Neural Network

Neural networks provide the biological functionality that mediates consciousness. The neural network includes the peripheral nervous system as well as the central nervous system. A small dynamic subset of neurons, principally in the thalamocortical system, and ever changing on a millisecond-by-millisecond basis as thoughts change, functions as the dynamic core mediating consciousness. Most of the remaining active neurons serve in a supporting role.

Interactions between neurons and neuronal aggregations are most accurately described as networks    (a combination of hierarchies and parallel processors). (Greenfield; Centers of Mind, 40)

Brain Imaging helps to visualize neural networks

Complexity of the neural network is indicated in the diffusion tensor image of the human brain obtained from 3-tesla MRI sequences.

 

Link to — Human Brain, Diffusion Spectrum Image

Link to — Human Connectome Project Proposal Synopsis

Link to — Brain Fiber Pathways

Link to —  Cortical Layers Diagram

Link to —  Sensory Pathways

Link to —  Sensory Pathways Diagram

Link to —  (Rolls; Memory, Attention, and Decision-Making, ) includes info on brain connections

 

Research Study — Axonal Synapse Sorting in Medial Entorhinal Cortex

Research Study — Chained Activation of Neuronal Assemblies supports Major Cognitive Processes

Research Study — Thalamocortical Signals Selectively Amplified via Recurrent Inputs

Research Study — Attention Modulated Top-Down, Selectively Processes Information

Research study — Geometric Structure of the Brain Fiber Pathways

Research study — Connectivity of the Neural Network

Research study — Chromatin Remodeling Regulates Neural Coding

Research study — Astrocyte Local Position Cues Sensorimotor

Research study — Interneuron Activity-Dependent Transcriptional Switch

Research Study — Embryonic Development of Brain

 

Expansive research for the Neural Network

Link to — Cerebral Cortex of Mouse  — Allen Institute project

Link to — Human Connectome Project (HCP)

 

Research Study — Small-World Networks for Cortical Architecture

 

Link to — Convergence-Divergence Zones

Antonio Damasio's convergence zone architecture is a good way of conceptualizing the brain's neural network with its reentry and recursion functionality. The dynamic core of consciousness will be a sparse, widespread ever-changing subset of neural activity, integrating neural assemblies over a continuously advancing time window of  ~100 ms.

                   Reentrant Neural Circuitry

There is no other object in the universe so completely distinguished by reentrant circuitry as the human brain. (Edelman; Universe of Consciousness, 49)

Very fast oscillations (VFO) can occur nested with a slower oscillation, beta-2 (20-30 Hz). (Traub; Cortical Oscillations, 260)

Brain architecture is characterized by abundant reciprocal connections between cortical regions, recurrent pathways that permit feedback and reactivation of active areas, and lateral inhibition that focuses neural activity within active centers by inhibiting less active adjacent regions. (Johnston; Why We Feel, 123)

A highly developed system of excitatory recurrent collateral connections between nearby pyramidal cells is a hallmark of neocortical design. (Rolls & Deco; Noisy Brain, 30)

Thalamocortical system does not contain loops so much as highly connected layered local structures with massively reentrant connections. (Edelman; Bright Air, 117)

It is estimated that a large feedback loop covering the entire brain takes only five or six synapses. (Andreasen, Creating Brain, 59)

Prefrontal regions are reciprocally connected with temporal, parietal, and occipital cortices, where they receive higher-level visual, auditory, and somatosensory information. (Miller; Human Frontal Lobes, 49)

 

Perception-Action Cycle uses Reentry and Recursion

Joaquin Fuster’s Perception-Action Cycle uses reentry and recursion on all hierarchical layers between the frontal motor cortical areas and the posterior sensory cortical areas. Antonio Damasio's Convergence Zone Architecture and Joaquin Fuster’s Perception-Action Cycle likely function in joint combination in the neural network

 

Neural Network's Synaptic Efficacies and Sensory Input Patterns

The brain's neural network is a network of neurons connected by synapses with synaptic efficacies established by genetics and continuously modified by ongoing experience.

