Techniques for extracting single-trial activity patterns from large-scale neural recordings

SHRUNKNeural dynamics encoding risky choices during deliberation reveal separate choice subspaces

Human decision-making involves the coordinated activity of multiple brain areas, acting in concert, to enable humans to make choices. Most decisions are carried out under conditions of uncertainty, where the desired outcome may not be achieved if the wrong decision is made. In these cases, humans deliberate before making a choice. The neural dynamics underlying deliberation are unknown and intracranial recordings in clinical settings present a unique opportunity to record high temporal resolution electrophysiological data from many (hundreds) brain locations during behavior. Combined with dynamic systems modeling, these allow identification of latent brain states that describe the neural dynamics during decision-making, providing insight into these neural dynamics and computations. Results show that the neural dynamics underlying risky decisions, but not decisions without risk, converge to separate subspaces depending on the subject's preferred choice and that the degree of overlap between these subspaces declines as choice approaches, suggesting a network level representation of evidence accumulation. These results bridge the gap between regression analyses and data driven models of latent states and suggest that during risky decisions, deliberation and evidence accumulation toward a final decision are represented by the same neural dynamics, providing novel insights into the neural computations underlying human choice.

Chaotic recurrent neural networks for brain modelling: A review

Even in the absence of external stimuli, the brain is spontaneously active. Indeed, most cortical activity is internally generated by recurrence. Both theoretical and experimental studies suggest that chaotic dynamics characterize this spontaneous activity. While the precise function of brain chaotic activity is still puzzling, we know that chaos confers many advantages. From a computational perspective, chaos enhances the complexity of network dynamics. From a behavioural point of view, chaotic activity could generate the variability required for exploration. Furthermore, information storage and transfer are maximized at the critical border between order and chaos. Despite these benefits, many computational brain models avoid incorporating spontaneous chaotic activity due to the challenges it poses for learning algorithms. In recent years, however, multiple approaches have been proposed to overcome this limitation. As a result, many different algorithms have been developed, initially within the reservoir computing paradigm. Over time, the field has evolved to increase the biological plausibility and performance of the algorithms, sometimes going beyond the reservoir computing framework. In this review article, we examine the computational benefits of chaos and the unique properties of chaotic recurrent neural networks, with a particular focus on those typically utilized in reservoir computing. We also provide a detailed analysis of the algorithms designed to train chaotic RNNs, tracing their historical evolution and highlighting key milestones in their development. Finally, we explore the applications and limitations of chaotic RNNs for brain modelling, consider their potential broader impacts beyond neuroscience, and outline promising directions for future research.

Humans actively reconfigure neural task states

The ability to switch between tasks is a core component of adaptive cognition, yet a mechanistic understanding of this capacity has remained elusive. In particular, there are longstanding debates over the extent to which task-switching is primarily influenced by interference from previous tasks or active preparation for upcoming tasks. We advance this debate by modeling the evolution of task representations in human electroencephalographic recordings as linear dynamical systems over a high-dimensional latent space. Using control theoretic analyses of dynamical systems fit to both brains and task-optimized neural networks, we reveal neural signatures of active preparation that reconcile the role of interference and control in task switching. These findings inform a core debate in cognitive control and offer a promising new paradigm for human neuroimaging analysis.

A Stochastic Dynamic Operator Framework That Improves the Precision of Analysis and Prediction Relative to the Classical Spike-Triggered Average Method, Extending the Toolkit

Here we test the stochastic dynamic operator (SDO) as a new framework for describing physiological signal dynamics relative to spiking or stimulus events. The SDO is a natural extension of existing spike-triggered average (STA) or stimulus-triggered average techniques currently used in neural analysis. It extends the classic STA to cover state-dependent and probabilistic responses where STA may fail. In simulated data, SDO methods were more sensitive and specific than the STA for identifying state-dependent relationships. We have tested SDO analysis for interactions between electrophysiological recordings of spinal interneurons, single motor units, and aggregate muscle electromyograms (EMG) of major muscles in the spinal frog hindlimb. When predicting target signal behavior relative to spiking events, the SDO framework outperformed or matched classical spike-triggered averaging methods. SDO analysis permits more complicated spike-signal relationships to be captured, analyzed, and interpreted visually and intuitively. SDO methods can be applied at different scales of interest where spike-triggered averaging methods are currently employed, and beyond, from single neurons to gross motor behaviors. SDOs may be readily generated and analyzed using the provided SDO Analysis Toolkit We anticipate this method will be broadly useful for describing dynamical signal behavior and uncovering state-dependent relationships of stochastic signals relative to discrete event times.

Mixed Representations of Sound and Action in the Auditory Midbrain

Linking sensory input and its consequences is a fundamental brain operation. During behavior, the neural activity of neocortical and limbic systems often reflects dynamic combinations of sensory and task-dependent variables, and these "mixed representations" are suggested to be important for perception, learning, and plasticity. However, the extent to which such integrative computations might occur outside of the forebrain is less clear. Here, we conduct cellular-resolution two-photon Ca2+ imaging in the superficial "shell" layers of the inferior colliculus (IC), as head-fixed mice of either sex perform a reward-based psychometric auditory task. We find that the activity of individual shell IC neurons jointly reflects auditory cues, mice's actions, and behavioral trial outcomes, such that trajectories of neural population activity diverge depending on mice's behavioral choice. Consequently, simple classifier models trained on shell IC neuron activity can predict trial-by-trial outcomes, even when training data are restricted to neural activity occurring prior to mice's instrumental actions. Thus, in behaving mice, auditory midbrain neurons transmit a population code that reflects a joint representation of sound, actions, and task-dependent variables.

Computational role of structure in neural activity and connectivity

One major challenge of neuroscience is identifying structure in seemingly disorganized neural activity. Different types of structure have different computational implications that can help neuroscientists understand the functional role of a particular brain area. Here, we outline a unified approach to characterize structure by inspecting the representational geometry and the modularity properties of the recorded activity and show that a similar approach can also reveal structure in connectivity. We start by setting up a general framework for determining geometry and modularity in activity and connectivity and relating these properties with computations performed by the network. We then use this framework to review the types of structure found in recent studies of model networks performing three classes of computations.

Preparatory activity and the expansive null-space

The study of the cortical control of movement experienced a conceptual shift over recent decades, as the basic currency of understanding shifted from single-neuron tuning towards population-level factors and their dynamics. This transition was informed by a maturing understanding of recurrent networks, where mechanism is often characterized in terms of population-level factors. By estimating factors from data, experimenters could test network-inspired hypotheses. Central to such hypotheses are 'output-null' factors that do not directly drive motor outputs yet are essential to the overall computation. In this Review, we highlight how the hypothesis of output-null factors was motivated by the venerable observation that motor-cortex neurons are active during movement preparation, well before movement begins. We discuss how output-null factors then became similarly central to understanding neural activity during movement. We discuss how this conceptual framework provided key analysis tools, making it possible for experimenters to address long-standing questions regarding motor control. We highlight an intriguing trend: as experimental and theoretical discoveries accumulate, the range of computational roles hypothesized to be subserved by output-null factors continues to expand.

