Noise-modulated multistable synapses in a Wilson-Cowan-based model of plasticity.

Frequency-dependent plasticity refers to changes in synaptic strength in response to different stimulation frequencies. Resonance is a factor known to be of importance in such frequency dependence, however, the role of neural noise in the process remains elusive. Considering the brain is an inherently noisy system, understanding its effects may prove beneficial in shaping therapeutic interventions based on non-invasive brain stimulation protocols. The Wilson-Cowan (WC) model is a well-established model to describe the average dynamics of neural populations and has been shown to exhibit bistability in the presence of noise. However, the important question of how the different stable regimes in the WC model can affect synaptic plasticity when cortical populations interact has not yet been addressed. Therefore, we investigated plasticity dynamics in a WC-based model of interacting neural populations coupled with activity-dependent synapses in which a periodic stimulation was applied in the presence of noise of controlled intensity. The results indicate that for a narrow range of the noise variance, synaptic strength can be optimized. In particular, there is a regime of noise intensity for which synaptic strength presents a triple-stable state. Regulating noise intensity affects the probability that the system chooses one of the stable states, thereby controlling plasticity. These results suggest that noise is a highly influential factor in determining the outcome of plasticity induced by stimulation.

High-frequency oscillations in the ripple bands and amplitude information coding: Toward a biomarker of maximum entropy in the preictal signals.

Intracranial electroencephalography (iEEG) can directly record local field potentials (LFPs) from a large set of neurons in the vicinity of the electrode. To search for possible epileptic biomarkers and to determine the epileptogenic zone that gives rise to seizures, we investigated the dynamics of basal and preictal signals. For this purpose, we explored the dynamics of the recorded time series for different frequency bands considering high-frequency oscillations (HFO) up to 240 Hz. We apply a Hilbert transform to study the amplitude and phase of the signals. The dynamics of the different frequency bands in the time causal entropy-complexity plane, H × C, is characterized by comparing the dynamical evolution of the basal and preictal time series. As the preictal states evolve closer to the time in which the epileptic seizure starts, the, H × C, dynamics changes for the higher frequency bands. The complexity evolves to very low values and the entropy becomes nearer to its maximal value. These quasi-stable states converge to equiprobable states when the entropy is maximal, and the complexity is zero. We could, therefore, speculate that in this case, it corresponds to the minimization of Gibbs free energy. In this case, the maximum entropy is equivalent to the principle of minimum consumption of resources in the system. We can interpret this as the nature of the system evolving temporally in the preictal state in such a way that the consumption of resources by the system is minimal for the amplitude in frequencies between 220-230 and 230-240 Hz.

GABA Modulates Frequency-Dependent Plasticity in Humans.

Frequency-dependent reorganization of the primary somatosensory cortex, together with perceptual changes, arises following repetitive sensory stimulation. Here, we investigate the role of GABA in this process. We co-stimulated two finger tips and measured GABA and Glx using magnetic resonance (MR) spectroscopy at the beginning and end of the stimulation. Participants performed a perceptual learning task before and after stimulation. There were 2 sessions with stimulation frequency either at or above the resonance frequency of the primary somatosensory cortex (23 and 39 Hz, respectively). Perceptual learning occurred following above resonance stimulation only, while GABA reduced during this condition. Lower levels of early GABA were associated with greater perceptual learning. One possible mechanism underlying this finding is that cortical disinhibition "unmasks" lateral connections within the cortex to permit adaptation to the sensory environment. These results provide evidence in humans for a frequency-dependent inhibitory mechanism underlying learning and suggest a mechanism-based approach for optimizing neurostimulation frequency.

Pattern Separation Underpins Expectation-Modulated Memory.