Sensory input for the neural network is a multimodal input of the five senses activating millions of neurons in ever changing spatial and temporal patterns.

A sensory input pattern propagates into the brain's established synaptic efficacies pattern, activating a pattern of synaptic efficacies closely conforming to the sensory input pattern.

Neural networks can be thought of as pattern associators, which link an input pattern with the most appropriate output pattern. (Arbib, Handbook of Brain Theory; Anderson; Associative Networks, 102)

Feedforward inhibition serves to impose a temporal framework on a target area on the basis of inputs received. (Andersen; Hippocampus Book, 299)

The brain’s neural mechanisms respond to this activated pattern of synaptic efficacies, resulting in perception, consciousness, inscribing synaptic memory, and all of the other functions.

Minds emerge when the activity of small circuits is organized across large networks so as to compose momentary patterns.  The patterns represent things and events located outside the brain, even in the body or in the external world, but some patterns also represent the brain's own processing of other patterns.  The term map applies to all of those representational patterns. (Damasio; Self Comes to Mind, 18)

"Standard" models of neuronal networks are based on the assumption that timing is determined by membrane depolarization and the pattern of synaptic inputs. (Traub; Cortical Oscillations, 60)

Because principal neurons frequently discharge in bursts of action potentials, the degree of postsynaptic facilitation or depression during such bursts may contain much of the information transmitted through the network. (Andersen; Hippocampus Book, 211)

By mapping the body in an integrated manner, the brain manages to create a critical component of what will become the self. (Damasio; Self Comes to Mind, 92)

Body mapping is a key to the elucidation of the problem of consciousness. (Damasio; Self Comes to Mind, 92)

 

Researach Study — Diverse Coupling of Neurons to Populations of Neurons

Researach Study — Lattice System of Functionally Distinct Cell Types in the Neocortex

Researach Study — Neural Circuit Activity Regulated via Protein Npas4

Researach Study — Axonal Remodeling by Visual Stimulation

 

Neural Network Connectivity Determined by Synaptic Efficacies

Activated neural network connectivity patterns are determined by synaptic efficacies. The neural network is never connected in any deterministic sense. A neuron does not fire deterministically based upon input signals from a fixed set of neurons. A neuron fires whenever its quorum of input signals (excitatory and inhibitory) exceeds its firing threshold in a ~2 ms time window.

Central Nervous System Organized in Successive Plate-Like Stages

Many parts of the central nervous system are organized in successive plate-like stages.(diagram) (Hubel: Eye, Brain, and Vision, 24)

The initial stages of the mammalian visual system have the plate-like organization often found in the Central Nervous System. The first three stages are housed in the retina;     the remainder are in the brain: in the lateral genetic geniculate bodies and the stages beyond in the cortex.(diagram) (Hubel: Eye, Brain, and Vision, 27)

 

Neural Network Functions As a "Dot Product" of Input Signal Vector and Synaptic Efficacies Vector

Metaphorically, much of the functioning of the neural network can be understood as the "dot product" of an input signal vector and the vector of synaptic efficacies. (Dot product is a standard operator for vector computations.)

The dot product operation is the simplest type of processing that might be performed by neuron as it involves taking a dot product of the incoming firing rates with the receiving synaptic weights to obtain the activation of the neuron. (Rolls & Deco; Noisy Brain, 62)

Dot product encoding ensures that the simple emergent properties of associative neuronal networks such as generalization, completion, and graceful degradation can be realized very naturally and simply. (Rolls & Deco; Noisy Brain, 62)

Sparse but Widespread Network of Neurons

Crick and Koch anticipate that at any given moment the neural correlate of consciousness (NCC) will be comprised of a sparse but widespread network of neurons, whose activity will stand out above background neuron firing for at least 100-200 milliseconds.  (Zeman; Consciousness, 289)

The gossamer web of neural network activity comprising the dynamic core at any instant will be widespread and sparse but nonetheless include many millions of axons.  The pattern of neural activity representing a perception will overlay the ensemble of synaptic strengths comprising the self.  In other words, a perception is comprised of a pattern of neural activity that can be understood as the dot product of an input signal vector and the composite vector of synaptic efficacies. In the these terms, the self is comprised of the composite vector of synaptic efficacies, which are established by genetics and a lifetime of experience operating on synaptic plasticity. The emerging property of consciousness is associated with the dynamic core.

