Computational Neuroscience

What is Computational Neuroscience?

Computational neuroscience uses mathematical models, algorithms, and simulations to study the nervous system. Researchers compare models with experimental data to investigate how neurons and networks process information and contribute to behavior.

Key Concepts and Techniques in Computational Neuroscience

• Neural Modeling

  Neural modeling involves the creation of mathematical or computational representations of neurons, neural circuits, or entire brain regions. These models can be used to simulate the behavior of individual neurons or networks of neurons, investigate the properties of neural systems, and test hypotheses about neural function and organization.

• Data Analysis and Machine Learning

  Computational neuroscience often involves the analysis of large-scale experimental data, such as neuronal recordings or brain imaging data. Researchers use statistical methods and machine learning to look for patterns in recordings and imaging data.

• Simulation and Visualization

  Simulation and visualization are important tools in computational neuroscience, allowing researchers to test the behavior of their models under various conditions and to visualize the spatial and temporal patterns of neural activity. High-performance computing resources are often used to run large-scale simulations of complex neural systems.

• Network Science

  Network science is an interdisciplinary field that provides a framework for studying the structure, function, and dynamics of complex networks, including neural networks. Computational neuroscientists use network science techniques to investigate the organization and connectivity of neural systems, as well as the principles that govern their information processing capabilities.

Major Areas of Research in Computational Neuroscience

• Neural Coding and Decoding

  Neural coding and decoding are fundamental questions in computational neuroscience, focusing on how information is represented and processed in neural systems. Researchers develop models to understand how neurons encode sensory information, represent internal states, and generate motor outputs, as well as how these neural codes can be decoded and interpreted.

• Neural Dynamics and Computation

  Researchers model how neural activity changes over time, including oscillations and the interactions that produce them.

• Learning and Plasticity

  Learning and plasticity are central topics in computational neuroscience, with research focusing on the mechanisms by which neural systems adapt and change in response to experience or injury. This includes the development of models that capture synaptic plasticity, the formation and modification of neural connections, and the emergence of functional organization in neural networks.

• Cognitive and Systems Neuroscience

  Models let researchers make specific predictions about perception, attention, memory, decision-making, and language, then compare those predictions with observations.

Applications of Computational Neuroscience

• Neuroinformatics

  Neuroinformatics is a subfield of computational neuroscience that focuses on the development of databases, tools, and standards for managing, sharing, and analyzing neuroscience data. This includes the creation of shared data repositories, standardized data formats, and software tools for data analysis and visualization.

• Brain-Inspired Artificial Intelligence

  Computational neuroscience has informed the development of artificial intelligence algorithms, such as artificial neural networks and deep learning. Some artificial neural networks draw inspiration from biological systems, but modern AI models are often substantial abstractions rather than faithful replicas of neural structure or function.

• Neuroprosthetics and Brain-Computer Interfaces

  Insights from computational neuroscience have informed the design of neuroprosthetic devices and brain-computer interfaces, which allow for direct communication between the brain and external devices. These technologies have potential applications in areas such as assistive devices for individuals with disabilities, neurorehabilitation, and human-computer interaction.

• Neuromorphic Engineering

  Neuromorphic engineering is an interdisciplinary field that aims to develop hardware and software systems that emulate the structure and function of biological neural networks. Neuromorphic systems use ideas from neural computation to pursue efficient processing for tasks such as robotics and sensory computation. Performance and energy advantages depend on the hardware, model and benchmark.

Sources and evidence

• Hans Ekkehard Plesser et al.: Building on models—a perspective for computational neuroscience. A model is an abstraction whose biological fit and performance must be tested; the paper is a workshop perspective.