Spike Train Patterns Research Articles

Effects of nonlinear membrane capacitance in the Hodgkin-Huxley model of action potential on the spike train patterns of a single neuron

The membrane capacitance has been shown to have a nonlinear dependence on the cell membrane potential in various types of cells. But the role of the nonlinear membrane capacitance in neurons has not been studied in detail. Herein, by considering the membrane capacitance to be a nonlinear parameter, we have explored the behavior of the cell membrane in three different types of neurons, i.e., squid giant neuron, rodent hippocampal interneuron, and rodent cortical neuron. The Hodgkin-Huxley equation of action potential was modified accordingly and simulated computationally. Our simulated results suggest that the action potential amplitude of a neuron almost remains the same for some duration when the voltage dependence parameter of the nonlinear capacitance increases up to a certain range, the initiation of the next action potential is delayed and the reduction in spike frequencies occurs in comparison to constant membrane capacitance. This indicates the importance of nonlinearity in membrane capacitance. Simultaneously the inter-spike interval (ISI) changes with the nonlinear membrane capacitance parameter. The gating dynamics show changes mainly in the activation current while the membrane capacitance is considered to be nonlinear. The above-mentioned computational results are primarily predictive pending experimental verification.

Capturing spike train temporal pattern with wavelet average coefficient for brain machine interface

Motor brain machine interfaces (BMIs) directly link the brain to artificial actuators and have the potential to mitigate severe body paralysis caused by neurological injury or disease. Most BMI systems involve a decoder that analyzes neural spike counts to infer movement intent. However, many classical BMI decoders (1) fail to take advantage of temporal patterns of spike trains, possibly over long time horizons; (2) are insufficient to achieve good BMI performance at high temporal resolution, as the underlying Gaussian assumption of decoders based on spike counts is violated. Here, we propose a new statistical feature that represents temporal patterns or temporal codes of spike events with richer description—wavelet average coefficients (WAC)—to be used as decoder input instead of spike counts. We constructed a wavelet decoder framework by using WAC features with a sliding-window approach, and compared the resulting decoder against classical decoders (Wiener and Kalman family) and new deep learning based decoders ( Long Short-Term Memory) using spike count features. We found that the sliding-window approach boosts decoding temporal resolution, and using WAC features significantly improves decoding performance over using spike count features.

Binless Kernel Machine: Modeling Spike Train Transformation for Cognitive Neural Prostheses.

Modeling spike train transformation among brain regions helps in designing a cognitive neural prosthesis that restores lost cognitive functions. Various methods analyze the nonlinear dynamic spike train transformation between two cortical areas with low computational eficiency. The application of a real-time neural prosthesis requires computational eficiency, performance stability, and better interpretation of the neural firing patterns that modulate target spike generation. We propose the binless kernel machine in the point-process framework to describe nonlinear dynamic spike train transformations. Our approach embeds the binless kernel to eficiently capture the feedforward dynamics of spike trains and maps the input spike timings into reproducing kernel Hilbert space (RKHS). An inhomogeneous Bernoulli process is designed to combine with a kernel logistic regression that operates on the binless kernel to generate an output spike train as a point process. Weights of the proposed model are estimated by maximizing the log likelihood of output spike trains in RKHS, which allows a global-optimal solution. To reduce computational complexity, we design a streaming-based clustering algorithm to extract typical and important spike train features. The cluster centers and their weights enable the visualization of the important input spike train patterns that motivate or inhibit output neuron firing. We test the proposed model on both synthetic data and real spike train data recorded from the dorsal premotor cortex and the primary motor cortex of a monkey performing a center-out task. Performances are evaluated by discrete-time rescaling Kolmogorov-Smirnov tests. Our model outperforms the existing methods with higher stability regardless of weight initialization and demonstrates higher eficiency in analyzing neural patterns from spike timing with less historical input (50%). Meanwhile, the typical spike train patterns selected according to weights are validated to encode output spike from the spike train of single-input neuron and the interaction of two input neurons.