Pattern separation and completion are fundamental hippocampal computations supporting memory encoding and retrieval. However, despite extensive exploration of these processes, it remains unclear whether and how top-down processes adaptively modulate the dynamics between these computations. Here we examine the role of expectation in shifting the hippocampus to perform pattern separation. In a behavioral task, 29 participants (7 males) learned a cue-object category contingency. Then, at encoding, one-third of the cues preceding the to-be-memorized objects, violated the studied rule. At test, participants performed a recognition task with old objects (targets) and a set of parametrically manipulated (very similar to dissimilar) foils for each object. Accuracy was found to be better for foils of high similarity to targets that were contextually unexpected at encoding compared with expected ones. Critically, there were no expectation-driven differences for targets and low similarity foils. To further explore these effects, we implemented a computational model of the hippocampus, performing the same task as the human participants. We used representational similarity analysis to examine how top-down expectation interacts with bottom-up perceptual input, in each layer. All subfields showed more dissimilar representations for unexpected items, with dentate gyrus (DG) and CA3 being more sensitive to expectation violation than CA1. Again, representational differences between expected and unexpected inputs were prominent for moderate to high levels of input similarity. This effect diminished when inputs from DG and CA3 into CA1 were lesioned. Overall, these novel findings strongly suggest that pattern separation in DG/CA3 underlies the effect that violation of expectation exerts on memory.SIGNIFICANCE STATEMENT What makes some events more memorable than others is a key question in cognitive neuroscience. Violation of expectation often leads to better memory performance, but the neural mechanism underlying this benefit remains elusive. In a behavioral study, we found that memory accuracy is enhanced selectively for unexpected highly similar foils, suggesting expectation violation does not enhance memory indiscriminately, but specifically aids the disambiguation of overlapping inputs. This is further supported by our subsequent investigation using a hippocampal computational model, revealing increased representational dissimilarity for unexpected highly similar foils in DG and CA3. These convergent results provide the first evidence that pattern separation plays an explicit role in supporting memory for unexpected information.

Natural Language Statistical Features of LSTM-Generated Texts.

Long short-term memory (LSTM) networks have recently shown remarkable performance in several tasks that are dealing with natural language generation, such as image captioning or poetry composition. Yet, only few works have analyzed text generated by LSTMs in order to quantitatively evaluate to which extent such artificial texts resemble those generated by humans. We compared the statistical structure of LSTM-generated language to that of written natural language, and to those produced by Markov models of various orders. In particular, we characterized the statistical structure of language by assessing word-frequency statistics, long-range correlations, and entropy measures. Our main finding is that while both LSTM- and Markov-generated texts can exhibit features similar to real ones in their word-frequency statistics and entropy measures, LSTM-texts are shown to reproduce long-range correlations at scales comparable to those found in natural language. Moreover, for LSTM networks, a temperature-like parameter controlling the generation process shows an optimal value-for which the produced texts are closest to real language-consistent across different statistical features investigated.

Evidence for frequency-dependent cortical plasticity in the human brain.

Frequency-dependent plasticity (FDP) describes adaptation at the synapse in response to stimulation at different frequencies. Its consequence on the structure and function of cortical networks is unknown. We tested whether cortical "resonance," favorable stimulation frequencies at which the sensory cortices respond maximally, influenced the impact of FDP on perception, functional topography, and connectivity of the primary somatosensory cortex using psychophysics and functional imaging (fMRI). We costimulated two digits on the hand synchronously at, above, or below the resonance frequency of the somatosensory cortex, and tested subjects' accuracy and speed on tactile localization before and after costimulation. More errors and slower response times followed costimulation at above- or below-resonance, respectively. Response times were faster after at-resonance costimulation. In the fMRI, the cortical representations of the two digits costimulated above-resonance shifted closer, potentially accounting for the poorer performance. Costimulation at-resonance did not shift the digit regions, but increased the functional coupling between them, potentially accounting for the improved response time. To relate these results to synaptic plasticity, we simulated a network of oscillators incorporating Hebbian learning. Two neighboring patches embedded in a cortical sheet, mimicking the two digit regions, were costimulated at different frequencies. Network activation outside the stimulated patches was greatest at above-resonance frequencies, reproducing the spread of digit representations seen with fMRI. Connection strengths within the patches increased following at-resonance costimulation, reproducing the increased fMRI connectivity. We show that FDP extends to the cortical level and is influenced by cortical resonance.

Modulation of Fast Narrowband Oscillations in the Mouse Retina and dLGN According to Background Light Intensity.

