What Is Computational Neuroscience?

Computational neuroscience is the use of mathematical analysis and computer simulation to simulate and study the nervous system at different levels: from the real biophysical models of neurons, their dynamic interactions and the learning of neural networks, to the organization and nerves of the brain Quantitative theory of type computing, etc., to understand the brain from a computational perspective, study the nature and capabilities of non-programmatic, adaptive, and brain-style information processing, and explore new types of information processing mechanisms and approaches to create the brain. Its development will have an important impact on intelligent science, information science, cognitive science, and neuroscience ^[1] .

Computational Neuroscience

Studies on the brain and nervous system have a long history. By the end of the 18th century, people recognized that the brain was divided into different parts and performed different functions. Cajal founded the neuron theory in 1891, thinking that the entire nervous system is composed of relatively independent nerve cells in structure. Based on the Cajal neuron theory, Sherrington proposed the concept of synapses between neurons in 1906. In the 1920s, Adrian proposed neural action potentials. In 1943, McCulloch and Pitts proposed the MP neural network model. In 1949 Hebb proposed the rules for neural network learning. In the 1950s Rosenblatt proposed the Perception model. Since the 1980s, research in neural computing has made progress. Hopfield introduced the Lyapunov function (called the "calculated energy function") to give the network stability criterion, which has a direct correspondence with VLSI, laying the foundation for the development of neural computers. At the same time, it can also be used for associative memory and optimization calculations, opening up new ways for neural networks to be used in computers. Amari has done a lot of research on the basic mathematical theory of neural networks, including statistical neurodynamics, the dynamics theory of neural fields, associative memory, and has made some foundational work especially in the area of information geometry. Computational neuroscience research seeks to embody the following basic characteristics of the human brain: The cerebral cortex is a vastly connected and complex system; the calculation of the human brain is based on large-scale parallel simulation processing; "Guest error" and association ability, good at generalization, analogy and promotion; Brain function is restricted by innate factors, but acquired factors such as experience, learning and training play an important role, which shows that the human brain is very strong Self-organizing and adaptive. Many intellectual activities of humans are not carried out by logical reasoning, but are formed by training.

At present, the understanding of how the human brain works is still very superficial, and the research on computational neuroscience is still insufficient. We are facing a new field full of unknowns, and we must explore more deeply in the basic principles and computational theory. Through the analysis and study of the structure, information processing, memory, and learning mechanism of the human nervous system, simulations are performed on the mechanism of human brain work, and new ideas and methods of intelligent science are proposed.

The scientific problems of computational neuroscience are as follows:

• The basic process of neural activity: study the neuronal ion channels and their regulation, synaptic transmission and its regulation, neuronal receptors and signal transduction, and the synchronization mechanism of neural activity.

• Computational model of a single neuron: A single neuron is the basic unit of a neural network. It consists of a neuronal body, dendrites, and axons. Neurons are connected by synapses.

• Neural Mechanisms of Learning and Memory: The nervous system changes its structure and function due to activities and environmental factors. This change is the basis for advanced brain functions such as learning and memory. Investigate the mechanisms and learning rules that produce this plasticity, especially synaptic plasticity. Study the neuron circuit information coding and processing mechanism.

• Molecular mechanism of neuron and nervous system development: Nerve cells are differentiated from neural stem cells during brain development, and then gradually form complex and sophisticated brains through migration, growth of protrusions, and formation of synaptic interconnections. To study the neurotrophic factors that regulate the differentiation of neural stem cells, maintain the survival of nerve cells, regulate the migration of nerve cells, the growth of neurites and the formation of synapses, and study their role and mechanism.

• Neurotransmitters: Study the composition of neurotransmitters, the synthesis, maintenance, release of neurotransmitters and their interaction with receptors.

^[1]

Cognitive neuroscience views that specific brain locations are responsible for specific cognitive functions. This view is derived from many different theories, such as

The origin of cognitive neuroscience has a lot to do with Phrenology. Craniophysiology is essentially a

French experimental psychologist Florence

Studies by European scientists such as John Hughlings Jackson have brought locationism back to the mainstream. Jackson's research is particularly focused on patients with brain injuries who have symptoms of epilepsy. He found that when patients have seizures, they often cause the same

French neuroscientist Paul Broca reported a patient's symptoms in 1861. The patient could understand the language, but couldn't speak, and could only pronounce the word "tan." The patient was later found in his left brain

In 1870, German physicians Eduard Hitzig and Gustav Fritsch published their findings in animal experiments. They apply electrical current to different parts of the dog's cerebral cortex, which can cause different corresponding actions. From this, they believe that the performance of behavior is derived from the operation of brain cells. German

In the early 20th century,

On September 11, 1956, the Conference on Cognitive Science was

Before the 1980s, there was little interaction between the two fields of neuroscience and cognitive psychology. In the late 1970s, the term "cognitive neuroscience" was born in the back seat of a taxi and was co-founded by George A. Miller and Michael Gazzaniga. Cognitive neuroscience begins

Attention-conscious decision-making judgment learning memory