Mixed representations of sound and action in the auditory midbrain

Linking sensory input and its consequences is a fundamental brain operation. Accordingly, neural activity of neo-cortical and limbic systems often reflects dynamic combinations of sensory and behaviorally relevant variables, and these "mixed representations" are suggested to be important for perception, learning, and plasticity. However, the extent to which such integrative computations might occur in brain regions upstream of the forebrain is less clear. Here, we conduct cellular-resolution 2-photon Ca2+ imaging in the superficial "shell" layers of the inferior colliculus (IC), as head-fixed mice of either sex perform a reward-based psychometric auditory task. We find that the activity of individual shell IC neurons jointly reflects auditory cues and mice's actions, such that trajectories of neural population activity diverge depending on mice's behavioral choice. Consequently, simple classifier models trained on shell IC neuron activity can predict trial-by-trial outcomes, even when training data are restricted to neural activity occurring prior to mice's instrumental actions. Thus in behaving animals, auditory midbrain neurons transmit a population code that reflects a joint representation of sound and action.

Compression strategies for large-scale electrophysiology data

Objective.With the rapid adoption of high-density electrode arrays for recording neural activity, electrophysiology data volumes within labs and across the field are growing at unprecedented rates. For example, a one-hour recording with a 384-channel Neuropixels probe generates over 80 GB of raw data. These large data volumes carry a high cost, especially if researchers plan to store and analyze their data in the cloud. Thus, there is a pressing need for strategies that can reduce the data footprint of each experiment.Approach.Here, we establish a set of benchmarks for comparing the performance of various compression algorithms on experimental and simulated recordings from Neuropixels 1.0 (NP1) and 2.0 (NP2) probes.Main results.For lossless compression, audio codecs (FLACandWavPack) achieve compression ratios (CRs) 6% higher for NP1 and 10% higher for NP2 than the best general-purpose codecs, at the expense of decompression speed. For lossy compression, theWavPackalgorithm in 'hybrid mode' increases the CR from 3.59 to 7.08 for NP1 and from 2.27 to 7.04 for NP2 (compressed file size of ∼14% for both types of probes), without adverse effects on spike sorting accuracy or spike waveforms.Significance.Along with the tools we have developed to make compression easier to deploy, these results should encourage all electrophysiologists to apply compression as part of their standard analysis workflows.

Neural ensemble dynamics in trunk and hindlimb sensorimotor cortex encode for the control of postural stability

The cortex has a disputed role in monitoring postural equilibrium and intervening in cases of major postural disturbances. Here, we investigate the patterns of neural activity in the cortex that underlie neural dynamics during unexpected perturbations. In both the primary sensory (S1) and motor (M1) cortices of the rat, unique neuronal classes differentially covary their responses to distinguish different characteristics of applied postural perturbations; however, there is substantial information gain in M1, demonstrating a role for higher-order computations in motor control. A dynamical systems model of M1 activity and forces generated by the limbs reveals that these neuronal classes contribute to a low-dimensional manifold comprised of separate subspaces enabled by congruent and incongruent neural firing patterns that define different computations depending on the postural responses. These results inform how the cortex engages in postural control, directing work aiming to understand postural instability after neurological disease.

Synchrony-Division Neural Multiplexing: An Encoding Model

Cortical neurons receive mixed information from the collective spiking activities of primary sensory neurons in response to a sensory stimulus. A recent study demonstrated an abrupt increase or decrease in stimulus intensity and the stimulus intensity itself can be respectively represented by the synchronous and asynchronous spikes of S1 neurons in rats. This evidence capitalized on the ability of an ensemble of homogeneous neurons to multiplex, a coding strategy that was referred to as synchrony-division multiplexing (SDM). Although neural multiplexing can be conceived by distinct functions of individual neurons in a heterogeneous neural ensemble, the extent to which nearly identical neurons in a homogeneous neural ensemble encode multiple features of a mixed stimulus remains unknown. Here, we present a computational framework to provide a system-level understanding on how an ensemble of homogeneous neurons enable SDM. First, we simulate SDM with an ensemble of homogeneous conductance-based model neurons receiving a mixed stimulus comprising slow and fast features. Using feature-estimation techniques, we show that both features of the stimulus can be inferred from the generated spikes. Second, we utilize linear nonlinear (LNL) cascade models and calculate temporal filters and static nonlinearities of differentially synchronized spikes. We demonstrate that these filters and nonlinearities are distinct for synchronous and asynchronous spikes. Finally, we develop an augmented LNL cascade model as an encoding model for the SDM by combining individual LNLs calculated for each type of spike. The augmented LNL model reveals that a homogeneous neural ensemble model can perform two different functions, namely, temporal- and rate-coding, simultaneously.

Thin flexible arrays for long-term multi-electrode recordings in macaque primary visual cortex

Objective.Basic, translational and clinical neuroscience are increasingly focusing on large-scale invasive recordings of neuronal activity. However, in large animals such as nonhuman primates and humans-in which the larger brain size with sulci and gyri imposes additional challenges compared to rodents, there is a huge unmet need to record from hundreds of neurons simultaneously anywhere in the brain for long periods of time. Here, we tested the electrical and mechanical properties of thin, flexible multi-electrode arrays (MEAs) inserted into the primary visual cortex of two macaque monkeys, and assessed their magnetic resonance imaging (MRI) compatibility and their capacity to record extracellular activity over a period of 1 year.Approach.To allow insertion of the floating arrays into the visual cortex, the 20 by 100µm²shafts were temporarily strengthened by means of a resorbable poly(lactic-co-glycolic acid) coating.Main results. After manual insertion of the arrays, theex vivoandin vivoMRI compatibility of the arrays proved to be excellent. We recorded clear single-unit activity from up to 50% of the electrodes, and multi-unit activity (MUA) on 60%-100% of the electrodes, which allowed detailed measurements of the receptive fields and the orientation selectivity of the neurons. Even 1 year after insertion, we obtained significant MUA responses on 70%-100% of the electrodes, while the receptive fields remained remarkably stable over the entire recording period.Significance.Thus, the thin and flexible MEAs we tested offer several crucial advantages compared to existing arrays, most notably in terms of brain tissue compliance, scalability, and brain coverage. Future brain-machine interface applications in humans may strongly benefit from this new generation of chronically implanted MEAs.

Integrated neural dynamics of sensorimotor decisions and actions

Recent theoretical models suggest that deciding about actions and executing them are not implemented by completely distinct neural mechanisms but are instead two modes of an integrated dynamical system. Here, we investigate this proposal by examining how neural activity unfolds during a dynamic decision-making task within the high-dimensional space defined by the activity of cells in monkey dorsal premotor (PMd), primary motor (M1), and dorsolateral prefrontal cortex (dlPFC) as well as the external and internal segments of the globus pallidus (GPe, GPi). Dimensionality reduction shows that the four strongest components of neural activity are functionally interpretable, reflecting a state transition between deliberation and commitment, the transformation of sensory evidence into a choice, and the baseline and slope of the rising urgency to decide. Analysis of the contribution of each population to these components shows meaningful differences between regions but no distinct clusters within each region, consistent with an integrated dynamical system. During deliberation, cortical activity unfolds on a two-dimensional "decision manifold" defined by sensory evidence and urgency and falls off this manifold at the moment of commitment into a choice-dependent trajectory leading to movement initiation. The structure of the manifold varies between regions: In PMd, it is curved; in M1, it is nearly perfectly flat; and in dlPFC, it is almost entirely confined to the sensory evidence dimension. In contrast, pallidal activity during deliberation is primarily defined by urgency. We suggest that these findings reveal the distinct functional contributions of different brain regions to an integrated dynamical system governing action selection and execution.

