Spatial patterns in a network composed of neurons with different excitabilities induced by autapse

It has been identified that autapse can modulate dynamics of single neurons and spatial patterns of neuronal networks. In the present paper, based on the results that autapse can induce type II excitability changed to type I excitability, spatial pattern transitions are simulated in a two-dimensional neuronal network composed of excitatory coupled neurons with autapse which can induce excitability transition. Different spatial patterns including random-like pattern, irregular wave, regular wave, and nearly synchronous behavior are simulated with increasing the percentage (σ) of neurons with type I excitability. When noise is introduced, spiral waves are induced. By calculating signal-to-noise ratio from the spatial structure function and the mean firing probability of neurons, regular waves and spiral waves exhibit optimal spatial correlation, implying the occurrence of spatial coherence resonance phenomenon. The changes of mean firing probability of neurons show that different firing frequency between type I excitability and type II excitability may be an important factor to modulate the spatial patterns. The results are helpful to understand the spatial patterns including spiral waves observed in the biological experiment on the rat cortex perfused with drugs which can induce single neurons changed from type II excitability to type I excitability and block the inhibitory couplings between neurons. The excitability transition, absence of inhibitory coupling, noise as well as the autapse are important factors to modulate the spatial patterns including spiral waves.     

Neuronal-like Irregular Spiking Dynamics in Highly Volatile Memristive Intermediate-scale AgPt-Nanoparticle Assemblies

Neuromorphic computing seeks functional materials capable of emulating brain-like dynamics to solve computational problems with time and energy efficiency, outclassing current transistor-based hardware architectures. Major efforts are focused on integrating memristive devices into highly regular circuits (i.e., crossbar arrays), where the information representation in individual memristive devices is closely oriented toward the behavior of artificial neurons. However, artificial neurons are rather rigid mathematical concepts than realistic projections of complex neuronal dynamics. Neuroscience suggests that highly efficient information representation on the level of individual neurons relies on dynamical features such as excitatory and inhibitory contributions, irregularity of firing patterns, and temporal correlations. Here, a conductive atomic force microscopy approach is applied to probe the memristive dynamics of nanoscale assemblies of AgPt-nanoparticles at the stability border of the conducting state, where physical forces causing the formation and decay of filamentary structures appear to be balanced. This unveils a dynamic regime, where the memristive response is governed by irregular firing patterns. The significance of such a dynamical regime is motivated by close similarities to excitation and inhibition-governed behavior in biological neuronal systems, which is crucial to tune biological neuronal systems into a state most suitable for information representation and computation.     

Phoneme representation and classification in primary auditory cortex

A controversial issue in neurolinguistics is whether basic neural auditory representations found in many animals can account for human perception of speech. This question was addressed by examining how a population of neurons in the primary auditory cortex (A1) of the naive awake ferret encodes phonemes and whether this representation could account for the human ability to discriminate them. When neural responses were characterized and ordered by spectral tuning and dynamics, perceptually significant features including formant patterns in vowels and place and manner of articulation in consonants, were readily visualized by activity in distinct neural subpopulations. Furthermore, these responses faithfully encoded the similarity between the acoustic features of these phonemes. A simple classifier trained on the neural representation was able to simulate human phoneme confusion when tested with novel exemplars. These results suggest that A1 responses are sufficiently rich to encode and discriminate phoneme classes and that humans and animals may build upon the same general acoustic representations to learn boundaries for categorical and robust sound classification.     

The role of neural architecture and the speed of signal propagation in the process of synchronization of bursting neurons

Synchronized neuronal activity has been observed at all levels of human and any other nervous systems and was suggested as particularly relevant in information processing and coding. In the present paper we investigate the synchronization of bursting neuronal activity. Motivated by the fact that in neural systems the interplay between the network structure and the dynamics taking place on it is closely interrelated, we develop a spatial network representation of neural architecture in which we can tune the network organization between a scale-free network with dominating long-range connections and a homogeneous network with mostly adjacent neurons connected. Our results reveal that the most synchronized response is obtained for the intermediate regime where long- as well as short-range connections constitute the neural architecture. Moreover, the optimal response is additionally enhanced when the speed of signal propagation is optimized.     

