Dynamics of information and emergent computation in generic neural microcircuit models

Numerous methods have already been developed to estimate the information contained in single spike trains. In this article we explore efficient methods for estimating the information contained in the simultaneous firing activity of hundreds of neurons. Obviously such methods are needed to analyze data from multi-unit recordings. We test these methods on generic neural microcircuit models consisting of 800 neurons, and analyze the temporal dynamics of information about preceding spike inputs in such circuits. It turns out that information spreads with high speed in such generic neural microcircuit models, thereby supporting—without the postulation of any additional neural or synaptic mechanisms—the possibility of ultra-rapid computations on the first input spikes.     

Input prediction and autonomous movement analysis in recurrent circuits of spiking neurons

Temporal integration of information and prediction of future sensory inputs are assumed to be important computational tasks of generic cortical microcircuits. It has remained open how cortical microcircuits could possibly achieve this, especially since they consist--in contrast to most neural network models--of neurons and synapses with heterogeneous dynamic responses. It turns out, however, that the diversity of computational units increases the capability of microcircuit models for temporal integration. Furthermore the prediction of future input may be rather easy for such circuits since it suffices to train the readouts from such microcircuits. In this article we show that very simple readouts from a generic recurrently connected circuit of integrate-and-fire neurons with diverse dynamic synapses can be trained in an unsupervised manner to predict movements of different objects, that move within an unlimited number of combinations of speed, angle, and offset over a simulated sensory field. The autonomously trained microcircuit model is also able to compute the direction of motion, which is a computationally difficult problem ('aperture problem') since it requires disambiguation of local sensory readings through the context of other sensory readings at the current and preceding moments. Furthermore the same circuit can be trained simultaneously in a supervised manner to also report the shape and velocity of the moving object. Finally it is shown that the trained neural circuit supports novelty detection and the generation of 'imagined movements'. Altogether the results of this article suggest that it is not necessary to construct specific and biologically unrealistic neural circuit models for specific sensory processing tasks, since 'found' generic cortical microcircuit models in combination with very simple perceptron-like readouts can easily be trained to solve such computational tasks.     

Edge of chaos and prediction of computational performance for neural circuit models.

We analyze in this article the significance of the edge of chaos for real-time computations in neural microcircuit models consisting of spiking neurons and dynamic synapses. We find that the edge of chaos predicts quite well those values of circuit parameters that yield maximal computational performance. But obviously it makes no prediction of their computational performance for other parameter values. Therefore, we propose a new method for predicting the computational performance of neural microcircuit models. The new measure estimates directly the kernel property and the generalization capability of a neural microcircuit. We validate the proposed measure by comparing its prediction with direct evaluations of the computational performance of various neural microcircuit models. The proposed method also allows us to quantify differences in the computational performance and generalization capability of neural circuits in different dynamic regimes (UP- and DOWN-states) that have been demonstrated through intracellular recordings in vivo.     

Generation of spike latency tuning by thalamocortical circuits in auditory cortex

In many sensory systems, the latency of spike responses of individual neurons is found to be tuned for stimulus features and proposed to be used as a coding strategy. Whether the spike latency tuning is simply relayed along sensory ascending pathways or generated by local circuits remains unclear. Here, in vivo whole-cell recordings from rat auditory cortical neurons in layer 4 revealed that the onset latency of their aggregate thalamic input exhibited nearly flat tuning for sound frequency, whereas their spike latency tuning was much sharper with a broadly expanded dynamic range. This suggests that the spike latency tuning is not simply inherited from the thalamus, but can be largely reconstructed by local circuits in the cortex. Dissecting of thalamocortical circuits and neural modeling further revealed that broadly tuned intracortical inhibition prolongs the integration time for spike generation preferentially at off-optimal frequencies, while sharply tuned intracortical excitation shortens it selectively at the optimal frequency. Such push and pull mechanisms mediated likely by feedforward excitatory and inhibitory inputs respectively greatly sharpen the spike latency tuning and expand its dynamic range. The modulation of integration time by thalamocortical-like circuits may represent an efficient strategy for converting information spatially coded in synaptic strength to temporal representation.     

From Neuroelectrodynamics to Thinking Machines

Natural systems can provide excellent solutions to build artificial intelligent systems. The brain represents the best model of computation that leads to general intelligent action. However, current mainstream models reflect a weak understanding of computations performed in the brain that is translated in a failure of building powerful thinking machines. Specifically, temporal reductionist neural models elude the complexity of information processing since spike timing models reinforce the idea of neurons that compress temporal information and that computation can be reduced to a communication of information between neurons. The active brain dynamics and neuronal data analyses reveal multiple computational levels where information is intracellularly processed in neurons. New experimental findings and theoretical approach of neuroelectrodynamics challenge current models as they now stand and advocate for a change in paradigm for bio-inspired computing machines.     

