Search

Penetrating flexible electrode arrays can simultaneously record thousands of individual neurons in the brains of live animals. However, it has been challenging to spatially map and longitudinally monitor the dynamics of large three-dimensional neural networks. Here we show that optimized ultraflexible electrode arrays distributed across multiple cortical regions in head-fixed mice and in freely moving rats allow for months-long stable electrophysiological recording of several thousand neurons at densities of about 1,000 neural units per cubic millimetre. The chronic recordings enhanced decoding accuracy during optogenetic stimulation and enabled the detection of strongly coupled neuron pairs at the million-pair and millisecond scales, and thus the inference of patterns of directional information flow. Longitudinal and volumetric measurements of neural couplings may facilitate the study of large-scale neural circuits.

Understanding the complexities of behavior is necessary to interpret neurophysiological data and establish animal models of neuropsychiatric disease. This understanding requires knowledge of the underlying information-processing structure-something often hidden from direct observation. Commonly, one assumes that behavior is solely governed by the experimenter-controlled rules that determine tasks. For example, differences in tasks that require memory of past actions are often interpreted as exclusively resulting from differences in memory. However, such assumptions are seldom tested. Here, we provide a comprehensive examination of multiple processes that contribute to behavior in a prevalent experimental paradigm. Using a combination of behavioral automation, hypothesis-driven trial design, and reinforcement learning modeling, we show that rats learn a spatial alternation task consistent with their drawing upon spatial preferences in addition to memory. Our approach also distinguishes learning based on established preferences from generalization of task structure, providing further insights into learning dynamics.

Executing memory-guided behavior requires both the storage of information about experience and the later recall of that information to inform choices. Awake hippocampal replay, when hippocampal neural ensembles briefly reactivate a representation related to prior experience, has been proposed to critically contribute to these memory-related processes. However, it remains unclear whether awake replay contributes to memory function by promoting the storage of past experiences, by facilitating planning based on an evaluation of those experiences, or both. We designed a dynamic spatial task which promotes replay before a memory-based choice and assessed how the content of replay related to past and future behavior. We found that replay content was decoupled from subsequent choice and instead was enriched for representations of previously rewarded locations and places that had not been recently visited, indicating a role in memory storage rather than in directly guiding subsequent behavior.

The importance of training robust neural network grows as 3D data is increasingly utilized in deep learning for vision tasks in robotics, drone control, and autonomous driving. One commonly used 3D data type is 3D point clouds, which describe shape information. We examine the problem of creating robust models from the perspective of the attacker, which is necessary in understanding how neural networks can be exploited. We explore two categories of attacks: distributional attacks that involve imperceptible perturbations to the distribution of points, and shape attacks that involve deforming the shape represented by a point cloud. We explore three possible shape attacks for attacking 3D point cloud classification and show that some of them are able to be effective even against preprocessing steps, like the previously proposed point-removal defenses. (Source code available at https://github.com/Daniel-Liu-c0deb0t/Adversarial-point-perturbations-on-3D-objects).

Summary Executing memory-guided behavior requires storage of information about experience and later recall of that information to inform choices. Awake hippocampal replay, when hippocampal neural ensembles briefly reactivate a representation related to prior experience, has been proposed to critically contribute to these memory-related processes. However, it remains unclear whether awake replay contributes to memory function by promoting the storage of past experiences, facilitating planning based on evaluation of those experiences, or both. We designed a dynamic spatial task that promotes replay before a memory-based choice and assessed how the content of replay related to past and future behavior. We found that replay content was decoupled from subsequent choice and instead was enriched for representations of previously rewarded locations and places that had not been visited recently, indicating a role in memory storage rather than in directly guiding subsequent behavior.

