Computational Neuroscience
Computational Neuroscience (of English computation. Computation, information processing, and Neuroscience: Neuroscience, Brain Research) is an interdisciplinary branch of science that deals with the information processing properties of the nervous system. Information processing means the entire spectrum of brain functions of the various stages of processing of sensations up to cognitive functions such as learning, memory, decision-making, as well as the control of the motor system for execution of actions.
The main methodological tool of computational neuroscience is the mathematical modeling of components of the nervous system such as neurons, synapses and neural networks with the methods and findings of biophysics and the theory of dynamic and complex systems. These models are often simulated due to their complexity in the computer. In addition, the computational neuroscience and neural analysis of experimental data. In all these approaches, a close cooperation between experimentally working scientists from the disciplines of biology, medicine, psychology, and physics, as well as theorists from mathematics, physics and computer science is required. The experimental data provide both the basis for the models (eg electrophysiological properties of neurons and synapses, network structures in real neural networks ) as well as the opportunity to test their predictions, such as on certain dynamic or information-processing properties. The models make it possible, in turn, often varied and, for example, to organize seemingly contradictory results of the experiments systematically and identify by mathematical analysis and simulation of complex relationships that are difficult or impossible to detect without this method.
Object modeling are structures on all sizes and complexity scales, ranging from biophysical simulations of the molecular dynamics of certain ion channels and neurotransmitter, using models of individual nerve cells to complex network models, simulate the interactions between brain regions. Depending on the question, these models may have very different levels of abstraction, that is, either to be closely applied to experimental data or rather illustrate the general principles and structures and formalize, which were obtained from the experiments.
Computational Neuroscience can be up demarcated to some extent compared with connectionist theories of psychology, pure learning theories such as machine learning and artificial neural networks, as well as the field of psychiatric computer science, although these areas in part have parallel histories and partly also pursue similar goals. Modeling approaches of computational neuroscience have to make claim to depict certain aspects of the neuronal structures of biologically realistic and direct predictions about such experiments. Connectionist models follow a similar forecast target at the level of psychophysical experiments, but have only a limited right to biological realism, which is limited to the structure of connections and the ability to learn. The same applies to learning theories, but which are often used in addition to purely technical purposes, such as for the prediction of complex time series or for pattern recognition in images. In this application-oriented areas, the analogy to the brain plays only a minor role, an understanding of human information processing is not sought. The neuro- computer science finally accepts, following her name, informatics perspective on the neuroscience one. This includes among other things the development of databases, data structures and standards for efficient storage, archiving and exchange of experimental data, and the development of software both for modeling of neural systems (eg Neuron, Genesis, NEST) and the acquisition and analysis of experimental data. More abstract approaches such as artificial neural networks and machine learning neuro- computer science are attributed sometimes also.
Research Topics
An early neuron model (1952 ), which, partly modified, often is the basis of today's software, is the Hodgkin -Huxley model. Starting from a description of the key ion channels affected by electrical properties of the cell membrane of the neurons in the form of an equivalent circuit diagram modeling the generation of action potentials. The mathematical methods used in the various models are mainly from the theory of dynamical systems. In part, the erratic behavior of neurons ( eg in the range of the threshold potential ) is carried by bifurcations account.
Examples of the application of such models are the description of the cells in the basal ganglia, with the aim to develop new therapies for Parkinson's disease in which the individual cells and modeling (such as the software neuron possible) is important and experiments, complex cognitive processes such as the Stroop test with the Emergent program to describe, with additional effects such as the Hebbian learning rule play a role, but individual cells are significantly more simplified due to the considered number.
History
The term "Computational Neuroscience" was introduced in 1985 by Eric L. Schwartz. Schwartz had organized a conference in Carmel, California this year at the request of the Systems Development Foundation. This had to give objective, an overview of a branch of science that was previously associated with a number of different terms such as " neural modeling ", " brain theory " or " neural networks ". The contributions to this conference were published in 1990 in a book called "Computational Neuroscience".
The early history of this area is closely associated with the names of scientists like Louis Lapique, Alan Lloyd Hodgkin and Andrew Fielding Huxley, Wilfrid Rall, David H. Hubel and Torsten N. Wiesel, and David Marr.
Lapique led 1907, the integrate- and-fire neuron model is a representative because of its simplicity remains one of the most popular models of computational neuroscience. Nearly 50 years later, Hodgkin and Huxley studied the experimentally highly accessible giant axon of the squid and their investigations initiated from the first biophysical model of the action potential from ( Hodgkin -Huxley model ), which she published in 1952. Rall extended this model to the cable theory, which laid the foundation for neuron models that are composed of spatially extended compartments (soma, axon, dendrite ). Today, such models are used to morphologically exact simulation example used using neuron.
Hubel and Wiesel conducted research into the cells of the primary visual cortex, the first area of the cerebral cortex that receives visual information from the retina. They discovered, among other things, that the cells of the primary visual cortex reflect not only the spatial structure of the image on the retina, but also the spatial orientation of the perceived objects can be read. Both Hodgkin and Huxley as well as Hubel and Wiesel received the Nobel Prize for Medicine (1963 and 1981 ) for her work.
Marr's work focused on the interactions between neurons of different areas such as the hippocampus and cerebral cortex. He presented a theory of vision, which is based on the principles of electronic data processing in the computer. He is considered one of the founders of neuroscience computer science.