Linking cortical spike pattern codes to auditory perception

Abstract Neurometric analysis has proven to be a powerful tool for studying links between neural activity and perception, especially in visual and somatosensory cortices, but conventional neurometrics are based on a simplistic rate-coding hypothesis that is clearly at odds with the rich and complex temporal spiking patterns evoked by many natural stimuli. In this study, we investigated the possible relationships between temporal spike pattern codes in the primary auditory cortex (A1) and the perceptual detection of subtle changes in the temporal structure of a natural sound. Using a two-alternative forced-choice oddity task, we measured the ability of human listeners to detect local time reversals in a marmoset twitter call. We also recorded responses of neurons in A1 of anesthetized and awake ferrets to these stimuli, and analyzed these responses using a novel neurometric approach that is sensitive to temporal discharge patterns. We found that although spike count-based neurometrics were inadequate to account for behavioral performance on this auditory task, neurometrics based on the temporal discharge patterns of populations of A1 units closely matched the psychometric performance curve, but only if the spiking patterns were resolved at temporal resolutions of 20 msec or better. These results demonstrate that neurometric discrimination curves can be calculated for temporal spiking patterns, and they suggest that such an extension of previous spike count-based approaches is likely to be essential for understanding the neural correlates of the perception of stimuli with a complex temporal structure.     

Neural timing nets.

Formulations of artificial neural networks are directly related to assumptions about neural coding in the brain. Traditional connectionist networks assume channel-based rate coding, while time-delay networks convert temporally-coded inputs into rate-coded outputs. Neural timing nets that operate on time structured input spike trains to produce meaningful time-structured outputs are proposed. Basic computational properties of simple feedforward and recurrent timing nets are outlined and applied to auditory computations. Feed-forward timing nets consist of arrays of coincidence detectors connected via tapped delay lines. These temporal sieves extract common spike patterns in their inputs that can subserve extraction of common fundamental frequencies (periodicity pitch) and common spectrum (timbre). Feedforward timing nets can also be used to separate time-shifted patterns, fusing patterns with similar internal temporal structure and spatially segregating different ones. Simple recurrent timing nets consisting of arrays of delay loops amplify and separate recurring time patterns. Single- and multichannel recurrent timing nets are presented that demonstrate the separation of concurrent, double vowels. Timing nets constitute a new and general neural network strategy for performing temporal computations on neural spike trains: extraction of common periodicities, detection of recurring temporal patterns, and formation and separation of invariant spike patterns that subserve auditory objects.     

Deterministic neural dynamics transmitted through neural networks

Precise spatiotemporal sequences of neuronal discharges (i.e., intervals between epochs repeating more often than expected by chance), have been observed in a large set of experimental electrophysiological recordings. Sensitivity to temporal information, by itself, does not demonstrate that dynamics embedded in spike trains can be transmitted through a neural network. This study analyzes how synaptic transmission through three archetypical types of neurons (regular-spiking, thalamo-cortical and resonator), simulated by a simple spiking model, can affect the transmission of precise timings generated by a nonlinear deterministic system (i.e., the Zaslavskii mapping in the present study). The results show that cells with subthreshold oscillations (resonators) are very sensitive to stochastic inputs, and are not a good candidate for transmitting temporally coded information. Thalamo-cortical neurons may transmit very well temporal patterns in the absence of background activity, but jitter accumulates along the synaptic chain. Conversely, we observed that cortical regular-spiking neurons can propagate filtered temporal information in a reliable way through the network, and with high temporal accuracy. We discuss the results in the general framework of neural dynamics and brain theories.     

A 4-dimensional representation of antennal lobe output based on an ensemble of characterized projection neurons

A central problem facing studies of neural encoding in sensory systems is how to accurately quantify the extent of spatial and temporal responses. In this study, we take advantage of the relatively simple and stereotypic neural architecture found in invertebrates. We combine standard electrophysiological techniques, recently developed population analysis techniques, and novel anatomical methods to form an innovative 4-dimensional view of odor output representations in the antennal lobe of the moth Manduca sexta. This novel approach allows quantification of olfactory responses of characterized neurons with spike time resolution. Additionally, arbitrary integration windows can be used for comparisons with other methods such as imaging. By assigning statistical significance to changes in neuronal firing, this method can visualize activity across the entire antennal lobe. The resulting 4-dimensional representation of antennal lobe output complements imaging and multi-unit experiments yet provides a more comprehensive and accurate view of glomerular activation patterns in spike time resolution.     