3D object classification and segmentation using deep neural networks has been extremely successful. As the problem of identifying 3D objects has many safety-critical applications, the neural networks have to be robust against adversarial changes to the input data set. There is a growing body of research on generating human-imperceptible adversarial attacks and defenses against them in the 2D image classification domain. However, 3D objects have various differences with 2D images, and this specific domain has not been rigorously studied so far. We present a preliminary evaluation of adversarial attacks on deep 3D point cloud classifiers, namely PointNet and PointNet++, by evaluating both white-box and black-box adversarial attacks that were proposed for 2D images and extending those attacks to reduce the perceptibility of the perturbations in 3D space. We also show the high effectiveness of simple defenses against those attacks by proposing new defenses that exploit the unique structure of 3D point clouds. Finally, we attempt to explain the effectiveness of the defenses through the intrinsic structures of both the point clouds and the neural network architectures. Overall, we find that networks that process 3D point cloud data are weak to adversarial attacks, but they are also more easily defensible compared to 2D image classifiers. Our investigation will provide the groundwork for future studies on improving the robustness of deep neural networks that handle 3D data., Abridged version accepted at the 2019 IEEE International Conference on Image Processing (ICIP). Source code: https://github.com/Daniel-Liu-c0deb0t/3D-Neural-Network-Adversarial-Attacks

Cognitive faculties such as imagination, planning, and decision-making entail the ability to represent hypothetical experience. Crucially, animal behavior in natural settings implies that the brain can represent hypothetical future experience not only quickly but also constantly over time, as external events continually unfold. To determine how this is possible, we recorded neural activity in the hippocampus of rats navigating a maze with multiple spatial paths. We found neural activity encoding two possible future scenarios (two upcoming maze paths) in constant alternation at 8Hz: one scenario per ∼125-ms cycle. Further, we found that the underlying dynamics of cycling (both inter- and intra-cycle dynamics) generalized across qualitatively different representational correlates (location and direction). Notably, cycling occurred across moving behaviors, including during running. These findings identify a general dynamic process capable of quickly and continually representing hypothetical experience, including that of multiple possible futures.

Cognitive faculties such as imagination, planning, and decision-making require the ability to represent alternative scenarios. In animals, split-second decision-making implies that the brain can represent alternatives at a commensurate speed. Yet despite this insight, it has remained unknown whether there exists neural activity that can consistently represent alternatives in

Sharp-wave ripple (SWR) events in the hippocampus replay millisecond-timescale patterns of place cell activity related to the past experience of an animal. Interrupting SWR events leads to learning and memory impairments, but how the specific patterns of place cell spiking seen during SWRs contribute to learning and memory remains unclear. A deeper understanding of this issue will require the ability to manipulate SWR events based on their content. Accurate real-time decoding of SWR replay events requires new algorithms that are able to estimate replay content and the associated uncertainty, along with software and hardware that can execute these algorithms for biological interventions on a millisecond timescale. Here we develop an efficient estimation algorithm to categorize the content of replay from multiunit spiking activity. Specifically, we apply real-time decoding methods to each SWR event and then compute the posterior probability of the replay feature. We illustrate this approach by classifying SWR events from data recorded in the hippocampus of a rat performing a spatial memory task into four categories: whether they represent outbound or inbound trajectories and whether the activity is replayed forward or backward in time. We show that our algorithm can classify the majority of SWR events in a recording epoch within 20 ms of the replay onset with high certainty, which makes the algorithm suitable for a real-time implementation with short latencies to incorporate into content-based feedback experiments.

Real-time, closed-loop experiments can uncover causal relationships between specific neural activity and behavior. An important advance in realizing this is the marked point process filtering framework which utilizes the “mark” or the waveform features of unsorted spikes, to construct a relationship between these features and behavior, which we call the encoding model. This relationship is not fixed, because learning changes coding properties of individual neurons, and electrodes can physically move during the experiment, changing waveform characteristics. We introduce a sequential, Bayesian encoding model which allows incorporation of new information on the fly to adapt the model in real time. A possible application of this framework is to the decoding of the contents of hippocampal ripples in rats exploring a maze. During physical exploration, we observe the marks and positions at which they occur, to update the encoding model, which is employed to decode contents of ripples when rats stop moving, and switch back to updating the model once the rat starts moving again.