Energy-efficient firing patterns with sparse bursts in the Chay neuron model

The energy efficiency of neural signal transmission is thought to be an important constraint in the nervous system. It is generally measured as the energy consumed per unit of information. Most of the previous studies have demonstrated this efficiency by focusing on single action potentials. However, neural information is more likely to be encoded by a spike train rather than by a single spike. To date, how the energy efficiency is dependent on patterns of spike trains is still unclear. In this study, we examined the energy efficiency of various firing patterns simulated by the Chay neuron model, including relatively high-frequency activities with massive spikes, medium-frequency activities with a moderate number of spikes, and low-frequency activities with rare spikes. Our results indicate that medium-frequency patterns are more energy efficient than both the high-frequency and low-frequency patterns. The most efficient medium-frequency pattern is a sparse burst firing (SBF) pattern because it consumes minimal energy and transmits an amount of neural information comparable to that of high-frequency patterns which consume much more energy. SBF patterns minimize energy consumption not only by producing fewer spikes than high-frequency patterns, but more importantly, also by consuming available energy sources, namely the potential energy stored in ionic concentration gradients, in a balanced way. Furthermore, with fewer spikes, the irregular spike trains of SBF patterns with short bursts and single spikes maximize the neural information that they carry, leading to higher energy efficiency. Thus, the sensory system may give priority to limiting energy costs over maximizing information to achieve greater energy efficiency.

Modeling fMRI BOLD signals and temporal mismatches in the cerebellar cortex

To understand brain activity relating neurons to circuits to learning and behavior, we explored a bottom-up computational reconstruction of population signals arising from cerebellum granular layer. As a first implementation, using bio-realistic computational models of cerebellum granule cell, in vivo spike train patterns were computed and then translated into functional Magnetic Resonance Imaging, Blood Oxygen-Level Dependent (BOLD) signals. The BOLD response was generated from averaged activity arising from center-surround organization modeled by using excitatory-inhibitory ratios related to experimental data. The averaged responses were converted to BOLD signals using the balloon and modified Windkessel models. Although both models generated BOLD responses corresponding to neural activity, the temporal mismatch was attributed to the response by the delayed compliance parameter in the Windkessel model. The modeling suggests that experimental variability observed in the cerebellar micro-zones could be related to compliance chances, activation patterns and number of neurons. Although detailed neuro-vasculature information was not modeled, the advantage in this methodology is that cerebellar cortex may allow seemingly linear transformations of underlying spiking that could be then used to validate network reconstructions.

A new approach for estimation of spiketrain patterns in basal ganglia.

The pathophysiological model of dystonia proposed that in addition to reduced firing rate in the internal pallidum, changes in the pattern may also play a role in disease manifestation. While common methods for patterns separation depends on arbitrary spiketrain parameters, we considered the new method for neural patterns based on spike density histograms and hierarchical clustering of real datasets. We used the single unit activity recordings from the globus pallidus external (GPe) and the globus pallidus internal (GPi) from 10 cervical dystonia (CD), 7 segmental dystonia (SD) and 8 generalized dystonia (GD) patients undergoing deep brain stimulation surgery. Using novel method, we separated three patterns of activity: burst, burst-like and tonic. Using this separation, we revealed the differences both in firing rate and pattern distribution between dystonia patients. We have shown the suitability of the proposed method for pattern clusterization on real data and assume that further application of this method would facilitate more detailed study of the neural activity in the basal ganglia and the search for neurophysiological biomarkers of movement disorders.

Spike-based encoding and learning of spectrum features for robust sound recognition

Abstract Biological evidence suggests that local time-frequency (LTF) information can be utilized to improve the recognition rate of sounds in the presence of noise. However, most of conventional methods use stationary (frequency-based) features which are not robust to noise, as each stationary feature contains a mixture of spectral information from both noise and signal. This paper proposes a spike-timing based model to encode and learn the LTF features extracted from sound spectrogram using spiking neural networks (SNNs), named LTF-SNN. In this model, we encode the reliable LTF features into spike train patterns and train with different spike-based learning rules. We analyze the efficacy of the spike-based feature encoding method and the recognition performance of the model by using two classes of SNN learning algorithms: ReSuMe and Tempotron. Utilizing the temporal coding and learning, networks of spiking neurons can effectively perform robust sound recognition tasks. Experimental results demonstrate that the model achieves superior performance in mismatched conditions compared with benchmark approaches.