Probing the Relationship Between Latent Linear Dynamical Systems and Low-Rank Recurrent Neural Network Models

A large body of work has suggested that neural populations exhibit low-dimensional dynamics during behavior. However, there are a variety of different approaches for modeling low-dimensional neural population activity. One approach involves latent linear dynamical system (LDS) models, in which population activity is described by a projection of low-dimensional latent variables with linear dynamics. A second approach involves low-rank recurrent neural networks (RNNs), in which population activity arises directly from a low-dimensional projection of past activity. Although these two modeling approaches have strong similarities, they arise in different contexts and tend to have different domains of application. Here we examine the precise relationship between latent LDS models and linear low-rank RNNs. When can one model class be converted to the other, and vice versa? We show that latent LDS models can only be converted to RNNs in specific limit cases, due to the non-Markovian property of latent LDS models. Conversely, we show that linear RNNs can be mapped onto LDS models, with latent dimensionality at most twice the rank of the RNN. A surprising consequence of our results is that a partially observed RNN is better represented by an LDS model than by an RNN consisting of only observed units.

Neural Algorithms and Circuits for Motor Planning

The brain plans and executes volitional movements. The underlying patterns of neural population activity have been explored in the context of movements of the eyes, limbs, tongue, and head in nonhuman primates and rodents. How do networks of neurons produce the slow neural dynamics that prepare specific movements and the fast dynamics that ultimately initiate these movements? Recent work exploits rapid and calibrated perturbations of neural activity to test specific dynamical systems models that are capable of producing the observed neural activity. These joint experimental and computational studies show that cortical dynamics during motor planning reflect fixed points of neural activity (attractors). Subcortical control signals reshape and move attractors over multiple timescales, causing commitment to specific actions and rapid transitions to movement execution. Experiments in rodents are beginning to reveal how these algorithms are implemented at the level of brain-wide neural circuits.

Cortical Control of Virtual Self-Motion Using Task-Specific Subspaces

Brain-machine interfaces (BMIs) for reaching have enjoyed continued performance improvements, yet there remains significant need for BMIs that control other movement classes. Recent scientific findings suggest that the intrinsic covariance structure of neural activity depends strongly on movement class, potentially necessitating different decode algorithms across classes. To address this possibility, we developed a self-motion BMI based on cortical activity as monkeys cycled a hand-held pedal to progress along a virtual track. Unlike during reaching, we found no high-variance dimensions that directly correlated with to-be-decoded variables. This was due to no neurons having consistent correlations between their responses and kinematic variables. Yet we could decode a single variable-self-motion-by nonlinearly leveraging structure that spanned multiple high-variance neural dimensions. Resulting online BMI-control success rates approached those during manual control. These findings make two broad points regarding how to build decode algorithms that harmonize with the empirical structure of neural activity in motor cortex. First, even when decoding from the same cortical region (e.g., arm-related motor cortex), different movement classes may need to employ very different strategies. Although correlations between neural activity and hand velocity are prominent during reaching tasks, they are not a fundamental property of motor cortex and cannot be counted on to be present in general. Second, although one generally desires a low-dimensional readout, it can be beneficial to leverage a multidimensional high-variance subspace. Fully embracing this approach requires highly nonlinear approaches tailored to the task at hand, but can produce near-native levels of performance.SIGNIFICANCE STATEMENT Many brain-machine interface decoders have been constructed for controlling movements normally performed with the arm. Yet it is unclear how these will function beyond the reach-like scenarios where they were developed. Existing decoders implicitly assume that neural covariance structure, and correlations with to-be-decoded kinematic variables, will be largely preserved across tasks. We find that the correlation between neural activity and hand kinematics, a feature typically exploited when decoding reach-like movements, is essentially absent during another task performed with the arm: cycling through a virtual environment. Nevertheless, the use of a different strategy, one focused on leveraging the highest-variance neural signals, supported high performance real-time brain-machine interface control.

Statistical neuroscience in the single trial limit

Individual neurons often produce highly variable responses over nominally identical trials, reflecting a mixture of intrinsic 'noise' and systematic changes in the animal's cognitive and behavioral state. Disentangling these sources of variability is of great scientific interest in its own right, but it is also increasingly inescapable as neuroscientists aspire to study more complex and naturalistic animal behaviors. In these settings, behavioral actions never repeat themselves exactly and may rarely do so even approximately. Thus, new statistical methods that extract reliable features of neural activity using few, if any, repeated trials are needed. Accurate statistical modeling in this severely trial-limited regime is challenging, but still possible if simplifying structure in neural data can be exploited. We review recent works that have identified different forms of simplifying structure - including shared gain modulations across neural subpopulations, temporal smoothness in neural firing rates, and correlations in responses across behavioral conditions - and exploited them to reveal novel insights into the trial-by-trial operation of neural circuits.

Monosynaptic inference via finely-timed spikes

Observations of finely-timed spike relationships in population recordings have been used to support partial reconstruction of neural microcircuit diagrams. In this approach, fine-timescale components of paired spike train interactions are isolated and subsequently attributed to synaptic parameters. Recent perturbation studies strengthen the case for such an inference, yet the complete set of measurements needed to calibrate statistical models is unavailable. To address this gap, we study features of pairwise spiking in a large-scale in vivo dataset where presynaptic neurons were explicitly decoupled from network activity by juxtacellular stimulation. We then construct biophysical models of paired spike trains to reproduce the observed phenomenology of in vivo monosynaptic interactions, including both fine-timescale spike-spike correlations and firing irregularity. A key characteristic of these models is that the paired neurons are coupled by rapidly-fluctuating background inputs. We quantify a monosynapse's causal effect by comparing the postsynaptic train with its counterfactual, when the monosynapse is removed. Subsequently, we develop statistical techniques for estimating this causal effect from the pre- and post-synaptic spike trains. A particular focus is the justification and application of a nonparametric separation of timescale principle to implement synaptic inference. Using simulated data generated from the biophysical models, we characterize the regimes in which the estimators accurately identify the monosynaptic effect. A secondary goal is to initiate a critical exploration of neurostatistical assumptions in terms of biophysical mechanisms, particularly with regards to the challenging but arguably fundamental issue of fast, unobservable nonstationarities in background dynamics.

Causal Network Inference for Neural Ensemble Activity

Interactions among cellular components forming a mesoscopic scale brain network (microcircuit) display characteristic neural dynamics. Analysis of microcircuits provides a system-level understanding of the neurobiology of health and disease. Causal discovery aims to detect causal relationships among variables based on observational data. A key barrier in causal discovery is the high dimensionality of the variable space. A method called Causal Inference for Microcircuits (CAIM) is proposed to reconstruct causal networks from calcium imaging or electrophysiology time series. CAIM combines neural recording, Bayesian network modeling, and neuron clustering. Validation experiments based on simulated data and a real-world reaching task dataset demonstrated that CAIM accurately revealed causal relationships among neural clusters.