There is no real structural top in neuronal hierarchy. (Buzsáki; Rhythms of the Brain, 334)

The top end of computation in neural networks is generally heralded by time, marked by inhibition of activity, rather than by some defined anatomical boundary. (Buzsáki; Rhythms of the Brain, 334)

Mediating Circuits and Modulatory Circuits

Two types of circuits in the brain.   (1) mediating circuits produce behaviors, (2) modulatory circuits act on the mediating circuits, regulating the strength of their synaptic connections. (Kandel; Search of Memory, 224)

Brain Operates As a Reality Emulator

The brain operates as a reality emulator. (Llinás; I of the Vortex, 13)

The brain fashions an internal model of the external world as a basis for prediction and exploration of alternatives. (e.g., mental exploration of possible move sequences in chess) (Holland; Hidden Order, 33)

Events in the outside world are represented by patterns of activity in neocortical areas. (Marrs 1971 model)   (Andersen; Hippocampus Book, 734)

Association is the most natural form of neural network computation. Neural networks can be thought of as pattern associators, which link an input pattern with the most appropriate output pattern. (Anderson; Associative Networks, 102)

Our brains are belief engines, evolved pattern recognition machines that connect the dots and create meaning out of the patterns that we think we see in nature. (Shermer; Believing Brain, 59)

Patternicity -- the tendency to find meaningful patterns in both meaningful and meaningless noise. (Shermer; Believing Brain, 60)

Brain Networks Are Constantly Active

Brain networks are constantly active, even when we are resting or sleeping. (Andreasen, Creating Brain, 59)

Projection neurons tend to be idle in the absence of inputs. Inhibitory interneurons are often active all the time. (LeDoux; Synaptic Self, 50)

The fast dynamics of local recurrent networks may be facilitated by the fact that cortical neurons in the awake state have a low spontaneous rate of firing (a few Hz),    which means that any small additional input    may produce some spikes sooner than otherwise would have occurred,    because some of the neurons may be very close to a threshold of firing. (Rolls; Memory, Attention, and Decision-Making, 26)

Rhythmic Neuronal Activity during Movement

Neural Population Dynamics -- Quasi-oscillatory neural responses are present when a monkey 'reaches'. Rotations of the population state are a prominent feature of the cortical response during reaching. These population-level rotations are a relatively simple dynamical feature yet explain seemingly complex features of individual-neuron responses,

 

Local Neural Circuits Operate Asynchronously

Fundamentally, local neural circuits operate asynchronously. I'll outline here my speculation on asynchronous local circuits linked into coherence.

Even in the quiescent state, projection neurons by their stochastic properties are active intermittently. Interneurons provide a local network of inhibitory signals on the dendritic trees of projection neurons. This nascent borderline activity of local circuitry renders them ready to quickly become fully active and respond to associative activity of slight surges in circulating neural signals.

Orderings of computations is necessary for performance. Ordering does not take place by a strict serial organization. Instead, computations pass information back and forth to coordinate their results. (Koch, Neuronal Theories; Posner; Constructing Neuronal Theories of Mind, 198)

Precise connections exist between anatomically distant areas. A particular anatomical area is active whenever its computation is required. Since computations are often contingent on information from another area, information is fed back to reenter the critical areas. (Posner; Constructing Neuronal Theories of Mind, 198)

Asynchronous local circuits are drawn together in ever higher coherence by associatively linking with signals from other asynchronous circuits, enlarging to form the neural assemblies of Gestalt functionality, active neural firing rates that may dynamically persist a few tens of milliseconds as short-term memory. Much of the local circuitry operates via recurrent circuitry, thereby providing local memory via rehearsal.