Robust spike-train learning in spike-event based weight update

Supervised learning algorithms in a spiking neural network either learn a spike-train pattern for a single neuron receiving input spike-train from multiple input synapses or learn to output the first spike time in a feedforward network setting. In this paper, we build upon spike-event based weight update strategy to learn continuous spike-train in a spiking neural network with a hidden layer using a dead zone on–off based adaptive learning rate rule which ensures convergence of the learning process in the sense of weight convergence and robustness of the learning process to external disturbances. Based on different benchmark problems, we compare this new method with other relevant spike-train learning algorithms. The results show that the speed of learning is much improved and the rate of successful learning is also greatly improved.

Analysis of temporal firing patterns of primary afferent C-fibers for different sensations in mice

Some people with amputated limbs can benefit from neural prosthetics to restore tactile sensation through electrical stimulation of the afferent nerve. The temporal spike train pattern generated in healthy subject’s nerve by various types of somatosensation could provide key information to closely mimic natural sensations using electrical stimulation. However, the temporal firing patterns of peripheral sensory fibers have not been well understood yet. To interpret somatosensory spike trains, we performed ex vivo singlefiber recordings from the saphenous nerve in isolated skin-nerve preparations from mice. Some mechanically sensitive primary afferent C-fibers could also be activated by hot, cold, and itching stimuli, and we observed stimulus-specific firing patterns. These temporal patterns of the C-fibers for chemical stimuli were analyzed using a computational model based on quadruplets of spikes, which we classified into three groups of responses, i.e., capsaicin (hot), allyl-isothiocyanate (cold), and α-methyl-serotonin (itching). Each group of responses to the chemical stimuli was different from that evoked by mechanical stimuli. Therefore, these findings indicate that nontactile somatosensation can be decoded and used as input to a computerized system. Our quadruplet approach to the temporal patterns of spike trains contributes valuable insight to the identification of temporal profiles of other biological conditions.

Spike Timing Matters in Novel Neuronal Code Involved in Vibrotactile Frequency Perception

Skin vibrations sensed by tactile receptors contribute significantly to the perception of object properties during tactile exploration [1-4] and to sensorimotor control during object manipulation [5]. Sustained low-frequency skin vibration (<60Hz) evokes a distinct tactile sensation referred to as flutter whose frequency can be clearly perceived [6]. How afferent spiking activity translates into the perception of frequency is still unknown. Measures based on mean spike rates of neurons in the primary somatosensory cortex are sufficient to explain performance in some frequency discrimination tasks [7-11]; however, thereis emerging evidence that stimuli can be distinguished based also on temporal features of neural activity [12, 13]. Our study's advance is to demonstrate that temporal features are fundamental for vibrotactile frequency perception. Pulsatile mechanical stimuli were used to elicit specified temporal spike train patterns in tactile afferents, and subsequently psychophysical methods were employed to characterize human frequency perception. Remarkably, the most salient temporal feature determining vibrotactile frequency was not the underlying periodicity but, rather, the duration of the silent gap between successive bursts of neural activity. This burst gap code for frequency represents a previously unknown form ofneural coding in the tactile sensory system, whichparallels auditory pitch perception mechanisms based on purely temporal information where longer inter-pulse intervals receive higher perceptual weights than short intervals [14]. Our study also demonstrates that human perception of stimuli can be determined exclusively by temporal features of spike trains independent of the mean spike rate and without contribution from population response factors.