Inferring Neuronal Couplings From Spiking Data Using a Systematic Procedure With a Statistical Criterion

Recent remarkable advances in experimental techniques have provided a background for inferring neuronal couplings from point process data that include a great number of neurons. Here, we propose a systematic procedure for pre- and postprocessing generic point process data in an objective manner to handle data in the framework of a binary simple statistical model, the Ising or generalized McCulloch-Pitts model. The procedure has two steps: (1) determining time bin size for transforming the point process data into discrete-time binary data and (2) screening relevant couplings from the estimated couplings. For the first step, we decide the optimal time bin size by introducing the null hypothesis that all neurons would fire independently, then choosing a time bin size so that the null hypothesis is rejected with the strict criteria. The likelihood associated with the null hypothesis is analytically evaluated and used for the rejection process. For the second postprocessing step, after a certain estimator of coupling is obtained based on the preprocessed data set (any estimator can be used with the proposed procedure), the estimate is compared with many other estimates derived from data sets obtained by randomizing the original data set in the time direction. We accept the original estimate as relevant only if its absolute value is sufficiently larger than those of randomized data sets. These manipulations suppress false positive couplings induced by statistical noise. We apply this inference procedure to spiking data from synthetic and in vitro neuronal networks. The results show that the proposed procedure identifies the presence or absence of synaptic couplings fairly well, including their signs, for the synthetic and experimental data. In particular, the results support that we can infer the physical connections of underlying systems in favorable situations, even when using a simple statistical model.

Frequency-separated principal component analysis of cortical population activity

A hallmark of neocortical activity is the presence of low-dimensional fluctuations in firing rate that are coordinated across neurons. However, the impact of these fluctuations on sensory processing remains unclear. Here, we examined fluctuations in populations of orientation-selective neurons from anesthetized macaque primary visual cortex (V1) during stimulus viewing as well as spontaneous activity. We introduce a novel approach termed frequency-separated principal component analysis (FS-PCA) to characterize these fluctuations. This method unveiled a distribution of components with a broad range of frequencies whose eigenvalues and variance followed an approximate power law. During stimulus viewing, subpopulations of V1 neurons correlated either positively or negatively with low-dimensional fluctuations. These two subpopulations displayed distinct activation properties and noise correlations in response to sensory input. Together, results suggest that slow, low-dimensional fluctuations in V1 population activity shape the response of individual neurons to oriented stimuli and may impact the transmission of sensory information to downstream regions of the primary visual system.NEW & NOTEWORTHY A method termed frequency-separated principal component analysis (FS-PCA) is introduced for analyzing populations of simultaneously recorded neurons. This framework extends standard principal component analysis by extracting components of activity delimited to specific frequency bands. FS-PCA revealed that circuits of the primary visual cortex generate a broad range of components dominated by low-frequency activity. Furthermore, low-dimensional fluctuations in population activity modulated the response of individual neurons to sensory input.

Identifying nonlinear dynamical systems via generative recurrent neural networks with applications to fMRI

A major tenet in theoretical neuroscience is that cognitive and behavioral processes are ultimately implemented in terms of the neural system dynamics. Accordingly, a major aim for the analysis of neurophysiological measurements should lie in the identification of the computational dynamics underlying task processing. Here we advance a state space model (SSM) based on generative piecewise-linear recurrent neural networks (PLRNN) to assess dynamics from neuroimaging data. In contrast to many other nonlinear time series models which have been proposed for reconstructing latent dynamics, our model is easily interpretable in neural terms, amenable to systematic dynamical systems analysis of the resulting set of equations, and can straightforwardly be transformed into an equivalent continuous-time dynamical system. The major contributions of this paper are the introduction of a new observation model suitable for functional magnetic resonance imaging (fMRI) coupled to the latent PLRNN, an efficient stepwise training procedure that forces the latent model to capture the 'true' underlying dynamics rather than just fitting (or predicting) the observations, and of an empirical measure based on the Kullback-Leibler divergence to evaluate from empirical time series how well this goal of approximating the underlying dynamics has been achieved. We validate and illustrate the power of our approach on simulated 'ground-truth' dynamical systems as well as on experimental fMRI time series, and demonstrate that the learnt dynamics harbors task-related nonlinear structure that a linear dynamical model fails to capture. Given that fMRI is one of the most common techniques for measuring brain activity non-invasively in human subjects, this approach may provide a novel step toward analyzing aberrant (nonlinear) dynamics for clinical assessment or neuroscientific research.

Effective Dimensionality Reduction for Visualizing Neural Dynamics by Laplacian Eigenmaps

With the development of neural recording technology, it has become possible to collect activities from hundreds or even thousands of neurons simultaneously. Visualization of neural population dynamics can help neuroscientists analyze large-scale neural activities efficiently. In this letter, Laplacian eigenmaps is applied to this task for the first time, and the experimental results show that the proposed method significantly outperforms the commonly used methods. This finding was confirmed by the systematic evaluation using nonhuman primate data, which contained the complex dynamics well suited for testing. According to our results, Laplacian eigenmaps is better than the other methods in various ways and can clearly visualize interesting biological phenomena related to neural dynamics.

Thirst regulates motivated behavior through modulation of brainwide neural population dynamics

Physiological needs produce motivational drives, such as thirst and hunger, that regulate behaviors essential to survival. Hypothalamic neurons sense these needs and must coordinate relevant brainwide neuronal activity to produce the appropriate behavior. We studied dynamics from ~24,000 neurons in 34 brain regions during thirst-motivated choice behavior in 21 mice as they consumed water and became sated. Water-predicting sensory cues elicited activity that rapidly spread throughout the brain of thirsty animals. These dynamics were gated by a brainwide mode of population activity that encoded motivational state. After satiation, focal optogenetic activation of hypothalamic thirst-sensing neurons returned global activity to the pre-satiation state. Thus, motivational states specify initial conditions that determine how a brainwide dynamical system transforms sensory input into behavioral output.

Structure and variability of delay activity in premotor cortex

Voluntary movements are widely considered to be planned before they are executed. Recent studies have hypothesized that neural activity in motor cortex during preparation acts as an ‘initial condition’ which seeds the proceeding neural dynamics. Here, we studied these initial conditions in detail by investigating 1) the organization of neural states for different reaches and 2) the variance of these neural states from trial to trial. We examined population-level responses in macaque premotor cortex (PMd) during the preparatory stage of an instructed-delay center-out reaching task with dense target configurations. We found that after target onset the neural activity on single trials converges to neural states that have a clear low-dimensional structure which is organized by both the reach endpoint and maximum speed of the following reach. Further, we found that variability of the neural states during preparation resembles the spatial variability of reaches made in the absence of visual feedback: there is less variability in direction than distance in neural state space. We also used offline decoding to understand the implications of this neural population structure for brain-machine interfaces (BMIs). We found that decoding of angle between reaches is dependent on reach distance, while decoding of arc-length is independent. Thus, it might be more appropriate to quantify decoding performance for discrete BMIs by using arc-length between reach end-points rather than the angle between them. Lastly, we show that in contrast to the common notion that direction can better be decoded than distance, their decoding capabilities are comparable. These results provide new insights into the dynamical neural processes that underline motor control and can inform the design of BMIs.