Optimal control of neuronal activity

We investigate the optimal control of neuronal spiking activity for neurons receiving a class of random synaptic inputs, characterized by a positive parameter alpha. Optimal control signals and optimal variances are found exactly for the diffusion process approximating an integrate and fire model. When synaptic inputs are "sub-Poisson" (alpha<0.5), we find that the optimal synaptic input is a delta function (corresponding to bang-bang control) and the optimal signal is not unique. Poisson synaptic input is the critical case: The control signal is unique, but the control signal is still a delta function. For "supra-Poisson" (alpha>0.5) inputs, the optimal control is smooth and unique. The optimal variance obtained in the current paper sets the lowest possible bound in controlling the stochasticity of neuronal activity. We also discuss how to implement the optimal control signal for certain model neurons.     

Cortical representation of spatiotemporal pattern of firing evoked by echolocation signals: population encoding of target features in real time.

Target perception in echolocating bats entails the generation of an acoustic image of the target in the auditory cortex. By integrating information conveyed in the sequence of acoustic echoes, the population of cortical neurons in hypothesized to encode different target features based on its spatiotemporal pattern of neural-spike firing during the course of echolocation. A biologically plausible approach to the cortical representation of target features is employed by using electrophysiological data recorded from the auditory cortex of the FM bat, Myotis lucifugus. A single-neuron model of delay-sensitive neurons is first approximated by the formulation of a Gaussian function with different variables to represent the delay-tuning properties of individual cortical neurons. A cortical region consisting of delay-sensitive neurons organized topographically according to best frequency (i.e., tontopically organized) is then modeled with multiple layers of the single-neuron model. A mechanism is developed to represent and encode the responses of these neurons based on time-dependent, incoming echo signals. The time-varying responses of the population of neurons are mapped spatially on the auditory-cortical surface as a cortical response map (CORMAP). The model is tested using phantom targets with single and multiple glints. These simulation results provide further validation of the current auditory framework as a biomimetic mechanism for capturing time-varying, acoustic stimuli impinging in the bat's ears, and the neural representation of acoustic stimulus features by saptiotemporal-firing patterns in the cortical population.     

Dynamical entropy production in spiking neuron networks in the balanced state

We demonstrate deterministic extensive chaos in the dynamics of large sparse networks of theta neurons in the balanced state. The analysis is based on numerically exact calculations of the full spectrum of Lyapunov exponents, the entropy production rate, and the attractor dimension. Extensive chaos is found in inhibitory networks and becomes more intense when an excitatory population is included. We find a strikingly high rate of entropy production that would limit information representation in cortical spike patterns to the immediate stimulus response.     

Neural control of vocalization: Respiratory and emotional influences

Previous research has shown that a region of the midbrain, the periaqueductal gray matter (PAG), is critical for vocalization. In this review, we describe the results of previous investigations in which we sought to find out how PAG neurons integrate the activity and precise timing of respiratory, laryngeal, and oral muscle activity for natural-sounding vocalization using the technique of excitatory amino acid microinjections in cats. In these studies, all surgical procedures were carried out under deep anaesthesia. In the precollicular decerebrate cat two general types of vocalization, classified as voiced and unvoiced, could be evoked by exciting neurons in the lateral part of the intermediate part of the PAG. The patterns of evoked electromyographic activity were strikingly similar to previously reported patterns of human muscle activity. Coordinated patterns of activity were evoked with just-threshold excitation leading to the conclusion that patterned muscle activity corresponding to the major categories of voiced and voiceless sound production are represented in the PAG. In a parallel series of human and animal experiments, we also determined that the speech and vocalization respiratory patterns are integrated and coordinated with afferent signals related to lung volume. These data have led to the proposal of a new hypothesis for the neural control of vocalization: that the PAG is a crucial brain site for mammalian voice production, not only in the production of emotional or involuntary sounds, but also as a generator of specific respiratory and laryngeal motor patterns essential for human speech and song     