A spiking network that learns to extract spike signatures from speech signals

Spiking neural networks (SNNs) with adaptive synapses reflect core properties of biological neural networks. Speech recognition, as an application involving audio coding and dynamic learning, provides a good test problem to study SNN functionality. We present a simple, novel, and efficient nonrecurrent SNN that learns to convert a speech signal into a spike train signature. The signature is distinguishable from signatures for other speech signals representing different words, thereby enabling digit recognition and discrimination in devices that use only spiking neurons. The method uses a small, nonrecurrent SNN consisting of Izhikevich neurons equipped with spike timing dependent plasticity (STDP) and biologically realistic synapses. This approach introduces an efficient and fast network without error-feedback training, although it does require supervised training. The new simulation results produce discriminative spike train patterns for spoken digits in which highly correlated spike trains belong to the same category and low correlated patterns belong to different categories. The proposed SNN is evaluated using a spoken digit recognition task where a subset of the Aurora speech dataset is used. The experimental results show that the network performs well in terms of accuracy rate and complexity.

EP 63. Spectral analysis and visualization of multi-unit activity in subthalamic nucleus in Parkinson’s as a tool for automated electrophysiological classification of basal ganglia structures during deep brain stimulation procedures

Introduction In deep brain stimulation (DBS) of the subthalamic nucleus (STN) intraoperative microelectrode recordings (MER) are a well-established method for electrophysiological stereotactic target identification. Despite the main electrophysiological properties of STN are already well described ( [Cagnan et al., 2011] , [Seifried et al., 2012] ), mapping of STN rely on visual and acoustic inspection of raw recordings of MER trajectories. In this study we analyzed spectral decomposition and spike train patterns retrospectively of MER in STN-DBS to proposing a simple and reliable electrophysiological classification system. Secondly we suggest visualization techniques to present electrophysiological results in a user-friendly way. Patients and methods We analyzed MER in 7 patients with bilateral STN-DBS including 26 MER trajectories retrospectively. All recordings were carried out on multiple trajectories 10 mm prior and 4–5 mm below MR-tomographic STN target along planed DBS lead trajectory. Data was recorded intraoperative by Medtronic Leadpoint system using sharp microTargeting electrodes (FHC, USA). Offline Multi-Unit activity analysis was realized with Fieldtrip-Toolbox and Matlab, Mathworks. Statistical analysis of power spectral density, noise level and spike interstimulus interval were obtained through non-parametric student’s t-test. In purpose of visualization we integrated the stereotactic surgical plan to the electrophysiological results. Results Spectral density analysis showed a strong increase in frequency bands of 300–1000 Hz within STN borders. Spike train patterns within STN were heterogeneously and differed in firing rate and interstimulus interval to the key anatomical structures, thalamus and substantia nigra. Conclusion Through combination of spectral density analysis and spike train analysis we propose a simple and reliable MER signal classification method, which can be used for automated electrophysiological mapping of STN. Three-dimensional visualization of mapping results can provide help for decision of final DBS electrode implantation site.

Temporal pattern of online communication spike trains in spreading a scientific rumor: how often, who interacts with whom?

We study complex time series (spike trains) of online user communication while spreading messages about the discovery of the Higgs boson in Twitter. We focus on online social interactions among users such as retweet, mention, and reply, and construct different types of active (performing an action) and passive (receiving an action) spike trains for each user. The spike trains are analyzed by means of local variation, to quantify the temporal behavior of active and passive users, as a function of their activity and popularity. We show that the active spike trains are bursty, independently of their activation frequency. For passive spike trains, in contrast, the local variation of popular users presents uncorrelated (Poisson random) dynamics. We further characterize the correlations of the local variation in different interactions. We obtain high values of correlation, and thus consistent temporal behavior, between retweets and mentions, but only for popular users, indicating that creating online attention suggests an alignment in the dynamics of the two interactions.

Distinguishing cognitive state with multifractal complexity of hippocampal interspike interval sequences.