Memories link information about specific experiences to more general knowledge that is abstracted from and contextualizes those experiences. Hippocampal-cortical activity patterns representing features of past experience are reinstated during awake memory reactivation events, but whether representations of both specific and general features of experience are simultaneously reinstated remains unknown. We examined hippocampal and prefrontal cortical firing patterns during memory reactivation in rats performing a well-learned foraging task with multiple spatial paths. We found that specific hippocampal place representations are preferentially reactivated with the subset of prefrontal cortical task representations that generalize across different paths. Our results suggest that hippocampal-cortical networks maintain links between stored representations for specific and general features of experience, which could support abstraction and task guidance in mammals., Memory representations in cortex and hippocampus are reactivated during sharp-wave ripple (SWR) events. Here, the authors show that, in a familiar environment, this activity preferentially links spatially selective hippocampal cells and task general PFC representations, pointing to a potential neural mechanism for generalization of individual experiences.

The brain is a massively interconnected neuronal network, organized into specialized circuits consisting of large ensembles of neurons distributed across anatomically connected regions. While circuit computations depend upon millisecond timescale interactions, the structure of the underlying networks are remodeled on timescales ranging from seconds to months. Current approaches lack the combination of resolution, spatial coverage, longevity, and stability to measure the detailed dynamics of these networks. Here we describe a large-scale, multisite recording platform that integrates polymer electrodes with a modular stacking headstage design supporting up to 1024 channels of recording in freely-behaving rats. We show that the integrated system can yield months-long recordings from hundreds of well-isolated units across multiple regions. Moreover, the recordings are stable enough to track a substantial fraction of single units for over a week. This platform enables large-scale electrophysiological interrogation of the function and evolution of distributed circuits throughout an animal9s adult life.

© 2015 Massachusetts Institute of Technology. Point process filters have been applied successfully to decode neural signals and track neural dynamics. Traditionally thesemethods assume that multiunit spiking activity has already been correctly spike-sorted. As a result, these methods are not appropriate for situations where sorting cannot be performed with high precision, such as real-time decoding for brain-computer interfaces. Because the unsupervised spike-sorting problem remains unsolved, we took an alternative approach that takes advantage of recent insights into clusterless decoding. Here we present a new point process decoding algorithm that does not require multiunit signals to be sorted into individual units. We use the theory of marked point processes to construct a function that characterizes the relationship between a covariate of interest (in this case, the location of a rat on a track) and features of the spike waveforms. In our example, we use tetrode recordings, and the marks represent a four-dimensional vector of the maximum amplitudes of the spike waveform on each of the four electrodes. In general, the marks may represent any features of the spike waveform. We then use Bayes's rule to estimate spatial location from hippocampal neural activity. We validate our approach with a simulation study and experimental data recorded in the hippocampus of a rat moving through a linear environment. Our decoding algorithm accurately reconstructs the rat's position from unsorted multiunit spiking activity. We then compare the quality of our decoding algorithm to that of a traditional spike-sorting and decoding algorithm. Our analyses show that the proposed decoding algorithm performs equivalent to or better than algorithms based on sorted single-unit activity. These results provide a path toward accurate real-time decoding of spiking patterns that could be used to carry out content-specific manipulations of population activity in hippocampus or elsewhere in the brain.

While ongoing experience proceeds continuously, memories of past experience are often recalled as episodes with defined beginnings and ends. The neural mechanisms that lead to the formation of discrete episodes from the stream of neural activity patterns representing ongoing experience are unknown. To investigate these mechanisms, we recorded neural activity in the rat hippocampus and prefrontal cortex, structures critical for memory processes. We show that during spatial navigation, hippocampal CA1 place cells maintain a continuous spatial representation across different states of motion (movement and immobility). In contrast, during sharp-wave ripples (SWRs), when representations of experience are transiently reactivated from memory, movement- and immobility-associated activity patterns are most often reactivated separately. Concurrently, distinct hippocampal reactivations of movement- or immobility-associated representations are accompanied by distinct modulation patterns in prefrontal cortex. These findings demonstrate a continuous representation of ongoing experience can be separated into independently reactivated memory representations.