Early studies of premotor cortex explored how individual neurons directly encode aspects of an upcoming movement during preparation. Recent developments have proposed that the dynamics of populations of neurons underlie motor control, and that neural activity during preparation serves to set up these dynamics. While the dynamics of motor control have been studied extensively, several aspects of preparatory activity remain unresolved. Here, we ask how the patterns of neural activity during preparation for different reaches are related to one another. We found that the neural activity during preparation for reaches to different targets has a clear ‘structure’. Additionally, we found that the activity on a given trial was predictive of the initial trajectory of the reach. Lastly, we assessed the implications of our findings for predicting upcoming movements from neural activity, as in brain-machine interfaces.

Dynamics of motor cortical activity during naturalistic feeding behavior

OBJECTIVE

The orofacial primary motor cortex (MIo) plays a critical role in controlling tongue and jaw movements during oral motor functions, such as chewing, swallowing and speech. However, the neural mechanisms of MIo during naturalistic feeding are still poorly understood. There is a strong need for a systematic study of motor cortical dynamics during feeding behavior.

APPROACH

To investigate the neural dynamics and variability of MIo neuronal activity during naturalistic feeding, we used chronically implanted micro-electrode arrays to simultaneously recorded ensembles of neuronal activity in the MIo of two monkeys (Macaca mulatta) while eating various types of food. We developed a Bayesian nonparametric latent variable model to reveal latent structures of neuronal population activity of the MIo and identify the complex mapping between MIo ensemble spike activity and high-dimensional kinematics.

MAIN RESULTS

Rhythmic neuronal firing patterns and oscillatory dynamics are evident in single-unit activity. At the population level, we uncovered the neural dynamics of rhythmic chewing, and quantified the neural variability at multiple timescales (complete feeding sequences, chewing sequence stages, chewing gape cycle phases) across food types. Our approach accommodates time-warping of chewing sequences and automatic model selection, and maps the latent states to chewing behaviors at fine timescales.

SIGNIFICANCE

Our work shows that neural representations of MIo ensembles display spatiotemporal patterns in chewing gape cycles at different chew sequence stages, and these patterns vary in a stage-dependent manner. Unsupervised learning and decoding analysis may reveal the link between complex MIo spatiotemporal patterns and chewing kinematics.

Probabilistic Encoding Models for Multivariate Neural Data

A key problem in systems neuroscience is to characterize how populations of neurons encode information in their patterns of activity. An understanding of the encoding process is essential both for gaining insight into the origins of perception and for the development of brain-computer interfaces. However, this characterization is complicated by the highly variable nature of neural responses, and thus usually requires probabilistic methods for analysis. Drawing on techniques from statistical modeling and machine learning, we review recent methods for extracting important variables that quantitatively describe how sensory information is encoded in neural activity. In particular, we discuss methods for estimating receptive fields, modeling neural population dynamics, and inferring low dimensional latent structure from a population of neurons, in the context of both electrophysiology and calcium imaging data.

Model-based targeted dimensionality reduction for neuronal population data

Summarizing high-dimensional data using a small number of parameters is a ubiquitous first step in the analysis of neuronal population activity. Recently developed methods use "targeted" approaches that work by identifying multiple, distinct low-dimensional subspaces of activity that capture the population response to individual experimental task variables, such as the value of a presented stimulus or the behavior of the animal. These methods have gained attention because they decompose total neural activity into what are ostensibly different parts of a neuronal computation. However, existing targeted methods have been developed outside of the confines of probabilistic modeling, making some aspects of the procedures ad hoc, or limited in flexibility or interpretability. Here we propose a new model-based method for targeted dimensionality reduction based on a probabilistic generative model of the population response data. The low-dimensional structure of our model is expressed as a low-rank factorization of a linear regression model. We perform efficient inference using a combination of expectation maximization and direct maximization of the marginal likelihood. We also develop an efficient method for estimating the dimensionality of each subspace. We show that our approach outperforms alternative methods in both mean squared error of the parameter estimates, and in identifying the correct dimensionality of encoding using simulated data. We also show that our method provides more accurate inference of low-dimensional subspaces of activity than a competing algorithm, demixed PCA.

The Theta Rhythm Mixes and Matches Gamma Oscillations Cycle by Cycle

In the hippocampus, gamma power modulation by the theta rhythm is interpreted as the signature of temporally coordinated inputs that reflect ongoing processing. In this issue of Neuron, Lopes-Dos-Santos et al. (2018) develop a new methodology demonstrating that theta cycles can be viewed as individual computational units characterized by typical gamma profiles.

Decoding Cognitive Processes from Neural Ensembles

An intrinsic difficulty in studying cognitive processes is that they are unobservable states that exist in between observable responses to the sensory environment. Cognitive states must be inferred from indirect behavioral measures. Neuroscience potentially provides the tools necessary to measure cognitive processes directly, but it is challenged on two fronts. First, neuroscientific measures often lack the spatiotemporal resolution to identify the neural computations that underlie a cognitive process. Second, the activity of a single neuron, which is the fundamental building block of neural computation, is too noisy to provide accurate measurements of a cognitive process. In this paper, I examine recent developments in neurophysiological recording and analysis methods that provide a potential solution to these problems.

Different population dynamics in the supplementary motor area and motor cortex during reaching

Neural populations perform computations through their collective activity. Different computations likely require different population-level dynamics. We leverage this assumption to examine neural responses recorded from the supplementary motor area (SMA) and motor cortex. During visually guided reaching, the respective roles of these areas remain unclear; neurons in both areas exhibit preparation-related activity and complex patterns of movement-related activity. To explore population dynamics, we employ a novel "hypothesis-guided" dimensionality reduction approach. This approach reveals commonalities but also stark differences: linear population dynamics, dominated by rotations, are prominent in motor cortex but largely absent in SMA. In motor cortex, the observed dynamics produce patterns resembling muscle activity. Conversely, the non-rotational patterns in SMA co-vary with cues regarding when movement should be initiated. Thus, while SMA and motor cortex display superficially similar single-neuron responses during visually guided reaching, their different population dynamics indicate they are likely performing quite different computations.

Linking dynamic patterns of neural activity in orbitofrontal cortex with decision making

Humans and animals demonstrate extraordinary flexibility in choice behavior, particularly when deciding based on subjective preferences. We evaluate options on different scales, deliberate, and often change our minds. Little is known about the neural mechanisms that underlie these dynamic aspects of decision-making, although neural activity in orbitofrontal cortex (OFC) likely plays a central role. Recent evidence from studies in macaques shows that attention modulates value responses in OFC, and that ensembles of OFC neurons dynamically signal different options during choices. When contexts change, these ensembles flexibly remap to encode the new task. Determining how these dynamic patterns emerge and relate to choices will inform models of decision-making and OFC function.