Fractality, represented as self-similar repeating patterns, is ubiquitous in nature and the brain. Dynamic patterns of hippocampal spike trains are known to exhibit multifractal properties during working memory processing; however, it is unclear whether the multifractal properties inherent to hippocampal spike trains reflect active cognitive processing. To examine this possibility, hippocampal neuronal ensembles were recorded from rats before, during and after a spatial working memory task following administration of tetrahydrocannabinol (THC), a memory-impairing component of cannabis. Multifractal detrended fluctuation analysis was performed on hippocampal interspike interval sequences to determine characteristics of monofractal long-range temporal correlations (LRTCs), quantified by the Hurst exponent, and the degree/magnitude of multifractal complexity, quantified by the width of the singularity spectrum. Our results demonstrate that multifractal firing patterns of hippocampal spike trains are a marker of functional memory processing, as they are more complex during the working memory task and significantly reduced following administration of memory impairing THC doses. Conversely, LRTCs are largest during resting state recordings, therefore reflecting different information compared to multifractality. In order to deepen conceptual understanding of multifractal complexity and LRTCs, these measures were compared to classical methods using hippocampal frequency content and firing variability measures. These results showed that LRTCs, multifractality, and theta rhythm represent independent processes, while delta rhythm correlated with multifractality. Taken together, these results provide a novel perspective on memory function by demonstrating that the multifractal nature of spike trains reflects hippocampal microcircuit activity that can be used to detect and quantify cognitive, physiological, and pathological states.

Extracting information in spike time patterns with wavelets and information theory.

We present a new method to assess the information carried by temporal patterns in spike trains. The method first performs a wavelet decomposition of the spike trains, then uses Shannon information to select a subset of coefficients carrying information, and finally assesses timing information in terms of decoding performance: the ability to identify the presented stimuli from spike train patterns. We show that the method allows: 1) a robust assessment of the information carried by spike time patterns even when this is distributed across multiple time scales and time points; 2) an effective denoising of the raster plots that improves the estimate of stimulus tuning of spike trains; and 3) an assessment of the information carried by temporally coordinated spikes across neurons. Using simulated data, we demonstrate that the Wavelet-Information (WI) method performs better and is more robust to spike time-jitter, background noise, and sample size than well-established approaches, such as principal component analysis, direct estimates of information from digitized spike trains, or a metric-based method. Furthermore, when applied to real spike trains from monkey auditory cortex and from rat barrel cortex, the WI method allows extracting larger amounts of spike timing information. Importantly, the fact that the WI method incorporates multiple time scales makes it robust to the choice of partly arbitrary parameters such as temporal resolution, response window length, number of response features considered, and the number of available trials. These results highlight the potential of the proposed method for accurate and objective assessments of how spike timing encodes information.

A fast BK-type KCa current acts as a postsynaptic modulator of temporal selectivity for communication signals.

Temporal patterns of spiking often convey behaviorally relevant information. Various synaptic mechanisms and intrinsic membrane properties can influence neuronal selectivity to temporal patterns of input. However, little is known about how synaptic mechanisms and intrinsic properties together determine the temporal selectivity of neuronal output. We tackled this question by recording from midbrain electrosensory neurons in mormyrid fish, in which the processing of temporal intervals between communication signals can be studied in a reduced in vitro preparation. Mormyrids communicate by varying interpulse intervals (IPIs) between electric pulses. Within the midbrain posterior exterolateral nucleus (ELp), the temporal patterns of afferent spike trains are filtered to establish single-neuron IPI tuning. We performed whole-cell recording from ELp neurons in a whole-brain preparation and examined the relationship between intrinsic excitability and IPI tuning. We found that spike frequency adaptation of ELp neurons was highly variable. Postsynaptic potentials (PSPs) of strongly adapting (phasic) neurons were more sharply tuned to IPIs than weakly adapting (tonic) neurons. Further, the synaptic filtering of IPIs by tonic neurons was more faithfully converted into variation in spiking output, particularly at short IPIs. Pharmacological manipulation under current- and voltage-clamp revealed that tonic firing is mediated by a fast, large-conductance Ca2+-activated K+ (KCa) current (BK) that speeds up action potential repolarization. These results suggest that BK currents can shape the temporal filtering of sensory inputs by modifying both synaptic responses and PSP-to-spike conversion. Slow SK-type KCa currents have previously been implicated in temporal processing. Thus, both fast and slow KCa currents can fine-tune temporal selectivity.

Extraction and restoration of hippocampal spatial memories with non-linear dynamical modeling.