Local asynchronous circuits associatively link together in a hierarchy or fractal arrangement of ever higher and more widely grouped reentry circuits. The wider ranging links of this nested fractal reentry circuitry provides the functionality for recursion and Bayesian inference.

Hierarchy and fractals of reentry and recurrent local circuitry produce the complex interactions resulting in "one over f" power spectrum on log-log plot of "pink noise."

 

Oscillatory Humming of Neural Network Activity

Neural network activity with its ubiquitous recurrent and reentrant circuitry generates electrical signals that can be detected on the scalp (EEG) and by invasive probes into the brain.  The spectral frequencies of these neural signals have a power law characteristic, displaying an inverse linear relationship on a log-log plot.  This kind of relationship is called "pink noise."

The amplitude of the EEG power spectrum increases as the frequency decreases.  This inverse relationship is expressed as the "one over f" power spectrum (also called "pink" noise). (Buzsáki - Rhythms of the Brain, 119)

Simultaneous processing of vast numbers of independent local circuits leads to gamma oscillations detected in EEG. Reentrant neural activity in more widespread neural circuits leads to lower frequency oscillations in the power spectrum.

Several frequency bands are often mentioned: Alpha frequencies, Delta frequencies, gamma frequencies, higher frequencies.

Object-Oriented-Programming Metaphor

Object Oriented Programming, used extensively in computer networking, may be a helpful metaphor for understanding local neural circuitry, where synaptically connected signaling circuitry, combined with local recurrent circuitry for the data storage of short-term memory, provide an adaptive functionality to be associatively combined with other local asynchronous neural modules, each with somewhat different functionality, into a dynamically coherent neuronal assembly, spanning a widespread but sparse dynamic core for consciousness.

Autoassociative Networks

Autoassociative networks can recall the appropriate memory from the network when provided with a fragment of one of the memories. This is called completion. (Rolls & Deco; Noisy Brain, 31)

By the associative property of memory, neural assemblies aggregate via the laws of Gestalts into the sparse but widespread neural assemblies of the dynamic core of consciousness.

 

Cognitive Networks -- Cognits

Joaquín Fuster has coined the term "cognit" as a generic term for any representation of knowledge in the cerebral cortex.

The cognit is made up of assemblies of neurons and connections between them. (Fuster; Cortex and Mind, 14)

Discussion of Cognitive Networks -- Cognits

 

Inhibitory Activity for Steady-State

Inhibitory activity to maintain a resting steady-state is a frequent theme in the neural network. In the basal ganglia, inhibitory activity suppresses FAPs, which always stand ready to initiate movement. To effect a movement commanded by the motor cortex, an excitatory signal stimulates an inhibitory path that inhibits the quiescent inhibitory activity of the basal ganglia. The result is that body movement can respond quickly, jack-in-the-box fashion, but dampened appropriately by the cerebellum.

Brain is a jack-in-the-box, loaded to the brim with spring-loaded plans of action.  (Hobson; Dreaming as Delirium, 117)   [Stereotyped motor programs]  [FAPs]

 

Research study — Interneurons Project Long-Range in Hippocampus and Entorhinal Cortex

Research study — Gap Junctions Compensate for Dendritic Integration

 

Plasticity and Memory of Neural Network -- Synapses

Neural Network — Plasticity

Neural Network — Short-Term Memory

 

 Dendritic Tree Synapses are Dynamically Reconfigured for the Dynamic Core

Dendritic trees contain spines, which may function to provide highly effective and modifiable chemical and electrical compartments that regulate synaptic efficacy, integration and plasticity.

 

Research Study — Dendritic Field Orientation Toward Active Axons in Developing Cortex —  regulation of dendrite orientation is conserved across species and cortical areas and shows how high-acuity sensory function may be achieved by the tuning of subcellular polarity to sources of high sensory activity.

Research study — Dendritic Spikes enhance Stimulus Selectivity

 

Neural signal processing in dendritic trees is being intensively researched.