Firing activity of complex space-clamped FitzHugh-Nagumo neural networks

We investigate how firing activity of complex neural networks depends on the random long-range connections and coupling strength. Network elements are described by excitable space-clamped FitzHugh-Nagumo (SCFHN) neurons with the values of parameters at which no firing activity occurs. It is found that for a given appropriate coupling strength C, there exists a critical fraction of random connections (or randomness) p^*, such that if p > p^* the firing neurons, which are absent in the nearest-neighbor network, occur. The firing activity becomes more frequent as randomness p is further increased. On the other hand, when the p is smaller, there are no active neurons in network, no matter what the value of C is. For a given larger p, there exist optimal coupling strength levels, where firing activity reaches its maximum. To the best of our knowledge, this is a novel mechanism for the emergence of firing activity in neurons.     

FitzHugh-Nagumo神经元网络的联想记忆与分割

以广泛讨论的Fitz Hugh-Nagumo神经元节点组成脉动神经元网络,从神经系统空时模式编码理论研究网络的记忆(或模式)存储与时间分割问题.给定一个输入模式,它是几种模式的叠加,网络能够以一部分神经元同步发放的形式一个接一个地分割出每一种模式.如果输入的模式有缺损,系统能够把它们恢复成原型,即神经网络的联想记忆功能.模拟需要调节耦合强度和噪声强度等参数使得网络在特定的参数值和中等强度噪声达到最优的时间分割,与广泛讨论的随机共振现象一致.     

Structured information in sparse-code metric neural networks

Sparse-code networks have retrieval abilities which are strongly dependent on the firing threshold for the neurons. If the connections are spatially uniform, the macroscopic properties of the network can be measured by the overlap between neurons and learned patterns, and by the global activity. However, for nonuniform networks, for instance small-world networks, the neurons can retrieve fragments of patterns without performing global retrieval. Local overlaps are needed to describe the network. We characterize the structure type of the neural states using a parameter that is related to fluctuations of the local overlaps, with distinction between bump and block phases. Simulation of neural dynamics shows a competition between localized (bump), structured (block) and global retrieval. When the network topology randomness increases, the phase-diagram shows a transition from local to global retrieval. Furthermore, the local phase splits into a bump phase for low activity and a block phase for high activity. A theoretical approach solves the asymptotic limit of the model, and confirms the simulation results which predicts the change of stability from bumps to blocks when the storage ratio increases.     

Information Hiding Based on Statistical Features of Self-Organizing Patterns

A computational technique for the determination of optimal hiding conditions of a digital image in a self-organizing pattern is presented in this paper. Three statistical features of the developing pattern (the Wada index based on the weighted and truncated Shannon entropy, the mean of the brightness of the pattern, and the p-value of the Kolmogorov-Smirnov criterion for the normality testing of the distribution function) are used for that purpose. The transition from the small-scale chaos of the initial conditions to the large-scale chaos of the developed pattern is observed during the evolution of the self-organizing system. Computational experiments are performed with the stripe-type patterns, spot-type patterns, and unstable patterns. It appears that optimal image hiding conditions are secured when the Wada index stabilizes after the initial decline, the mean of the brightness of the pattern remains stable before dropping down significantly below the average, and the p-value indicates that the distribution becomes Gaussian.     

Postnatal functional development of the dorsal and posteroventral cochlear nuclei of the cat

The postnatal development of firing patterns and response areas was determined for single neurons in the dorsal (DCN) and posteroventral ( PVCN ) cochlear nuclei of the kitten. Extracellular, single-unit responses to pure-tone stimulation were recorded in ketamine and sodium pentobarbital anesthetized kittens between the ages of 5 and 52 days. Within the first two weeks of postnatal life threshold is high, first-spike latency is long, and maximal discharge rate is low as compared to older kittens and adult cats. Prior to the end of the second postnatal week the tone-evoked temporal discharge patterns that characterize neurons of the DCN and PVCN in the adult cat are routinely recorded. These patterns, which appear within the first 50 ms of tonal stimulation, include the so-called " primarylike ," "chopper," " pauser ," "buildup," and "onset" types and their variants. In animals younger than about 10-12 days of age, the driven activity that occurs later than about 50 ms after stimulus onset often is not sustained, but breaks up during the stimulus into bursts that are separated by intervals of about 100-150 ms. Also within the first two weeks of postnatal life, many of the response-area properties of DCN and PVCN neurons are similar to those recorded in adult cats. The excitation and inhibition found within the so-called type II/III, type IV, and type V response areas of the adult occur in this early postnatal period. We conclude that many of the cellular mechanisms that underlie the temporal firing patterns and the organization of the response areas of DCN and PVCN neurons are active in the growing, differentiating cochlear nuclei and that the emergence of these mechanisms does not depend on afferent activity generated in the cochlear and auditory nerve by the animal's acoustic environment. Furthermore, if temporal firing patterns and response-area profiles remain relatively constant over the life span of the animal, then so must the spatial and temporal relationships of the inputs that produce and maintain them as these neurons, and the circuits of which they are a part, grow in size and complexity.     