Background light intensity (irradiance) substantially impacts the visual code in the early visual system at synaptic and single-neuron levels, but its influence on population activity is largely unexplored. We show that fast narrowband oscillations, an important feature of population activity, systematically increase in amplitude as a function of irradiance in both anesthetized and awake, freely moving mice and at the level of the retina and dorsal lateral geniculate nucleus (dLGN). Narrowband coherence increases with irradiance across large areas of the dLGN, but especially for neighboring units. The spectral sensitivity of these effects and their substantial reduction in melanopsin knockout animals indicate a contribution from inner retinal photoreceptors. At bright backgrounds, narrowband coherence allows pooling of single-unit responses to become a viable strategy for enhancing visual signals within its frequency range.

Cortical Resonance Frequencies Emerge from Network Size and Connectivity.

Neural oscillations occur within a wide frequency range with different brain regions exhibiting resonance-like characteristics at specific points in the spectrum. At the microscopic scale, single neurons possess intrinsic oscillatory properties, such that is not yet known whether cortical resonance is consequential to neural oscillations or an emergent property of the networks that interconnect them. Using a network model of loosely-coupled Wilson-Cowan oscillators to simulate a patch of cortical sheet, we demonstrate that the size of the activated network is inversely related to its resonance frequency. Further analysis of the parameter space indicated that the number of excitatory and inhibitory connections, as well as the average transmission delay between units, determined the resonance frequency. The model predicted that if an activated network within the visual cortex increased in size, the resonance frequency of the network would decrease. We tested this prediction experimentally using the steady-state visual evoked potential where we stimulated the visual cortex with different size stimuli at a range of driving frequencies. We demonstrate that the frequency corresponding to peak steady-state response inversely correlated with the size of the network. We conclude that although individual neurons possess resonance properties, oscillatory activity at the macroscopic level is strongly influenced by network interactions, and that the steady-state response can be used to investigate functional networks.

Bursting Neurons in the Hippocampal Formation Encode Features of LFP Rhythms.

Burst spike patterns are common in regions of the hippocampal formation such as the subiculum and medial entorhinal cortex (MEC). Neurons in these areas are immersed in extracellular electrical potential fluctuations often recorded as the local field potential (LFP). LFP rhythms within different frequency bands are linked to different behavioral states. For example, delta rhythms are often associated with slow-wave sleep, inactivity and anesthesia; whereas theta rhythms are prominent during awake exploratory behavior and REM sleep. Recent evidence suggests that bursting neurons in the hippocampal formation can encode LFP features. We explored this hypothesis using a two-compartment model of a bursting pyramidal neuron driven by time-varying input signals containing spectral peaks at either delta or theta rhythms. The model predicted a neural code in which bursts represented the instantaneous value, phase, slope and amplitude of the driving signal both in their timing and size (spike number). To verify whether this code is employed in vivo, we examined electrophysiological recordings from the subiculum of anesthetized rats and the MEC of a behaving rat containing prevalent delta or theta rhythms, respectively. In both areas, we found bursting cells that encoded information about the instantaneous voltage, phase, slope and/or amplitude of the dominant LFP rhythm with essentially the same neural code as the simulated neurons. A fraction of the cells encoded part of the information in burst size, in agreement with model predictions. These results provide in-vivo evidence that the output of bursting neurons in the mammalian brain is tuned to features of the LFP.

Thalamic neuron models encode stimulus information by burst-size modulation.

Thalamic neurons have been long assumed to fire in tonic mode during perceptive states, and in burst mode during sleep and unconsciousness. However, recent evidence suggests that bursts may also be relevant in the encoding of sensory information. Here, we explore the neural code of such thalamic bursts. In order to assess whether the burst code is generic or whether it depends on the detailed properties of each bursting neuron, we analyzed two neuron models incorporating different levels of biological detail. One of the models contained no information of the biophysical processes entailed in spike generation, and described neuron activity at a phenomenological level. The second model represented the evolution of the individual ionic conductances involved in spiking and bursting, and required a large number of parameters. We analyzed the models' input selectivity using reverse correlation methods and information theory. We found that n-spike bursts from both models transmit information by modulating their spike count in response to changes to instantaneous input features, such as slope, phase, amplitude, etc. The stimulus feature that is most efficiently encoded by bursts, however, need not coincide with one of such classical features. We therefore searched for the optimal feature among all those that could be expressed as a linear transformation of the time-dependent input current. We found that bursting neurons transmitted 6 times more information about such more general features. The relevant events in the stimulus were located in a time window spanning ~100 ms before and ~20 ms after burst onset. Most importantly, the neural code employed by the simple and the biologically realistic models was largely the same, implying that the simple thalamic neuron model contains the essential ingredients that account for the computational properties of the thalamic burst code. Thus, our results suggest the n-spike burst code is a general property of thalamic neurons.