Vigor of reaching movements: reward discounts the cost of effort

Making a movement may be thought of as an economic decision in which one spends effort to acquire reward. Time discounts reward, which predicts that the magnitude of reward should affect movement vigor: we should move faster, spending greater effort, when there is greater reward at stake. Indeed, saccade peak velocities are greater and reaction-times shorter when a target is paired with reward. In this study, we focused on human reaching and asked whether movement kinematics were affected by expectation of reward. Participants made out-and-back reaching movements to one of four quadrants of a 14-cm circle. During various periods of the experiment only one of the four quadrants was paired with reward, and the transition from reward to nonreward status of a quadrant occurred randomly. Our experiment design minimized dependence of reward on accuracy, granting the subjects wide latitude in self-selecting their movement speed, amplitude, and variability. When a quadrant was paired with reward, reaching movements had a shorter reaction time, higher peak velocity, and greater amplitude. Despite this greater vigor, movements toward the rewarded quadrant suffered from less variability: both reaction times and reach kinematics were less variable when there was expectation of reward. Importantly, the effect of reward on vigor was specific to the movement component that preceded the time of reward (outward reach), not the movement component that followed it (return reach). Our results suggest that expectation of reward not only increases vigor of human reaching but also decreases its variability. NEW & NOTEWORTHY Movements may be thought of as an economic transaction where the vigor of the movement represents the effort that the brain is willing to expend to acquire a rewarding state. We show that in reaching, reward discounts the cost of effort, producing movements with shorter reaction time, higher velocity, greater amplitude, and reduced reaction-time variability. These results complement earlier observations in saccades, suggesting a common principle of economics across modalities of motor control.

Computational Neuroscience: Mathematical and Statistical Perspectives

Mathematical and statistical models have played important roles in neuroscience, especially by describing the electrical activity of neurons recorded individually, or collectively across large networks. As the field moves forward rapidly, new challenges are emerging. For maximal effectiveness, those working to advance computational neuroscience will need to appreciate and exploit the complementary strengths of mechanistic theory and the statistical paradigm.

Neural mechanisms of movement planning: motor cortex and beyond

Neurons in motor cortex and connected brain regions fire in anticipation of specific movements, long before movement occurs. This neural activity reflects internal processes by which the brain plans and executes volitional movements. The study of motor planning offers an opportunity to understand how the structure and dynamics of neural circuits support persistent internal states and how these states influence behavior. Recent advances in large-scale neural recordings are beginning to decipher the relationship of the dynamics of populations of neurons during motor planning and movements. New behavioral tasks in rodents, together with quantified perturbations, link dynamics in specific nodes of neural circuits to behavior. These studies reveal a neural network distributed across multiple brain regions that collectively supports motor planning. We review recent advances and highlight areas where further work is needed to achieve a deeper understanding of the mechanisms underlying motor planning and related cognitive processes.

Fully integrated silicon probes for high-density recording of neural activity

Sensory, motor and cognitive operations involve the coordinated action of large neuronal populations across multiple brain regions in both superficial and deep structures. Existing extracellular probes record neural activity with excellent spatial and temporal (sub-millisecond) resolution, but from only a few dozen neurons per shank. Optical Ca2+ imaging offers more coverage but lacks the temporal resolution needed to distinguish individual spikes reliably and does not measure local field potentials. Until now, no technology compatible with use in unrestrained animals has combined high spatiotemporal resolution with large volume coverage. Here we design, fabricate and test a new silicon probe known as Neuropixels to meet this need. Each probe has 384 recording channels that can programmably address 960 complementary metal-oxide-semiconductor (CMOS) processing-compatible low-impedance TiN sites that tile a single 10-mm long, 70 × 20-μm cross-section shank. The 6 × 9-mm probe base is fabricated with the shank on a single chip. Voltage signals are filtered, amplified, multiplexed and digitized on the base, allowing the direct transmission of noise-free digital data from the probe. The combination of dense recording sites and high channel count yielded well-isolated spiking activity from hundreds of neurons per probe implanted in mice and rats. Using two probes, more than 700 well-isolated single neurons were recorded simultaneously from five brain structures in an awake mouse. The fully integrated functionality and small size of Neuropixels probes allowed large populations of neurons from several brain structures to be recorded in freely moving animals. This combination of high-performance electrode technology and scalable chip fabrication methods opens a path towards recording of brain-wide neural activity during behaviour.

Dynamic Structure of Neural Variability in the Cortical Representation of Speech Sounds

Accurate sensory discrimination is commonly believed to require precise representations in the nervous system; however, neural stimulus responses can be highly variable, even to identical stimuli. Recent studies suggest that cortical response variability decreases during stimulus processing, but the implications of such effects on stimulus discrimination are unclear. To address this, we examined electrocorticographic cortical field potential recordings from the human nonprimary auditory cortex (superior temporal gyrus) while subjects listened to speech syllables. Compared with a prestimulus baseline, activation variability decreased upon stimulus onset, similar to findings from microelectrode recordings in animal studies. We found that this decrease was simultaneous with encoding and spatially specific for those electrodes that most strongly discriminated speech sounds. We also found that variability was predominantly reduced in a correlated subspace across electrodes. We then compared signal and variability (noise) correlations and found that noise correlations reduce more for electrodes with strong signal correlations. Furthermore, we found that this decrease in variability is strongest in the high gamma band, which correlates with firing rate response. Together, these findings indicate that the structure of single-trial response variability is shaped to enhance discriminability despite non-stimulus-related noise.

SIGNIFICANCE STATEMENT

Cortical responses can be highly variable to auditory speech sounds. Despite this, sensory perception can be remarkably stable. Here, we recorded from the human superior temporal gyrus, a high-order auditory cortex, and studied the changes in the cortical representation of speech stimuli across multiple repetitions. We found that neural variability is reduced upon stimulus onset across electrodes that encode speech sounds.

Rapid face adaptation distributes representation in inferior-temporal cortex across time and neuronal dimensions

Neuronal networks of the brain adapt their information processing according to the history of stimuli. Whereas most studies have linked adaptation to repetition suppression, recurrent connections within a network and disinhibition due to adaptation predict more complex response patterns. The main questions of this study are as follows: what is the effect of the selectivity of neurons on suppression/enhancement of neural responses? What are the consequences of adaptation on information representation in neural population and the temporal structure of response patterns? We studied rapid face adaptation using spiking activities of neurons in the inferior-temporal (IT) cortex. Investigating the responses of neurons, within a wide range from negative to positive face selectivity, showed that despite the peak amplitude suppression in highly positive selective neurons, responses were enhanced in most other neurons. This enhancement can be attributed to disinhibition due to adaptation. Delayed and distributed responses were observed for positive selective neurons. Principal component analysis of the IT population responses over time revealed that repetition of face stimuli resulted in temporal decorrelation of the network activity. The contributions of the main and higher neuronal dimensions were changed under an adaptation condition, where more neuronal dimensions were used to encode repeated face stimuli.

In vitro biocompatibility and electrical stability of thick-film platinum/gold alloy electrodes printed on alumina

OBJECTIVE

High-density electrode arrays are a powerful tool in both clinical neuroscience and basic research. However, current manufacturing techniques require the use of specialised techniques and equipment, which are available to few labs. We have developed a high-density electrode array with customisable design, manufactured using simple printing techniques and with commercially available materials.