To build a cognitive prosthesis that can replace the memory function of the hippocampus, it is essential to model the input-output function of the damaged hippocampal region, so the prosthetic device can stimulate the downstream hippocampal region, e.g., CA1, with the output signal, e.g., CA1 spike trains, predicted from the ongoing input signal, e.g., CA3 spike trains, and the identified input-output function, e.g., CA3-CA1 model. In order for the downstream region to form appropriate long-term memories based on the restored output signal, furthermore, the output signal should contain sufficient information about the memories that the animal has formed. In this study, we verify this premise by applying regression and classification modelings of the spatio-temporal patterns of spike trains to the hippocampal CA3 and CA1 data recorded from rats performing a memory-dependent delayed non-match-to-sample (DNMS) task. The regression model is essentially the multiple-input, multiple-output (MIMO) non-linear dynamical model of spike train transformation. It predicts the output spike trains based on the input spike trains and thus restores the output signal. In addition, the classification model interprets the signal by relating the spatio-temporal patterns to the memory events. We have found that: (1) both hippocampal CA3 and CA1 spike trains contain sufficient information for predicting the locations of the sample responses (i.e., left and right memories) during the DNMS task; and more importantly (2) the CA1 spike trains predicted from the CA3 spike trains by the MIMO model also are sufficient for predicting the locations on a single-trial basis. These results show quantitatively that, with a moderate number of unitary recordings from the hippocampus, the MIMO non-linear dynamical model is able to extract and restore spatial memory information for the formation of long-term memories and thus can serve as the computational basis of the hippocampal memory prosthesis.

Discriminative and affective touch: sensing and feeling.

The multimodal properties of the human somatosensory system continue to be unravelled. There is mounting evidence that one of these submodalities-touch-has another dimension, providing not only its well-recognized discriminative input to the brain, but also an affective input. It has long been recognized that touch plays an important role in many forms of social communication and a number of theories have been proposed to explain observations and beliefs about the "power of touch." Here, we propose that a class of low-threshold mechanosensitive C fibers that innervate the hairy skin represent the neurobiological substrate for the affective and rewarding properties of touch.

Stochastic variational learning in recurrent spiking networks.

The ability to learn and perform statistical inference with biologically plausible recurrent networks of spiking neurons is an important step toward understanding perception and reasoning. Here we derive and investigate a new learning rule for recurrent spiking networks with hidden neurons, combining principles from variational learning and reinforcement learning. Our network defines a generative model over spike train histories and the derived learning rule has the form of a local Spike Timing Dependent Plasticity rule modulated by global factors (neuromodulators) conveying information about “novelty” on a statistically rigorous ground. Simulations show that our model is able to learn both stationary and non-stationary patterns of spike trains. We also propose one experiment that could potentially be performed with animals in order to test the dynamics of the predicted novelty signal.

Effect of Alzheimer's disease on the dynamical and computational characteristics of recurrent neural networks

Recurrent circuits of simple model neurons can provide the substrate for cognitive functions such as perception, memory, association, classification or prediction of dynamical systems [1-3]. In Alzheimer's disease (AD), the impairment of such functions is clearly correlated to synapse loss [4]. So far, the mechanisms underlying this correlation are only poorly understood. Here, we investigate how the loss of excitatory synapses in sparsely connected random networks of spiking excitatory and inhibitory neurons [5] alters their dynamical and computational characteristics. By means of simulations, we study the network response to noisy variations of multidimensional spike-train patterns. We find that the loss of excitatory synapses on excitatory neurons (decrease in excitatory-excitatory indegree; vertical arrow in Figure ​Figure1)1) lowers the network's sensitivity to small perturbations of time-varying inputs, reduces its ability to discriminate and improves its generalization capability [6]. Figure 1 Loss of excitatory-excitatory synapses (vertical arrow) impairs discrimination capability (gray coded). Recovery of discrimination capability by firing-rate homeostasis (scaling up remaining excitatory-excitatory synapses; horizontal arrow). A full recovery of the network performance can be achieved by firing-rate homeostasis, implemented by scaling up the remaining excitatory-excitatory synapses (horizontal arrow in Figure ​Figure1).1). Homeostasis may therefore explain the absence of clinical symptoms in early AD, despite cortical damage. The onset of clinical symptoms may result from an exhaustion of homeostatic resources.