Melanopsin-driven increases in maintained activity enhance thalamic visual response reliability across a simulated dawn.

Twice a day, at dawn and dusk, we experience gradual but very high amplitude changes in background light intensity (irradiance). Although we perceive the associated change in environmental brightness, the representation of such very slow alterations in irradiance by the early visual system has been little studied. Here, we addressed this deficit by recording electrophysiological activity in the mouse dorsal lateral geniculate nucleus under exposure to a simulated dawn. As irradiance increased we found a widespread enhancement in baseline firing that extended to units with ON as well as OFF responses to fast luminance increments. This change in baseline firing was equally apparent when the slow irradiance ramp appeared alone or when a variety of higher-frequency artificial or natural visual stimuli were superimposed upon it. Using a combination of conventional knockout, chemogenetic, and receptor-silent substitution manipulations, we continued to show that, over higher irradiances, this increase in firing originates with inner-retinal melanopsin photoreception. At the single-unit level, irradiance-dependent increases in baseline firing were strongly correlated with improvements in the amplitude of responses to higher-frequency visual stimuli. This in turn results in an up to threefold increase in single-trial reliability of fast visual responses. In this way, our data indicate that melanopsin drives a generalized increase in dorsal lateral geniculate nucleus excitability as dawn progresses that both conveys information about changing background light intensity and increases the signal:noise for fast visual responses.

Phase-locking of bursting neuronal firing to dominant LFP frequency components.

Neuronal firing in the hippocampal formation relative to the phase of local field potentials (LFP) has a key role in memory processing and spatial navigation. Firing can be in either tonic or burst mode. Although bursting neurons are common in the hippocampal formation, the characteristics of their locking to LFP phase are not completely understood. We investigated phase-locking properties of bursting neurons using simulations generated by a dual compartmental model of a pyramidal neuron adapted to match the bursting activity in the subiculum of a rat. The model was driven with stochastic input signals containing a power spectral profile consistent with physiologically relevant frequencies observed in LFP. The single spikes and spike bursts fired by the model were locked to a preferred phase of the predominant frequency band where there was a peak in the power of the driving signal. Moreover, the preferred phase of locking shifted with increasing burst size, providing evidence that LFP phase can be encoded by burst size. We also provide initial support for the model results by analysing example data of spontaneous LFP and spiking activity recorded from the subiculum of a single urethane-anaesthetised rat. Subicular neurons fired single spikes, two-spike bursts and larger bursts that locked to a preferred phase of either dominant slow oscillations or theta rhythms within the LFP, according to the model prediction. Both power-modulated phase-locking and gradual shift in the preferred phase of locking as a function of burst size suggest that neurons can use bursts to encode timing information contained in LFP phase into a spike-count code.

Melanopsin-driven light adaptation in mouse vision.

BACKGROUND: In bright light, mammals use a distinct photopigment (melanopsin) to measure irradiance for centrally mediated responses such as circadian entrainment. We aimed to determine whether the information generated by melanopsin is also used by the visual system as a signal for light adaptation. To this end, we compared retinal and thalamic responses to a range of artificial and natural visual stimuli presented using spectral compositions that either approximate the mouse's experience of natural daylight ("daylight") or are selectively depleted of wavelengths to which melanopsin is most sensitive ("mel-low"). RESULTS: We found reproducible and reversible changes in the flash electroretinogram between daylight and mel-low. Simultaneous recording in the dorsal lateral geniculate nucleus (dLGN) revealed that these reflect changes in feature selectivity of visual circuits in both temporal and spatial dimensions. A substantial fraction of units preferred finer spatial patterns in the daylight condition, while the population of direction-sensitive units became tuned to faster motion. The dLGN contained a richer, more reliable encoding of natural scenes in the daylight condition. These effects were absent in mice lacking melanopsin. CONCLUSIONS: The feature selectivity of many neurons in the mouse dLGN is adjusted according to a melanopsin-dependent measure of environmental brightness. These changes originate, at least in part, within the retina. Melanopsin performs a role analogous to a photographer's light meter, providing an independent measure of irradiance that determines optimal setting for visual circuits.