Link to — Neural Network - Dendritic Tree Research

 

Many Local Functional Modules

Metaphor of Object-Oriented Programming — I believe there are many tiny local functional modules operating in parallel and hierarchically with recurrent and reentrant connections operating recursively.  Antonio Damasio’s dispositions operating with convergence-divergence zones are a good way of conceptualizing the autonomous functionality of these local networks.

 

             Research study — Visual Cortex Functional Microcircuits

 

Neurogenesis in Adult

Research study — Hippocampal Neurogenesis in Adult

 

Three Major Topological Arrangements in the Brain

(1) Thalamocortical system, (2) Parallel, unidirectional pathways through the basal ganglia and through the cerebellum, (3) Fan-out meshwork of diffusely projecting neuromodulatory neurons emanating from brain stem nuclei.

Three major topological systems – (diagram) (Edelman; Wider than the Sky, 26)

Frontal Cortex Connected to Subcortical Structures

Frontal-cortical regions are connected to a complex circuitry of subcortical structures. (Miller; Human Frontal Lobes, 14)

White Matter of Connecting Nerve Fibers Beneath Gray Matter of Cortex

White matter lies underneath the cortex and is made up of nerve fibers connecting the cortex with the rest of the nervous system. Prefrontal white matter is disproportionately larger in humans than other primates, which suggests a higher degree of connectivity in this part of the brain. (Gazzaniga; Human, 21)

Higher-order association areas tend to have the greatest density of commissural projections, whereas fewer interhemispheric connections are present between primary sensory and motor cortices. (Miller; Human Frontal Lobes, 51)

Research Study — Language—Associated Gene SRPX2 Regulates Synapse Formation

 

Thalamocortical system

The thalamocortical system, by its hublike organization, allows radial communication of the thalamic nuclei with all aspects of the cortex. These cortical regions include the sensory, motor, and associational areas. These areas subserve a feedforward/feedback, reverberating flow of information. (Llinás; I of the Vortex, 126)

Thalamocortical system evolved to receive signals from sensory receptor sheets and give signals to voluntary muscles. Thalamocortical system is very fast in its responses (milliseconds to seconds), although its synaptic connections undergo some changes that last a lifetime. Cerebral cortex is arranged as a set of maps, which receive inputs via the thalamus. Thalamocortical system does not contain loops so much as highly connected layered local structures with massively reentrant connections. (Edelman, Bright Air, , 117)

Most neurons in the thalamocortical system receive signals from other neurons, rather than directly from sensory inputs. (Edelman; Universe of Consciousness, 137)

Prewired by nature

Brain is prewired (i.e., patterns of synaptic efficacies) by nature and for the most part genetically determined in embryonic development and early childhood: vision (Hubel and Wiesel); somatosensory (Mountcastle); language (Chomsky).  (Llinás 193)

Precise point-to-point wiring     cannot occur;    the variation is too great    for the information stored in the genome. (Edelman; Bright Air, 25)

Organization of interneurons in the spinal cord for vertebrates is quite complex. (Kandel; Principles of Neural Science, 1248)

Complex Networks

The concept that physical systems, made up of a large number of interacting subunits, obey universal laws that are independent of the microscopic details is a relatively recent breakthrough in statistical physics. (Buzsáki; Rhythms of the Brain, 127)

Nervous systems are organized on multiple scales, from synaptic connections between single cells, to the organization of cell populations within individual anatomical regions, and finally to large-scale architecture brain regions and their interconnecting pathways. (Sporns; Networks of the Brain, 35)

The modularity of brain's architecture effectively insulates functionally bound subsystems from spreading perturbations due to small fluctuations in structure or dynamics. (Sporns; Networks of the Brain, 71)

The response of a neuron depends on the immediate discharge history of the neuron and the long-term history of the connectivity of the network in which it is embedded. (Buzsáki; Rhythms of the Brain, 127)

Even a weak transient local perturbation can invade large parts of the network and have a long-lasting effect, whereas myriads of other inputs remain ignored. (Buzsáki; Rhythms of the Brain, 127)

Although neuronal networks of the brain are in perpetual flux, due to their time-dependent state changes, the firing patterns of neurons are constrained by the past history of the network.  Complex networks have memory. (Buzsáki; Rhythms of the Brain, 127)