A few strong connections: optimizing information retention in neuronal avalanches

Background  

How living neural networks retain information is still incompletely understood. Two prominent ideas on this topic have developed in parallel, but have remained somewhat unconnected. The first of these, the "synaptic hypothesis," holds that information can be retained in synaptic connection strengths, or weights, between neurons. Recent work inspired by statistical mechanics has suggested that networks will retain the most information when their weights are distributed in a skewed manner, with many weak weights and only a few strong ones. The second of these ideas is that information can be represented by stable activity patterns. Multineuron recordings have shown that sequences of neural activity distributed over many neurons are repeated above chance levels when animals perform well-learned tasks. Although these two ideas are compelling, no one to our knowledge has yet linked the predicted optimum distribution of weights to stable activity patterns actually observed in living neural networks.     

Neural network models of sound localization based on directional filtering by the pinna.

Three-layer neural-network functions were developed to transform spectral representations of pinna-filtered stimuli at the input to a space-mapped representation of sound-source direction at the output. The inputs are modeled after transfer functions of the external ear of the cat; the output is modeled on the spatial sensitivity of superior colliculus neurons. Network solutions are obtained by backpropagation and by a method that enforces uniform task distribution in the hidden layer of the model. Solutions are characterized using bandlimited inputs to study the relative strength of potential sound localization cues in various frequency regions. This analysis suggests that the frequency region containing the first spectral notch (5-18 kHz) provides the best localization cues. Response properties of model neurons were studied using input patterns modeled after auditory nerve response profiles to pure tones at various frequencies and sound levels. The response properties of hidden layer model neurons resemble cochlear nucleus types III and IV and their composites. Neurons in both hidden and output layers show the properties of spectral notch detectors. Although neural networks have limitations as models of real neural systems, the results illustrate how they can provide insight into the computation of complex transformations in the nervous system.     

The neural processing of taste

Although there have been many recent advances in the field of gustatory neurobiology, our knowledge of how the nervous system is organized to process information about taste is still far from complete. Many studies on this topic have focused on understanding how gustatory neural circuits are spatially organized to represent information about taste quality (e.g., "sweet", "salty", "bitter", etc.). Arguments pertaining to this issue have largely centered on whether taste is carried by dedicated neural channels or a pattern of activity across a neural population. But there is now mounting evidence that the timing of neural events may also importantly contribute to the representation of taste. In this review, we attempt to summarize recent findings in the field that pertain to these issues. Both space and time are variables likely related to the mechanism of the gustatory neural code: information about taste appears to reside in spatial and temporal patterns of activation in gustatory neurons. What is more, the organization of the taste network in the brain would suggest that the parameters of space and time extend to the neural processing of gustatory information on a much grander scale.     