Quantifying the information in the long-range order of words: semantic structures and universal linguistic constraints.

We review some recent progress on the characterisation of long-range patterns of word use in language using methods from information theory. In particular, two levels of structure in language are considered. The first level corresponds to the patterns of words usage over different contextual domains. A direct application of information theory to quantify the specificity of words across different sections of a linguistic sequence leads to a measure of semantic information. Moreover, a natural scale emerges that characterises the typical size of semantic structures. Since the information measure is made up of additive contributions from individual words, it is possible to rank the words according to their overall weight in the total information. This allows the extraction of keywords most relevant to the semantic content of the sequence without any prior knowledge of the language. The second level considered is the complex structure of correlations among words in linguistic sequences. The degree of order in language can be quantified by means of the entropy. Reliable estimates of the entropy were obtained from corpora of texts from several linguistic families by means of lossless compression algorithms. The value of the entropy fluctuates across different languages since it depends on linguistic organisation at various levels. However, when a measure of relative entropy that specifically quantifies the degree of word ordering in language is estimated, it presents an almost constant value over all the linguistic families studied. This suggests that the entropy of word ordering is a novel quantitative linguistic universal.

Keywords and Co-Occurrence Patterns in the Voynich Manuscript: An Information-Theoretic Analysis.

The Voynich manuscript has remained so far as a mystery for linguists and cryptologists. While the text written on medieval parchment -using an unknown script system- shows basic statistical patterns that bear resemblance to those from real languages, there are features that suggested to some researches that the manuscript was a forgery intended as a hoax. Here we analyse the long-range structure of the manuscript using methods from information theory. We show that the Voynich manuscript presents a complex organization in the distribution of words that is compatible with those found in real language sequences. We are also able to extract some of the most significant semantic word-networks in the text. These results together with some previously known statistical features of the Voynich manuscript, give support to the presence of a genuine message inside the book.

Linking dynamical and functional properties of intrinsically bursting neurons.

Several studies have shown that bursting neurons can encode information in the number of spikes per burst: As the stimulus varies, so does the length of individual bursts. There presented stimuli, however, vary substantially among different sensory modalities and different neurons.The goal of this paper is to determine which kind of stimulus features can be encoded in burst length, and how those features depend on the mathematical properties of the underlying dynamical system.We show that the initiation and termination of each burst is triggered by specific stimulus features whose temporal characteristsics are determined by the types of bifurcations that initiate and terminate firing in each burst. As only a few bifurcations are possible, only a restricted number of encoded features exists. Here we focus specifically on describing parabolic, square-wave and elliptic bursters. We find that parabolic bursters, whose firing is initiated and terminated by saddle-node bifurcations, behave as prototypical integrators: Firing is triggered by depolarizing stimuli, and lasts for as long as excitation is prolonged. Elliptic bursters, contrastingly, constitute prototypical resonators, since both the initiating and terminating bifurcations possess well-defined oscillation time scales. Firing is therefore triggered by stimulus stretches of matching frequency and terminated by a phase-inversion in the oscillation. The behavior of square-wave bursters is somewhat intermediate, since they are triggered by a fold bifurcation of cycles of well-defined frequency but are terminated by a homoclinic bifurcation lacking an oscillating time scale. These correspondences show that stimulus selectivity is determined by the type of bifurcations. By testing several neuron models, we also demonstrate that additional biological properties that do not modify the bifurcation structure play a minor role in stimulus encoding. Moreover, we show that burst-length variability (and thereby, the capacity to transmit information) depends on a trade-off between the variance of the external signal driving the cell and the strength of the slow internal currents modulating bursts. Thus, our work explicitly links the computational properties of bursting neurons to the mathematical properties of the underlying dynamical systems.