Spike Train Coactivity Encodes Learned Natural Stimulus Invariances in Songbird Auditory Cortex

The capacity for sensory systems to encode relevant information that is invariant to many stimulus changes is central to normal, real-world, cognitive function. This invariance is thought to be reflected in the complex spatiotemporal activity patterns of neural populations, but our understanding of population-level representational invariance remains coarse. Applied topology is a promising tool to discover invariant structure in large datasets. Here, we use topological techniques to characterize and compare the spatiotemporal pattern of coactive spiking within populations of simultaneously recorded neurons in the secondary auditory region caudal medial neostriatum of European starlings (Sturnus vulgaris). We show that the pattern of population spike train coactivity carries stimulus-specific structure that is not reducible to that of individual neurons. We then introduce a topology-based similarity measure for population coactivity that is sensitive to invariant stimulus structure and show that this measure captures invariant neural representations tied to the learned relationships between natural vocalizations. This demonstrates one mechanism whereby emergent stimulus properties can be encoded in population activity, and shows the potential of applied topology for understanding invariant representations in neural populations.SIGNIFICANCE STATEMENT Information in neural populations is carried by the temporal patterns of spikes. We applied novel mathematical tools from the field of algebraic topology to quantify the structure of these temporal patterns. We found that, in a secondary auditory region of a songbird, these patterns reflected invariant information about a learned stimulus relationship. These results demonstrate that topology provides a novel approach for characterizing neural responses that is sensitive to invariant relationships that are critical for the perception of natural stimuli.     

A diversity of synaptic filters are created by temporal summation of excitation and inhibition

Temporal filtering is a fundamental operation of nervous systems. In peripheral sensory systems, the temporal pattern of spiking activity can encode various stimulus qualities, and temporal filtering allows postsynaptic neurons to detect behaviorally relevant stimulus features from these spike trains. Intrinsic excitability, short-term synaptic plasticity, and voltage-dependent dendritic conductances have all been identified as mechanisms that can establish temporal filtering behavior in single neurons. Here we show that synaptic integration of temporally summating excitation and inhibition can establish diverse temporal filters of presynaptic input. Mormyrid electric fish communicate by varying the intervals between electric organ discharges. The timing of each discharge is coded by peripheral receptors into precisely timed spikes. Within the midbrain posterior exterolateral nucleus, temporal filtering by individual neurons results in selective responses to a particular range of presynaptic interspike intervals. These neurons are diverse in their temporal filtering properties, reflecting the wide range of intervals that must be detected during natural communication behavior. By manipulating presynaptic spike timing with high temporal resolution, we demonstrate that tuning to behaviorally relevant patterns of presynaptic input is similar in vivo and in vitro. We reveal that GABAergic inhibition plays a critical role in establishing different temporal filtering properties. Further, our results demonstrate that temporal summation of excitation and inhibition establishes selective responses to high and low rates of synaptic input, respectively. Simple models of synaptic integration reveal that variation in these two competing influences provides a basic mechanism for generating diverse temporal filters of synaptic input.     

Dopamine modulation of spike dynamics in bursting neurons

The pyloric network of the lobster stomatogastric ganglion is a prime example of an oscillatory neural circuit. In our previous study on the firing patterns of pyloric neurons we observed characteristic temporal structures termed 'interspike interval (ISI) signatures' which were found to depend on the synaptic connectivity of the network. Dopamine, a well-known modulator of the pyloric network, is known to affect inhibitory synapses so it might also tune the fine temporal structure of intraburst spikes, a phenomenon not previously investigated. In the recent work we study the DA modulation of ISI patterns of two identified pyloric neurons in normal conditions and after blocking their glutamatergic synaptic connections. Dopamine (10-50 microM) strongly regularizes the firing of the lateral pyloric (LP) and pyloric dilator (PD) neurons by increasing the reliability of recurrent spike patterns. The most dramatic effect is observed in the LP, where precisely replicated spike multiplets appear in a normally 'noisy' neuron. The DA-induced regularization of intraburst spike patterns requires functional glutamatergic inputs to the LP neuron and this effect cannot be mimicked by simple intracellular depolarization. Inhibitory synaptic inputs arriving before the bursts are important factors in shaping the intraburst spike dynamics of both the PD and the LP neurons. Our data reveal a novel aspect of chemical neuromodulation in oscillatory neural networks. This effect sets in at concentrations lower than those affecting the overall burst pattern of the network. The sensitivity of intraburst spike dynamics to preceding synaptic inputs also suggests a novel method of temporal coding in neural bursters.     

