Hardware implementation of the ring generator with tunable frequency based on electronic neurons

Nikita M. Egorov,Marina V. Sysoeva,Vladimir I. Ponomarenko,Ilya V. Sysoev
2023-05-08
Abstract:Constructing electronic models of neurons has several applications including reproducing dynamics of biological neurons and their networks and neuroprosthetics. In the brain, most neurons themselves are in a non-oscillatory mode, and brain rhythms arise due to their collective dynamics. In this case, very small ensembles of neurons can act as rhythm generators. such ensembles can be constructed and study within the framework of a radiophysical experiment. In this work, a circuit of a ring rhythm generator was created from several (from two to eight) FitzHugh--Nagumo electronic oscillators with electronic synapses (sigmoid coupling function and delay were implemented). Oscillatory modes were shown to be possible in this circuit as a result of collective dynamics, with the frequency being controlled by means of a delay in the synapse and/or changing the number of elements in the ring. For some parameter values, there were multistability modes, when the implementation of a specific oscillatory regime was determined by initial conditions or could be achieved by a short-term external driving. The constructed generator well models the switching of the main frequency in brain local field potentials at limbic epilepsy, but can be used independently as well.
Adaptation and Self-Organizing Systems
What problem does this paper attempt to address?
The problem that this paper attempts to solve is how to implement a ring - shaped generator with adjustable frequency through hardware, which is based on the electronic neuron model. Specifically, the authors constructed a circuit composed of several (from 2 to 8) FitzHugh - Nagumo electronic oscillators, and these oscillators are connected by electronic synapses (implementing the sigmoid coupling function and delay). The main objective of the study is to show that in such a circuit, due to the collective dynamics, oscillation patterns can be generated, and the frequency can be controlled by the delay in the synapses and / or by changing the number of elements in the ring. In addition, for certain parameter values, there are multistability patterns, that is, the realization of a specific oscillation pattern depends on the initial conditions, or can be achieved by a brief external drive. The core problem of the paper is to explore how to use this hardware - implemented ring - shaped generator to simulate the change of the main frequency of the local field potential in the brain during limbic seizures, and also to explore its application potential as an independent device. In this way, researchers hope to better understand the dynamic characteristics of neural networks, especially the behavior in pathological states, such as seizures, and the possibilities of these models in practical applications.