Information coding in a laminar computational model of cat primary visual cortex.

Neural populations across cortical layers perform different computational tasks. However, it is not known whether information in different layers is encoded using a common neural code or whether it depends on the specific layer. Here we studied the laminar distribution of information in a large-scale computational model of cat primary visual cortex. We analyzed the amount of information about the input stimulus conveyed by the different representations of the cortical responses. In particular, we compared the information encoded in four possible neural codes: (1) the information carried by the firing rate of individual neurons; (2) the information carried by spike patterns within a time window; (3) the rate-and-phase information carried by the firing rate labelled by the phase of the Local Field Potentials (LFP); (4) the pattern-and-phase information carried by the spike patterns tagged with the LFP phase. We found that there is substantially more information in the rate-and-phase code compared with the firing rate alone for low LFP frequency bands (less than 30 Hz). When comparing how information is encoded across layers, we found that the extra information contained in a rate-and-phase code may reach 90 % in Layer 4, while in other layers it reaches only 60 %, compared to the information carried by the firing rate alone. These results suggest that information processing in primary sensory cortices could rely on different coding strategies across different layers.

Universal entropy of word ordering across linguistic families.

BACKGROUND: The language faculty is probably the most distinctive feature of our species, and endows us with a unique ability to exchange highly structured information. In written language, information is encoded by the concatenation of basic symbols under grammatical and semantic constraints. As is also the case in other natural information carriers, the resulting symbolic sequences show a delicate balance between order and disorder. That balance is determined by the interplay between the diversity of symbols and by their specific ordering in the sequences. Here we used entropy to quantify the contribution of different organizational levels to the overall statistical structure of language. METHODOLOGY/PRINCIPAL FINDINGS: We computed a relative entropy measure to quantify the degree of ordering in word sequences from languages belonging to several linguistic families. While a direct estimation of the overall entropy of language yielded values that varied for the different families considered, the relative entropy quantifying word ordering presented an almost constant value for all those families. CONCLUSIONS/SIGNIFICANCE: Our results indicate that despite the differences in the structure and vocabulary of the languages analyzed, the impact of word ordering in the structure of language is a statistical linguistic universal.

Conversion of phase information into a spike-count code by bursting neurons.

Single neurons in the cerebral cortex are immersed in a fluctuating electric field, the local field potential (LFP), which mainly originates from synchronous synaptic input into the local neural neighborhood. As shown by recent studies in visual and auditory cortices, the angular phase of the LFP at the time of spike generation adds significant extra information about the external world, beyond the one contained in the firing rate alone. However, no biologically plausible mechanism has yet been suggested that allows downstream neurons to infer the phase of the LFP at the soma of their pre-synaptic afferents. Therefore, so far there is no evidence that the nervous system can process phase information. Here we study a model of a bursting pyramidal neuron, driven by a time-dependent stimulus. We show that the number of spikes per burst varies systematically with the phase of the fluctuating input at the time of burst onset. The mapping between input phase and number of spikes per burst is a robust response feature for a broad range of stimulus statistics. Our results suggest that cortical bursting neurons could play a crucial role in translating LFP phase information into an easily decodable spike count code.

Spike-phase coding boosts and stabilizes information carried by spatial and temporal spike patterns.

Several neural codes have been proposed in order to explain how neurons encode sensory information. Here we tested the hypothesis that different codes might be employed concurrently and provide complementary stimulus information. Quantifying the information encoded about natural sounds in the auditory cortex of alert animals, we found that temporal spike-train patterns and spatial populations were both highly informative. However, the relative phase of slow ongoing rhythms at which these (temporal or population) responses occurred provided much additional and complementary information. Such nested codes combining spike-train patterns with the phase of firing were not only most informative, but also most robust to sensory noise added to the stimulus. Our findings suggest that processing in sensory cortices could rely on the concurrent use of several codes that combine information across different spatiotemporal scales. In addition, they propose a role of slow cortical rhythms in stabilizing sensory representations by reducing effects of noise.