Principles of visual motion detection

Motion information is required for the solution of many complex tasks of the visual system such as depth perception by motion parallax and figure/ground discrimination by relative motion. However, motion information is not explicitly encoded at the level of the retinal input. Instead, it has to be computed from the time-dependent brightness patterns of the retinal image as sensed by the two-dimensional array of photoreceptors. Different models have been proposed which describe the neural computations underlying motion detection in various ways. To what extent do biological motion detectors approximate any of these models? As will be argued here, there is increasing evidence from the different disciplines studying biological motion vision, that, throughout the animal kingdom ranging from invertebrates to vertebrates including man, the mechanisms underlying motion detection can be attributed to only a few, essentially equivalent computational principles. Motion detection may, therefore, be one of the first examples in computational neurosciences where common principles can be found not only at the cellular level (e.g., dendritic integration, spike propagation, synaptic transmission) but also at the level of computations performed by small neural networks.     

Multineuronal vectorization is more efficient than time-segmental vectorization for information extraction from neuronal activities in the inferior temporal cortex

In order for patients with disabilities to control assistive devices with their own neural activity, multineuronal spike trains must be efficiently decoded because only limited computational resources can be used to generate prosthetic control signals in portable real-time applications. In this study, we compare the abilities of two vectorizing procedures (multineuronal and time-segmental) to extract information from spike trains during the same total neuron-seconds. In the multineuronal vectorizing procedure, we defined a response vector whose components represented the spike counts of one to five neurons. In the time-segmental vectorizing procedure, a response vector consisted of components representing a neuron’s spike counts for one to five time-segment(s) of a response period of 1 s. Spike trains were recorded from neurons in the inferior temporal cortex of monkeys presented with visual stimuli. We examined whether the amount of information of the visual stimuli carried by these neurons differed between the two vectorizing procedures. The amount of information calculated with the multineuronal vectorizing procedure, but not the time-segmental vectorizing procedure, significantly increased with the dimensions of the response vector. We conclude that the multineuronal vectorizing procedure is superior to the time-segmental vectorizing procedure in efficiently extracting information from neuronal signals.     

Neurons with stereotyped and rapid responses provide a reference frame for relative temporal coding in primate auditory cortex

The precise timing of spikes of cortical neurons relative to stimulus onset carries substantial sensory information. To access this information the sensory systems would need to maintain an internal temporal reference that reflects the precise stimulus timing. Whether and how sensory systems implement such reference frames to decode time-dependent responses, however, remains debated. Studying the encoding of naturalistic sounds in primate (Macaca mulatta) auditory cortex we here investigate potential intrinsic references for decoding temporally precise information. Within the population of recorded neurons, we found one subset responding with stereotyped fast latencies that varied little across trials or stimuli, while the remaining neurons had stimulus-modulated responses with longer and variable latencies. Computational analysis demonstrated that the neurons with stereotyped short latencies constitute an effective temporal reference for relative coding. Using the response onset of a simultaneously recorded stereotyped neuron allowed decoding most of the stimulus information carried by onset latencies and the full spike train of stimulus-modulated neurons. Computational modeling showed that few tens of such stereotyped reference neurons suffice to recover nearly all information that would be available when decoding the same responses relative to the actual stimulus onset. These findings reveal an explicit neural signature of an intrinsic reference for decoding temporal response patterns in the auditory cortex of alert animals. Furthermore, they highlight a role for apparently unselective neurons as an early saliency signal that provides a temporal reference for extracting stimulus information from other neurons.     

Real-time covariance computer for cell assemblies is based on neuronal principles

Progress in understanding neural information processing in the brain requires the ability to study the correlated activities of neuron groups. However, which neurons belong to a functional group? The answer to this question depends on the availability of multiple electrode arrays and real-time coupling analyzers. This paper describes such an instrument; electrode arrays from our group have been described elsewhere. A real-time covariance computer has been built embodying certain nearly universal neural principles. The spike trains from each recorded unit pass through leaky integrators which mimic the "postsynaptic potentials" (PSPs). The subsequent covariance calculations are performed by multiplications of all possible PSP-channel combinations. This is greatly simplified by the use of analog-to-stochastic converters, which encode the PSPs to stochastic two-level signals. Multiplications, thus, can simply be carried out by AND gates. With a second set of leaky integrators the participations of the single neurons in the joint activity of the group are derived. This operation mimics the spatial and temporal integration at the membranes of the dendritic tree and soma. For each neuron the actual coupling values are displayed in real time via LED bargraph displays. A 4-channel prototype of this instrument was used to analyze covariant activities in the cat's visual cortex.     