APPROACH

Electrode arrays were manufactured by thick-film printing a platinum-gold alloy (Pt/Au) and an insulating dielectric on 96% alumina ceramic plates. Arrays were conditioned in serum and serum-free conditions, with and without 1 kHz, 200 µA, charge balanced stimulation for up to 21 d. Array biocompatibility was assessed using an extract assay and a PC-12 cell contact assay. Electrode impedance, charge storage capacity and charge injection capacity were before and after array conditioning.

MAIN RESULTS

The manufactured Pt/Au electrodes have a highly porous surface and exhibit electrical properties comparable to arrays manufactured using alternative techniques. Materials used in array manufacture were found to be non-toxic to L929 fibroblasts by extract assay, and neuronal-like PC-12 cells adhered and extended neurites on the array surfaces. Arrays remained functional after long-term delivery of electrical pulses while exposed to protein-rich environments. Charge storage capacities and charge injection capacities increased following stimulation accounted for by an increase in surface index (real surface area) observed by vertical scanning interferometry. Further, we observed accumulation of proteins at the electrode sites following conditioning in the presence of serum.

SIGNIFICANCE

This study demonstrates the in vitro biocompatibility of commercially available thick-film printing materials. The printing technique is both simple and versatile, with layouts readily modified to produce customized electrode arrays. Thick-film electrode arrays are an attractive tool that may be implemented for general tissue engineering and neuroscience research.

RELATING ACCUMULATOR MODEL PARAMETERS AND NEURAL DYNAMICS

Accumulator models explain decision-making as an accumulation of evidence to a response threshold. Specific model parameters are associated with specific model mechanisms, such as the time when accumulation begins, the average rate of evidence accumulation, and the threshold. These mechanisms determine both the within-trial dynamics of evidence accumulation and the predicted behavior. Cognitive modelers usually infer what mechanisms vary during decision-making by seeing what parameters vary when a model is fitted to observed behavior. The recent identification of neural activity with evidence accumulation suggests that it may be possible to directly infer what mechanisms vary from an analysis of how neural dynamics vary. However, evidence accumulation is often noisy, and noise complicates the relationship between accumulator dynamics and the underlying mechanisms leading to those dynamics. To understand what kinds of inferences can be made about decision-making mechanisms based on measures of neural dynamics, we measured simulated accumulator model dynamics while systematically varying model parameters. In some cases, decision- making mechanisms can be directly inferred from dynamics, allowing us to distinguish between models that make identical behavioral predictions. In other cases, however, different parameterized mechanisms produce surprisingly similar dynamics, limiting the inferences that can be made based on measuring dynamics alone. Analyzing neural dynamics can provide a powerful tool to resolve model mimicry at the behavioral level, but we caution against drawing inferences based solely on neural analyses. Instead, simultaneous modeling of behavior and neural dynamics provides the most powerful approach to understand decision-making and likely other aspects of cognition and perception.

Neural Population Dynamics during Reaching Are Better Explained by a Dynamical System than Representational Tuning

Recent models of movement generation in motor cortex have sought to explain neural activity not as a function of movement parameters, known as representational models, but as a dynamical system acting at the level of the population. Despite evidence supporting this framework, the evaluation of representational models and their integration with dynamical systems is incomplete in the literature. Using a representational velocity-tuning based simulation of center-out reaching, we show that incorporating variable latency offsets between neural activity and kinematics is sufficient to generate rotational dynamics at the level of neural populations, a phenomenon observed in motor cortex. However, we developed a covariance-matched permutation test (CMPT) that reassigns neural data between task conditions independently for each neuron while maintaining overall neuron-to-neuron relationships, revealing that rotations based on the representational model did not uniquely depend on the underlying condition structure. In contrast, rotations based on either a dynamical model or motor cortex data depend on this relationship, providing evidence that the dynamical model more readily explains motor cortex activity. Importantly, implementing a recurrent neural network we demonstrate that both representational tuning properties and rotational dynamics emerge, providing evidence that a dynamical system can reproduce previous findings of representational tuning. Finally, using motor cortex data in combination with the CMPT, we show that results based on small numbers of neurons or conditions should be interpreted cautiously, potentially informing future experimental design. Together, our findings reinforce the view that representational models lack the explanatory power to describe complex aspects of single neuron and population level activity.

The question of how the brain generates movement has been extensively studied, yet multiple competing models exist. Representational approaches relate the activity of single neurons to movement parameters such as velocity and position, approaches useful for the decoding of movement intentions, while the dynamical systems approach predicts that neural activity should evolve in a predictable way based on population activity. Existing representational models cannot reproduce the recent finding in monkeys that predictable rotational patterns underlie motor cortex activity during reach initiation, a finding predicted by a dynamical model in which muscle activity is a direct combination of neural population rotations. However, previous simulations did not consider an essential aspect of representational models: variable time offsets between neurons and kinematics. Whereas these offsets reveal rotational patterns in the model, these rotations are statistically different from those observed in the brain and predicted by a dynamical model. Importantly, a simple recurrent neural network model also showed rotational patterns statistically similar to those observed in the brain, supporting the idea that dynamical systems-based approaches may provide a powerful explanation of motor cortex function.

A large field of view two-photon mesoscope with subcellular resolution for in vivo imaging

Imaging is used to map activity across populations of neurons. Microscopes with cellular resolution have small (.

Decoding subjective decisions from orbitofrontal cortex

When making a subjective choice, the brain must compute a value for each option and compare those values to make a decision. The orbitofrontal cortex (OFC) is critically involved in this process, but the neural mechanisms remain obscure, in part due to limitations in our ability to measure and control the internal deliberations that can alter the dynamics of the decision process. Here, we tracked the dynamics by recovering temporally precise neural states from multi-dimensional data in OFC. During individual choices, OFC alternated between states associated with the value of two available options, with dynamics that predicted whether a subject would decide quickly or vacillate between the two alternatives. Ensembles of value-encoding neurons contributed to these states, with individual neurons shifting activity patterns as the network evaluated each option. Thus, the mechanism of subjective decision-making involves the dynamic activation of OFC states associated with each choice alternative.

Investigating cell culture dynamics combining high density recordings with dimensional reduction techniques

High density multielectrode array recordings with CMOS-MEAs allow to monitor cell culture activity with unprecedent details respect to previous recording techniques. This is clarifying how network activity develops and is motivating the development of novel data analysis tools. Here, in order to advance in the exploitation of the richness of these large-scale array recordings, we introduce a principal component analysis approach that aims at improving on existing methodologies to describe neural activity events within large networks.

Representation of Muscle Synergies in the Primate Brain

Evidence suggests that the CNS uses motor primitives to simplify movement control, but whether it actually stores primitives instead of computing solutions on the fly to satisfy task demands is a controversial and still-unanswered possibility. Also in contention is whether these primitives take the form of time-invariant muscle coactivations ("spatial" synergies) or time-varying muscle commands ("spatiotemporal" synergies). Here, we examined forelimb muscle patterns and motor cortical spiking data in rhesus macaques (Macaca mulatta) handling objects of variable shape and size. From these data, we extracted both spatiotemporal and spatial synergies using non-negative decomposition. Each spatiotemporal synergy represents a sequence of muscular or neural activations that appeared to recur frequently during the animals' behavior. Key features of the spatiotemporal synergies (including their dimensionality, timing, and amplitude modulation) were independently observed in the muscular and neural data. In addition, both at the muscular and neural levels, these spatiotemporal synergies could be readily reconstructed as sequential activations of spatial synergies (a subset of those extracted independently from the task data), suggestive of a hierarchical relationship between the two levels of synergies. The possibility that motor cortex may execute even complex skill using spatiotemporal synergies has novel implications for the design of neuroprosthetic devices, which could gain computational efficiency by adopting the discrete and low-dimensional control that these primitives imply.