 

Link to diagram — Neural Network Multiple Reentry Loops

 

Local Network Connectivity

 

Research study — Network Local Synaptic Connections

Research study — Neural Network Local Circuit Connections

Research Study — Synapse Elimination via Astrocytes

Research study — Dendritic Discrimination of Temporal Input Sequences

Research study — Dendritic Microcircuit Computations

Research study — Cortical Connectivity with Axodendritic Overlap — Multiple types of axon overlap with the dendrites of cortical neurons. Some axons arise locally, whereas others ascend from the thalamus or descend from higher cortical areas.

 

Neural Network

A principle of cortical design is that it does have in part local connectivity, so that each part can have its own processing and storage, which may be triggered by other modules, but is a distinct operation from that which occurs simultaneously in other modules. (Rolls & Deco; Noisy Brain, 23)

This independent local processing and storage leads to asynchronous operation whereby recurrent and reentry signals at multiple spatial and time scales produce a spectral characteristic with an inverse relationship on a log-log plot.

Widespread circulation of neural signals in a more quiescent state leads to gross synchronization at lower frequencies.

In primary sensory cortical areas, the main extrinsic forward input is from the thalamus, and ends in layer 4, where synapses are formed onto spiny stellate cells.  These in turn project heavily onto pyramidal cells in layers 3 and 2, which in turn send projections forward to the next cortical area. (Rolls & Deco; Noisy Brain, 23)

The circuitry is more complex because the thalamic afferents also synapse onto the basal dendrites in or close to the layer 2 pyramidal cells, as well is onto layer 6 pyramidal cells and inhibitory interneurons. (Rolls & Deco; Noisy Brain, 23)

Forward corto-cortical afferents to a cortical area sometimes have a columnar pattern to their distribution, with a column width 200-300 µ in diameter. (Rolls & Deco; Noisy Brain, 25)

Even for modest firing frequencies such as 20 Hz, the action potential amplitude, measured about 300 µ from the soma,    attenuates to less than half of its amplitude at lower frequencies. (Andersen; Hippocampus Book, 146)

Cortical processing takes place through a hierarchy of cortical stages.  Convergence and competition are key aspects of the processing. (Rolls & Deco; Noisy Brain, 26)

Anatomical convergence of information from different sensory modalities at the hippocampus. (Andersen; Hippocampus Book, 738)

Cortical connectivity apparatus of the perception-action cycle is completed in both directions at every hierarchical level. (Fuster; Prefrontal Cortex, 360)

Orbitofrontal cortex, amygdala and hippocampus are stages of information processing at which the different sensory modalities (such as vision, hearing, touch, taste, and smell for the orbitofrontal cortex and amygdala) are brought together, so that correlations between inputs in different modalities can be detected in these regions, but not at prior cortical processing stages in each modality, as these cortical processing stages are mainly unimodal. (Rolls & Deco; Noisy Brain, 27)

Paths in and out of the cerebral cortex

 Layer IV is typically the input layer.  Layers V and VI are typically output layers. Layers I - III provide interconnections, sideways connections. (Calvin; Neil's Brain, 90)

Degeneracy, Redundancy, Fault Tolerance

The neural network has the property of degeneracy, redundancy, fault tolerance. No one neuron is essential and critical for the operation and functioning of the neural network.

Structurally variable but functionally equivalent networks, are an example of degeneracy,    defined as the capacity of systems to perform similar functions despite differences in the way they are configured and connected. (Sporns; Networks of the Brain, 68)

Degeneracy is widespread among biological systems    and can be found in molecular,    cellular,    and large-scale networks. (Sporns; Networks of the Brain, 69)

 

Populations of Neurons and Stochastic Behavior

Neural Network — Populations of Neurons

Neural Network — Stochastic Behavior

 

Rate Coding in Cortex

Neural Network — Rate Coding in Cortex

 

Hierarchical Networks and Autoassociation Attractive Systems

Neural Network — Hierarchically Arranged Networks

Neural Network — Autoassociation Attractor Systems

 

"The Self" of a Person -- Synaptic Molecular Signature

A person' s individuality as a person is represented in the neurons and synapses, thus forming the molecular signature, or "the self," of the person.