Predictive Visual Motion Extrapolation Emerges Spontaneously and without Supervision at Each Layer of a Hierarchical Neural Network with Spike-Timing-Dependent Plasticity

The fact that the transmission and processing of visual information in the brain takes time presents a problem for the accurate real-time localization of a moving object. One way this problem might be solved is extrapolation: using an object''s past trajectory to predict its location in the present moment. Here, we investigate how a simulated in silico layered neural network might implement such extrapolation mechanisms, and how the necessary neural circuits might develop. We allowed an unsupervised hierarchical network of velocity-tuned neurons to learn its connectivity through spike-timing-dependent plasticity (STDP). We show that the temporal contingencies between the different neural populations that are activated by an object as it moves causes the receptive fields of higher-level neurons to shift in the direction opposite to their preferred direction of motion. The result is that neural populations spontaneously start to represent moving objects as being further along their trajectory than where they were physically detected. Because of the inherent delays of neural transmission, this effectively compensates for (part of) those delays by bringing the represented position of a moving object closer to its instantaneous position in the world. Finally, we show that this model accurately predicts the pattern of perceptual mislocalization that arises when human observers are required to localize a moving object relative to a flashed static object (the flash-lag effect; FLE).SIGNIFICANCE STATEMENT Our ability to track and respond to rapidly changing visual stimuli, such as a fast-moving tennis ball, indicates that the brain is capable of extrapolating the trajectory of a moving object to predict its current position, despite the delays that result from neural transmission. Here, we show how the neural circuits underlying this ability can be learned through spike-timing-dependent synaptic plasticity and that these circuits emerge spontaneously and without supervision. This demonstrates how the neural transmission delays can, in part, be compensated to implement the extrapolation mechanisms required to predict where a moving object is at the present moment.     

A Fast and Simple Population Code for Orientation in Primate V1

Orientation tuning has been a classic model for understanding single-neuron computation in the neocortex. However, little is known about how orientation can be read out from the activity of neural populations, in particular in alert animals. Our study is a first step toward that goal. We recorded from up to 20 well isolated single neurons in the primary visual cortex of alert macaques simultaneously and applied a simple, neurally plausible decoder to read out the population code. We focus on two questions: First, what are the time course and the timescale at which orientation can be read out from the population response? Second, how complex does the decoding mechanism in a downstream neuron have to be to reliably discriminate between visual stimuli with different orientations? We show that the neural ensembles in primary visual cortex of awake macaques represent orientation in a way that facilitates a fast and simple readout mechanism: With an average latency of 30-80 ms, the population code can be read out instantaneously with a short integration time of only tens of milliseconds, and neither stimulus contrast nor correlations need to be taken into account to compute the optimal synaptic weight pattern. Our study shows that-similar to the case of single-neuron computation-the representation of orientation in the spike patterns of neural populations can serve as an exemplary case for understanding the computations performed by neural ensembles underlying visual processing during behavior.     

Dendritic plasticity in the adult neocortex.

Dendrites are a major determinant of how neurons integrate and process incoming information, and thus, they play a vital role in the functional properties of neural circuits. During the past 30 years, it has become clear that the functional organization of the adult neocortex (and other parts of the brain) is dynamic and can change in response to experimental manipulations that affect cortical activity patterns. A variety of changes in the cortical microcircuit, including changes in synaptic function, neuronal membrane properties, and axonal trajectories, are associated with these functional changes. In this review, the authors discuss changes in the structure of dendrites in the adult neocortex, the sorts of experimental manipulations that induce these changes, and some possible molecular mechanisms that may underlie the changes.     

Conditioned spikes: a simple and fast method to represent rates and temporal patterns in multielectrode recordings

Increasing evidence suggests that the brain utilizes distributed codes that can only be analyzed by simultaneously recording the activity of multiple neurons. This paper introduces a new methodology for studying neural ensemble recordings. The method uses a novel representation to provide complementary information about the stimuli which are contained in the temporal pattern of the spike sequence. By using this procedure, a high correlation of synchronized events with stimuli times is apparent. To quantify the results and to compare the performance of this method against the most traditional raster plot, we have used Fano factor and cross-correlation analysis. Our results suggest that several consecutive spikes from different neurons within an extended time window may encode behaviorally relevant information. We propose that this new representation, in addition to the other approaches currently used (standard raster plots, multivariate statistical methods, neuronal networks, information theory, etc.), can be a useful procedure to describe population spike dynamics.