SIGNIFICANCE STATEMENT

We studied the motor cortical and forearm muscular activity of rhesus macaques (Macaca mulatta) as they reached, grasped, and carried objects of varied shape and size. We applied non-negative matrix factorization separately to the cortical and muscular data to reduce their dimensionality to a smaller set of time-varying "spatiotemporal" synergies. Each synergy represents a sequence of cortical or muscular activity that recurred frequently during the animals' behavior. Salient features of the synergies (including their dimensionality, timing, and amplitude modulation) were observed at both the cortical and muscular levels. The possibility that the brain may execute even complex behaviors using spatiotemporal synergies has implications for neuroprosthetic algorithm design, which could become more computationally efficient by adopting the discrete and low-dimensional control that they afford.

Capturing the temporal evolution of choice across prefrontal cortex

Activity in prefrontal cortex (PFC) has been richly described using economic models of choice. Yet such descriptions fail to capture the dynamics of decision formation. Describing dynamic neural processes has proven challenging due to the problem of indexing the internal state of PFC and its trial-by-trial variation. Using primate neurophysiology and human magnetoencephalography, we here recover a single-trial index of PFC internal states from multiple simultaneously recorded PFC subregions. This index can explain the origins of neural representations of economic variables in PFC. It describes the relationship between neural dynamics and behaviour in both human and monkey PFC, directly bridging between human neuroimaging data and underlying neuronal activity. Moreover, it reveals a functionally dissociable interaction between orbitofrontal cortex, anterior cingulate cortex and dorsolateral PFC in guiding cost-benefit decisions. We cast our observations in terms of a recurrent neural network model of choice, providing formal links to mechanistic dynamical accounts of decision-making.

DOI:http://dx.doi.org/10.7554/eLife.11945.001

In 1848, a railroad worker named Phineas Gage suffered an accident that was to secure him a place in neuroscience lore. While constructing a new railway line, a mistimed explosion propelled an iron bar into the base of his skull, where it passed behind his left eye before exiting through the top of his head. Gage survived the accident, but those who knew him reported significant changes in his personality and behaviour.

Gage’s ability to make decisions was particularly impaired by his injury. Decision-making involves weighing up the costs and benefits associated with alternative courses of action. It entails looking into the future to decide whether an anticipated reward will justify the effort or expense necessary to obtain it. This process is dependent on a region of the brain called the prefrontal cortex, the area that sustained the most damage in Phineas Gage.

While many studies have shown correlations between activity in particular parts of prefrontal cortex and the outcome of decisions, little is known about how this activity evolves over time as a decision is made. To explore this process, Hunt et al. trained macaque monkeys to choose between pairs of images that were associated with specific rewards (quantities of fruit juice) and costs (either amounts of work or fixed delays).

Electrode recordings revealed changes in prefrontal activity that varied over time as the monkeys deliberated over each pair of images, choosing for example between a large reward after a long delay versus a smaller reward immediately. This activity was consistent with a mathematical model of decision-making, which also explains data from brain imaging experiments in humans. This provides an important link between human data and electrode recordings in animals.

However, some of the patterns of activity observed in both macaques and humans appeared to reflect the speed at which decisions were made, rather than the outcome of the decisions themselves. By extracting information about decision speed on each decision from each region, it was shown that communication between regions of prefrontal cortex changes when choices are between two different amounts of work, as opposed to two different delays. Further experiments are needed to explore this phenomenon and to determine how other brain regions interact with the prefrontal cortex to support the decision-making process.

DOI:http://dx.doi.org/10.7554/eLife.11945.002

Predicting Reaction Time from the Neural State Space of the Premotor and Parietal Grasping Network

Neural networks of the brain involved in the planning and execution of grasping movements are not fully understood. The network formed by macaque anterior intraparietal area (AIP) and hand area (F5) of the ventral premotor cortex is implicated strongly in the generation of grasping movements. However, the differential role of each area in this frontoparietal network is unclear. We recorded spiking activity from many electrodes in parallel in AIP and F5 while three macaque monkeys (Macaca mulatta) performed a delayed grasping task. By analyzing neural population activity during action preparation, we found that state space analysis of simultaneously recorded units is significantly more predictive of subsequent reaction times (RTs) than traditional methods. Furthermore, because we observed a wide variety of individual unit characteristics, we developed the sign-corrected average rate (SCAR) method of neural population averaging. The SCAR method was able to explain at least as much variance in RT overall as state space methods. Overall, F5 activity predicted RT (18% variance explained) significantly better than AIP (6%). The SCAR methods provides a straightforward interpretation of population activity, although other state space methods could provide richer descriptions of population dynamics. Together, these results lend support to the differential role of the parietal and frontal cortices in preparation for grasping, suggesting that variability in preparatory activity in F5 has a more potent effect on trial-to-trial RT variability than AIP.

SIGNIFICANCE STATEMENT

Grasping movements are planned before they are executed, but how is the preparatory activity in a population of neurons related to the subsequent reaction time (RT)? A population analysis of the activity of many neurons recorded in parallel in macaque premotor (F5) and parietal (AIP) cortices during a delayed grasping task revealed that preparatory activity in F5 could explain a threefold larger fraction of variability in trial-to-trial RT than AIP. These striking differences lend additional support to a differential role of the parietal and premotor cortices in grasp movement preparation, suggesting that F5 has a more direct influence on trial-to-trial variability and movement timing, whereas AIP might be more closely linked to overall movement intentions.

Prefrontal spatial working memory network predicts animal's decision making in a free choice saccade task

While neurons in the lateral prefrontal cortex (PFC) encode spatial information during the performance of working memory tasks, they are also known to participate in subjective behavior such as spatial attention and action selection. In the present study, we analyzed the activity of primate PFC neurons during the performance of a free choice memory-guided saccade task in which the monkeys needed to choose a saccade direction by themselves. In trials when the receptive field location was subsequently chosen by the animal, PFC neurons with spatially selective visual response started to show greater activation before cue onset. This result suggests that the fluctuation of firing before cue presentation prematurely biased the representation of a certain spatial location and eventually encouraged the subsequent choice of that location. In addition, modulation of the activity by the animal's choice was observed only in neurons with high sustainability of activation and was also dependent on the spatial configuration of the visual cues. These findings were consistent with known characteristics of PFC neurons in information maintenance in spatial working memory function. These results suggest that precue fluctuation of spatial representation was shared and enhanced through the working memory network in the PFC and could finally influence the animal's free choice of saccade direction. The present study revealed that the PFC plays an important role in decision making in a free choice condition and that the dynamics of decision making are constrained by the network architecture embedded in this cortical area.