scale connectomics datasets of the future.

No abstract

No abstract

As biological imaging datasets continue to grow in size, extracting information from large image volumes presents a computationally intensive challenge. State-of-the-art algorithms are almost entirely dominated by the use of convolutional neural network approaches that may be difficult to run at scale given schedule, cost, and resource limitations. We demonstrate a novel solution for high-resolution electron microscopy brain image volumes that permits the identification of individual neurons and synapses. Instead of conventional approaches whereby voxels are labelled according to the neuron or neuron segment to which they belong, we instead focus on extracting the underlying brain graph represented by synaptic connections between individual neurons while also identifying key features like skeleton similarity and path length. This graph represents a critical step and scaffold for understanding the structure of neuronal circuitry. Our approach recasts the segmentation problem to one of path-finding between keypoints (i.e., connectivity) in an information sharing framework using virtual agents. We create a family of sensors which follow local decision-making rules that perform computationally cheap operations on potential fields to perform tasks such as avoiding cell membranes and finding synapses. These enable a swarm of virtual agents to efficiently and robustly traverse three-dimensional datasets, create a sparse segmentation of pathways, and capture connectivity information. We achieve results that meet or exceed state-of-the-art performance at a substantially lower computational cost. This tool offers a categorically different approach to connectome estimation that can augment how we extract connectivity information at scale. Our method is generalizable and may be extended to biomedical imaging problems such as tracing the bronchial trees in lungs or road networks in natural images.