Study shows simple, computationally-light model can simulate complex
brain cell responses
Date:
April 18, 2022
Source:
Tokyo University of Science
Summary:
Studying how brain cells respond to signals from their neighbors
can aid the understanding of cognition and development. However,
experimentally measuring the brain's activity is complicated. Neuron
models provide a non-invasive way to investigate the brain, but
most existing models are either computationally intensive or
cannot model complex neuronal responses. Recently, a team has
used a computationally simple neuron model to simulate some of
the complex responses of neurons.
FULL STORY ==========================================================================
The brain is arguably the single most important organ in the human
body. It controls how we move, react, think and feel, and enables us to
have complex emotions and memories. The brain is composed of approximately
86 billion neurons that form a complex network. These neurons receive,
process, and transfer information using chemical and electrical signals.
========================================================================== Learning how neurons respond to different signals can further the
understanding of cognition and development and improve the management of disorders of the brain. But experimentally studying neuronal networks
is a complex and occasionally invasive procedure. Mathematical models
provide a non-invasive means to accomplish the task of understanding
neuronal networks, but most current models are either too computationally intensive, or they cannot adequately simulate the different types of
complex neuronal responses. In a recent study, published in Nonlinear
Theory and Its Applications, IEICE, a research team led by Prof. Tohru
Ikeguchi of Tokyo University of Science, has analyzed some of the
complex responses of neurons in a computationally simple neuron model,
the Izhikevich neuron model. "My laboratory is engaged in research on neuroscience and this study analyzes the basic mathematical properties
of a neuron model. While we analyzed a single neuron model in this study,
this model is often used in computational neuroscience, and not all of its properties have been clarified. Our study fills that gap," explains Prof.
Ikeguchi. The research team also comprised Mr. Yota Tsukamoto and PhD
student Ms. Honami Tsushima, also from Tokyo University of Science.
The responses of a neuron to a sinusoidal input (a signal shaped like a
sine wave, which oscillates smoothly and periodically) have been clarified experimentally. These responses can be either periodic, quasi-periodic,
or chaotic. Previous work on the Izhikevich neuron model has demonstrated
that it can simulate the periodic responses of neurons. "In this work, we analyzed the dynamical behavior of the Izhikevich neuron model in response
to a sinusoidal signal and found that it exhibited not only periodic
responses, but non- periodic responses as well," explains Prof. Ikeguchi.
The research team then quantitatively analyzed how many different types
of 'inter-spike intervals' there were in the dataset and then used it to distinguish between periodic and non-periodic responses. When a neuron
receives a sufficient amount of stimulus, it emits 'spikes,' thereby
conducting a signal to the next neuron. The inter-spike interval refers
to the interval time between two consecutive spikes.
They found that neurons provided periodic responses to signals that
had larger amplitudes than a certain threshold value and that signals
below this value induced non-periodic responses. They also analyzed the response of the Izhikevich neuron model in detail using a technique called 'stroboscopic observation points,' which helped them identify that the non-periodic responses of the Izhikevich neuron model were actually quasi-periodic responses.
When asked about the future implications of this study, Prof. Ikeguchi
says, "This study was limited to the model of a single neuron. In the
future, we will prepare many such models and combine them to clarify
how a neural network works. We will also prepare two types of neurons, excitatory and inhibitory neurons, and use them to mimic the actual
brain, which will help us understand principles of information processing
in our brain." The use of a simple model for accurate simulations of
neuronal response is a significant step forward in this exciting field
of research and illuminates the way towards the future understanding of cognitive and developmental disorders.
========================================================================== Story Source: Materials provided by Tokyo_University_of_Science. Note:
Content may be edited for style and length.
========================================================================== Journal Reference:
1. Yota Tsukamoto, Honami Tsushima, Tohru Ikeguchi. Non-periodic
responses
of the Izhikevich neuron model to periodic inputs. Nonlinear
Theory and Its Applications, IEICE, 2022; 13 (2): 367 DOI:
10.1587/nolta.13.367 ==========================================================================
Link to news story:
https://www.sciencedaily.com/releases/2022/04/220418093832.htm
--- up 7 weeks, 10 hours, 51 minutes
* Origin: -=> Castle Rock BBS <=- Now Husky HPT Powered! (1:317/3)