Abstract Coherent activations of brain neuron networks underlie many physiological functions associated with various behavioral states. These synchronous fluctuations in the electrical activity are also referred to as rhythms. At cellular level, rhythmicity can be induced by mechanisms intrinsic oscillations neurons or network circulation excitation between synaptically coupled neurons. One specific mechanism concerns astrocytes that accompany and coherently modulate synaptic contacts...

The concept of a tripartite synapse holds that astrocytes can affect both the pre- and post-synaptic compartments through Ca(2+)-dependent release gliotransmitters. Because astrocytic Ca(2+) transients usually last for few seconds, we assumed regulation synaptic transmission may also occur on scale seconds. Here, considered basic physiological functions synapses investigated at level neural network activity. firing dynamics individual neurons in spontaneous was described by Hodgkin-Huxley...

The mathematical model of the spiking neural network (SNN) supplied by astrocytes is investigated. are a specific type brain cells which not electrically excitable but induce chemical modulations neuronal firing. We analyze how influence images encoded in form dynamic pattern SNN. Serving at much slower time scale, astrocytic interacting with neurons can remarkably enhance image representation quality. dynamics affected noise distorting information image. demonstrate that activation...

The goal of this study is to propose a new reduced phenomenological model that describes the mean-field dynamics arising from neuron–glial interaction, taking into account short-term synaptic plasticity and recurrent connections in presence astrocytic modulation connection. Using computer simulation numerical methods nonlinear dynamics, it shown proposed reproduces rich set patterns population activity, including spiking, bursting chaotic temporal patterns. These can coexist for specific...

We investigated a mathematical model composed of spiking neural network (SNN) interacting with astrocytes. analysed how information content in the form two-dimensional images can be represented by an SNN spatiotemporal pattern. The includes excitatory and inhibitory neurons some proportion, sustaining excitation–inhibition balance autonomous firing. astrocytes accompanying each synapse provide slow modulation synaptic transmission strength. An image was uploaded to stimulation pulses...

This paper investigates various bifurcation scenarios of the appearance bursting activity in phenomenological mean-field model neuron–glial interactions. In particular, we show that homoclinic spiral attractors this system can be source several types with different properties.

Molecules of the extracellular matrix (ECM) can modulate efficacy synaptic transmission and neuronal excitability. These mechanisms are crucial for homeostatic regulation firing over extended timescales. In this study, we introduce a simple mathematical model spiking balanced by influence ECM. We consider neuron receiving random input in form Poisson spike trains ECM, which is modeled phenomenological variable involved two feedback mechanisms. One mechanism scales values conductance to...

Experimental studies highlight the important role of extracellular matrix (ECM) in regulation neuronal excitability and synaptic connectivity nervous system. In its turn, neural ECM is formed an activity-dependent manner. Its maturation closes so-called critical period development, stabilizing efficient configurations networks brain. locally remodeled by proteases secreted activated manner into space this process for physiological plasticity. We ask if remodeling may be exaggerated under...

From the standpoint of neurodynamics, gener-ation bursts in populations neurons has been studied depth, but extrasynaptic mechanisms generation remain practically unstudied. We propose a computationally efficient tripartite synapse model that demonstrates formation burst dynamics neural network.

Domain adaptation is a popular paradigm in modern machine learning which aims at tackling the problem of divergence (or shift) between labeled training and validation datasets (source domain) potentially large unlabeled dataset (target domain). The task to embed both into common space source informative for while target minimized. most domain solutions are based on neural networks that combine classification adversarial modules, frequently making them data-hungry difficult train. We present...

We propose a mathematical model of spiking neural network (SNN) that interacts with an active extracellular field formed by the brain matrix (ECM). The SNN exhibits irregular dynamics induced constant noise drive. Following neurobiological facts, neuronal firing leads to production ECM occupies space. In turn, components can modulate signaling and synaptic transmission, for example, through effect so-called scaling. By simulating model, we discovered ECM-mediated regulation activity promotes...

The effect of glial cells (astrocytes) on the synaptic dynamics interneuronal contacts and network is studied. model a connections with impulse neurons. Neurons have been discovered in that detect spontaneous transitions to generation mode when it comes typical epileptiform activity brain neural networks.

This work is driven by a practical question: corrections of Artificial Intelligence (AI) errors. These should be quick and non-iterative. To solve this problem without modification legacy AI system, we propose special 'external' devices, correctors. Elementary correctors consist two parts, classifier that separates the situations with high risk error from in which system works well new decision recommended for potential Input signals can inputs its internal signals, outputs. If intrinsic...

We consider an unstructured neuron network model composed of excitatory and inhibitory neurons. The synaptic connections are supplied with spike timing-dependent plasticity (STDP). take the STDP implemented using a memristor. In normal conditions, forms so-called bursting discharges typical living networks in dissociated neuronal cultures. Incorporating biologically inspired model, we demonstrate how memristive emulates plasticity, which is crucial for regulating synchronous brain activity....

The phenomenon of bursting activity has been discovered for a long time ago, but its origin is still unclear. We propose new phenomenological model on the basis short term synaptic plasticity (STSP), recurrent connections and neuron-glial interactions in order to understand patterns observed experiments. found that can produce are important understanding complex dynamics neuronal networks.

This study introduces a novel method for detecting the post-COVID state using ECG data. By leveraging convolutional neural network, we identify "cardiospikes" present in data of individuals who have experienced COVID-19 infection. With test sample, achieve an 87 percent accuracy these cardiospikes. Importantly, our research demonstrates that observed cardiospikes are not artifacts hardware-software signal distortions, but rather possess inherent nature, indicating their potential as markers...

We propose a new model for neuromorphic olfactory analyzer based on memristive synapses. The comprises layer of receptive neurons that perceive various odors and "decoder" recognize these odors. It is demonstrated connecting layers with synapses enables the training to two types odorants varying concentrations. In absence such synapses, does not exhibit specificity in recognizing odorants. recognition 'odorant' occurs through neural activity group decoder have acquired odorant learning...

In this study, we introduce an innovative hybrid artificial neural network model incorporating astrocyte-driven short-term memory. The combines a convolutional with dynamic models of synaptic plasticity and astrocytic modulation transmission. model's performance was evaluated using simulated data from visual change detection experiments conducted on mice. Comparisons were made between the proposed model, recurrent simulating memory based sustained activity, feedforward depression (STPNet)...

This work is driven by a practical question, corrections of Artificial Intelligence (AI) errors. Systematic re-training large AI system hardly possible. To solve this problem, special external devices, correctors, are developed. They should provide quick and non-iterative fix without modification legacy system. A common universal part the corrector classifier that separate undesired erroneous behavior from normal operation. Training such classifiers grand challenge at heart one- few-shot...