Author(s): Joo, Hannah R; Fan, Jiang Lan; Chen, Supin; Pebbles, Jeanine A; Liang, Hexin; Chung, Jason E; Yorita, Allison M; Tooker, Angela C; Tolosa, Vanessa M; Geaghan-Breiner, Charlotte; Roumis, Demetris K; Liu, Daniel F; Haque, Razi; Frank, Loren M | Abstract: ObjectiveElectrode arrays for chronic implantation in the brain are a critical technology in both neuroscience and medicine. Recently, flexible, thin-film polymer electrode arrays have shown promise in facilitating stable, single-unit recordings spanning months in rats. While array flexibility enhances integration with neural tissue, it also requires removal of the dura mater, the tough membrane surrounding the brain, and temporary bracing to penetrate the brain parenchyma. Durotomy increases brain swelling, vascular damage, and surgical time. Insertion using a bracing shuttle results in additional vascular damage and brain compression, which increase with device diameter; while a higher-diameter shuttle will have a higher critical load and more likely penetrate dura, it will damage more brain parenchyma and vasculature. One way to penetrate the intact dura and limit tissue compression without increasing shuttle diameter is to reduce the force required for insertion by sharpening the shuttle tip.ApproachWe describe a novel design and fabrication process to create silicon insertion shuttles that are sharp in three dimensions and can penetrate rat dura, for faster, easier, and less damaging implantation of polymer arrays. Sharpened profiles are obtained by reflowing patterned photoresist, then transferring its sloped profile to silicon with dry etches.Main resultsWe demonstrate that sharpened shuttles can reliably implant polymer probes through dura to yield high quality single unit and local field potential recordings for at least 95 days. On insertion directly through dura, tissue compression is minimal.SignificanceThis is the first demonstration of a rat dural-penetrating array for chronic recording. This device obviates the need for a durotomy, reducing surgical time and risk of damage to the blood-brain barrier. This is an improvement to state-of-the-art flexible polymer electrode arrays that facilitates their implantation, particularly in multi-site recording experiments. This sharpening process can also be integrated into silicon electrode array fabrication.

Next-generation sequencing technologies create large, multiplexed DNA sequences that require preprocessing before any further analysis. Part of this preprocessing includes demultiplexing and trimming sequences. Although there are many existing tools that can handle these preprocessing steps, they cannot be easily extended to new sequence schematics when new pipelines are developed. We present Fuzzysplit, a tool that relies on a simple declarative language to describe the schematics of sequences, which makes it incredibly adaptable to different use cases. In this paper, we explain the matching algorithms behind Fuzzysplit and we provide a preliminary comparison of its performance with other well-established tools. Overall, we find that its matching accuracy is comparable to previous tools.

The human cognitive process of situation awareness is limited to the amount of data and the level of complexity between the data elements. Situation assessment, encompassing automated threat and impact assessment, shall assist human analysts by estimating the critical activities and objects in an emerging situation. The existing work on situation assessment, while serving its individual purposes, is not driven explicitly by the need to enhance situation awareness. This article provides a summary of various related work, ranging from visualization to algorithmic threat projection, and describes a human-centered framework that associates situation assessment processes and models with requirements needed to enhance situation awareness.

The brain is a massively interconnected network of specialized circuits. Even primary sensory areas, once thought to support relatively simple, feed-forward processing, are now known to be parts of complex feedback circuits. All brain functions depend on millisecond timescale interactions across these brain networks. Current approaches cannot measure or manipulate such large-scale interactions. Here we demonstrate that polymer-based, penetrating, micro-electrode arrays can provide high quality neural recordings from awake, behaving animals over periods of months. Our results indicate that polymer electrodes are a viable substrate for the development of systems that can record from thousands of channels across months to years. This is our first step towards developing a 1000+ electrode system capable of providing high-quality, long-term neural recordings.