Physiological responses to the pulsation threshold paradigm. I: Pulsation threshold patterns do not reproduce physiological rate profiles of high-pass and low-pass noise maskers

This article compares psychophysical measures of human processing of acoustic stimuli with one neurophysiological representation (normalized discharge rate profiles) of those stimuli. Psychophysical pulsation threshold patterns (PTPs) were derived for high-pass and low-pass noise maskers. Spectral features of both maskers are clearly evident in the PTPs. However, while the representation of high-pass noise in the PTPs becomes sharper with increasing masker level, the representation of low-pass noise degenerates as masker level is increased. One assumption that has been used previously to interpret pulsation threshold data is that PTPs reflect the profile of activity in primary neural elements in response to the masking stimulus. To investigate this hypothesis, normalized-rate profiles of responses to both maskers were derived from populations of auditory-nerve fibers in cats. Normalized-rate profiles do not exhibit the same behavior as PTPs for high-pass noise maskers in that the neural representation of the band edge degenerates as sound level increases. Furthermore, the distinction between the passband and the stop band is lost in the neural rate profiles, whereas the distinction improves in the high-pass noise PTPs.     

Functional integral mean field expansions for nuclear many fermion systems

We present a careful analysis of the auxiliary field functional integral formalism for many fermion systems. We examine the limiting procedure used in construction of such integrals and show that a wide flexibility exists with respect to the choice of the one-body field representation upon which mean field expansions are made. We demonstrate the utility of this flexibility in the context of the evaluation of the grand canonical partition function. We examine the zero order. RPA and certain higher-order terms. The above-mentioned flexibility is reflected in the dependence of the results on a trial two-body interaction, different choices of which produce Hartree, Fock, HartreeFock or other forms of the mean field expansions. A standard variational procedure selects the Hartree-Fock as the optimal choice. With this choice we find certain corrections to previously reported RPA contribution for the Hartree mean field. We also indicate the relevance of our formulation for the recent applications of the functional integral mean field approach to nuclear dynamical problems.     

Autocalibrated parallel imaging reconstruction with sampling pattern optimization for GRASE: APIR4GRASE

PurposeTo reduce artifacts and scan time of GRASE imaging by selecting an optimal sampling pattern and jointly reconstructing gradient echo and spin echo images.MethodsWe jointly reconstruct images for the different echo types by considering these as additional virtual coil channels in the novel Autocalibrated Parallel Imaging Reconstruction with Sampling Pattern Optimization for GRASE (APIR4GRASE) method. Besides image reconstruction, we identify optimal sampling patterns for the acquisition. The selected optimal patterns were validated on phantom and in-vivo acquisitions. Comparison to the conventional GRASE without acceleration, and to the GRAPPA reconstruction with a single echo type was also performed.ResultsUsing identified optimal sampling patterns, APIR4GRASE eliminated modulation artifacts in both phantom and in-vivo experiments; mean square error (MSE) was reduced by 78% and 94%, respectively, compared to the conventional GRASE with similar scan time. Both artifacts and g-factor were reduced compared to the GRAPPA reconstruction with a single echo type.ConclusionAPIR4GRASE substantially improves the speed and quality of GRASE imaging over the state-of-the-art, and is able to reconstruct both spin echo and gradient echo images.     

Roughness sub-layer investigation of a turbulent boundary layer

Particle image velocimetry (PIV) measurements at relatively low Reynolds numbers in a turbulent boundary layer over a three-dimensional roughened surface, consisting of pyramidal rows, have been presented. Measurements have been taken in a streamwise wall-normal plane intercepting the apex of a row of pyramids and the diagonal of the square base. The results shown in this paper point out the non-homogeneity of the flow in the roughness sub-layer. The different flow behaviour along the ascendant and the descendent part of the pyramids and in the region between two consecutive pyramids has been visualised. Low values of the streamwise component of the mean velocity and high values of the streamwise and wall-normal component of the Reynolds normal stresses, turbulent kinetic energy and Reynolds shear stresses are present in the downstream part of the pyramids, near their base. The mean representation of the flow shows swirling patterns in correspondence of the top of the pyramids. It is suggested that this representation is produced mainly by vortical structures travelling along the wall, retarded and intensified when interacting with the flow around the roughness element and by swirling patterns originating from the interaction of the incoming flow with the pyramids. In order to understand the origin of the high turbulence activity observed downstream the pyramids, near their base, a conditional analysis based on the quadrant method has been performed. It has been shown that sweep events are the major contributor to the very high values of the Reynolds shear stresses observed in this region and are due to the combination of the vortical flow at the top of the pyramid and a large scale (with respect to the roughness element) inflow motion.