—With expanding network infrastructures, increasingvulnerabilities and uncertain malicious activities, cyber securityresearch has begun to provide situation assessment beyondIntrusion Detection Systems (IDSs). A key goal of cyber situationassessment is to efficiently and effectively project the likely futuretargets of ongoing multistage attacks. This work presents two en-semble techniques that combine real-time projection algorithmsmodeling the behavior, capability, and opportunity of maliciousactivities in a network. Sugeno fuzzy inference system and Trans-ferable Belief Model are used to combine supporting evidence andresolve conflicts between the algorithm outputs. The two ensembletechniques are analyzed and compared using simulated attackdatasets generated for varying network environments and attackparameters. The results are discussed to reveal the benefits andlimitations of individual algorithms and ensemble techniques. I. I NTRODUCTION Like other problem domains, computer network securityexhibits noisy observations due to not only malicious, butalso trusted and false positive activities. Due to the sheerquantity of observables from cyber security sensors, effectivecomputer security tools must be able to reduce the searchspace of observables by identifying the most malicious andimportant activities. Built upon various Intrusion DetectionSystems (IDSs), alert correlation and attack projection havebeen gaining interest from the research community. Alertcorrelation [1]–[4] seeks to intelligently associate observables.Attack projection [5]–[7] analyzes the aggregated alerts, re-ferred to as multistage attacks, and projects each into thefuture to estimate potentially threatened targets in a network.This paper examines attack projections algorithms that assessdifferent characteristics of multistage attacks and discusseshow ensemble approaches can benefit the projection process.Alert correlation and attack projection share the need tocharacterize or model the progression of cyber attacks. Areasonably administered computer network will require sophis-ticated hackers to perform multiple attack actions before reach-ing critical data or services. The complexity, the uncertainty,and the distributed nature of network and system configura-tions make the modeling of attack progression challenging.Past work on vulnerability trees [8] can be used to modelattacks where an attack may begin at a leaf node and progressto a single root goal. Such an approach may be impracticalbecause there can be a number of large trees that need to beimplemented to capture not only all possible goals, but alsodifferent paths through the network. In fact, attacks towarda single goal may not progress in a tree-like manner. Morecompact approaches have utilized directed acyclic graphs,called attack graphs, and applied Bayesian analysis [6], [9]–[11]. While theoretically sound, generating a comprehensiveset of attack graphs for a given network may be too challengingof a task in the real world. Lippmann [12] reviewed 16 paperson attack graph generation, and found that none has analyzedmore than 20 machines and none has considered a reasonablenumber of vulnerabilities and the complexity of firewall rules.Recognizing the challenge of generating attack graphs,Holsopple

Extensive discussions have taken place in recent year regarding impact assessment - what is it and how can we do it? It is especially intriguing in this modern era where non-traditional warfare has caused either information overload or limited understanding of adversary doctrines. This work provides a methodical discussion of key elements for the broad definition of impact assessment (IA). The discussion will start with a process flow involving components related to IA. Two key functional components, impact estimation and threat projection, are compared and illustrated in detail. These details include a discussion of when to model red and blue knowledge. Algorithmic approaches will be discussed, augmented with lessons learned from our IA development for cyber situation awareness. This paper aims at providing the community with a systematic understanding of IA and its open issues with specific examples.

Traditionally, experiments designed to study the role of specific spike patterns in learning and memory tasks take one of two forms, 1) observational studies that characterize statistical properties of neural activity during such tasks or 2) interventional studies that broadly alter neural activities over an entire neural population or brain region. This work is part of a larger project to allow investigators to manipulate neural populations in a content-specific way, altering spiking activity related to certain learning and memory patterns while leaving activity related to other patterns intact. One fundamental challenge of this work is to decode the information content of specific spike sequences in real-time. Previously, we have used point process theory to develop efficient decoding algorithms based on spike train observations. However these algorithms assume the spike trains have been accurately sorted ahead of time, which is not possible for real-time decoding. Here we present a new point process decoding algorithm that does not require multiunit signals to be sorted. We use the theory of marked point processes to characterize the relationship between the coding properties of multiunit activity and features of the spike waveforms [1-3]. Using Bayes’ rule, we compute the posterior distribution of a signal to decode given multiunit activity from a neural population. We first characterize the spiking activity of a neural population using the conditional intensity function for marked point processes. We then construct point process filters to iteratively calculate the full posterior density of a signal. We illustrate our approach with a simulation study as well as with experimental data recorded in the hippocampus of a rat performing a spatial memory task. Our decoding framework is used to reconstruct the animal’s position from unsorted multiunit spiking activity. We then compare the quality of fit of our decoding framework to that of a traditional spike-sorting and decoding framework. Our analyses show that the proposed decoding algorithm performs as well as or better than algorithms based on sorted single-unit activity. These results provide a mechanism for content-specific manipulations of population activity in hippocampus.