Brain can achieve the modification of body maps very rapidly, in the time scale of hundreds of milliseconds or less. (Damasio; Looking for Spinoza, 118)

Timescale for the brain to induce changes in the body-proper is seconds. It takes about one second for a long and often myelinated axon to convey signals to body parts. (Damasio; Looking for Spinoza, 118)

It takes a few seconds for a hormone to be released into the bloodstream and began to produce its cascade of subsequent effects. (Damasio; Looking for Spinoza, 118)

 

Research Study — Activity-Dependent Tanscription Factor

Research study — Development of Large-Scale Functional Brain Networks in Children

 

“Remembered Present” - Consciousness

"The Remembered Present" - dynamic interaction between memory and ongoing perception that gives rise to consciousness. Activity of the reticular thalamic nucleus gates various activity combinations of specific thalamic nuclei corresponding to different sensory modalities. Intralaminar thalamic nuclei, which send diffuse connections to most areas of the cortex, help to synchronize thalamocortical responses and regulate the overall levels of activity in these multiple reentrant systems. (Edelman; Wider than the Sky, 55)

 

Diagram of Convergence-Divergence Zones Architecture by Damasio

Diagram of Convergence-Divergence Zones Architecture Functionality

 

Research study — Neural Network Connectivity

 

Research study — Cortical Networks for Vision and Language

 

Research study — Cortical Connectivity with Axodendritic Overlap

 

Network Links

Autoassociation Attractor Systems

Hierarchically Arranged Network of Synaptic Connections

Neural Network Plasticity

Populations of Neurons

Short-Term Memory

Stochastic Behavior

Memory Trace Pattern

 

Dynamic Link Architecture -- von der Malsburg

Network theory — Dynamic Link Architecture

 

Basal Ganglia and Cortex

The large-scale organization of the basal ganglia can be viewed as a family of reentrant loops that are organized in parallel, each taking its origin from a particular set of functionally related cortical fields, passing through the functionally corresponding portions of the basal ganglia, and returning to parts of those same cortical fields by way of specific basal ganglia recipient zones in the dorsal thalamus. (Arbib, Handbook of Brain Theory; Alexander; Basal Ganglia, 139)

Hebbian Plasticity

Classical conditioning as Hebbian plasticity (LeDoux; Synaptic Self, 160)

LTP occurs in accumbens circuits. Dopamine facilitates Hebbian plasticity. (LeDoux; Synaptic Self, 250)

Hebb's cell assemblies were supposed to be the fundamental units of thought. (Waldrop; Complexity, 164)

Store information through synaptic change in the sensory regions in which it is processed. Plasticity and memory may share a common fundamental explanation in Hebb's rule. (Zeman; Consciousness, 205)

Neural Correlate of Consciousness (NCC) is a loosely linked but temporarily coherent network of neurons [Edelman’s Dynamic Core] around the brain, a grouping called a “cell assembly” by Donald Hebb. (Zeman; Consciousness, 288)

 

Limbic System and Cortex

Modular composition of the brain

Neural networks are greatly influenced by the modular composition of the brain. (Crick; Astonishing Hypothesis, 157, 158)  (Changeux; Neuronal Man, 54, 55, 56, 57)

Brain circuits can be thought of as hierarchically arranged circuits linked together by synaptic connections. (LeDoux; Synaptic Self, 50)

 

Several ways of classification:

excite, inhibit, modulate

neurotransmitters

projection neurons, interneurons

feedback connections (Koch 120)

Neural plasticity, network changes

Memory, perception

 

Columbus metaphor for growth cones. Don’t try to overextend this metaphor; just use it as a vague notion. In subsequent years, many other ships probed the lands. Only a few developed into active ports.

 

Research study — Grid Cells and Theta Oscillations

Research study — Grid Cells dissociated from Head Turning

 

Research study — Neural Network — Recent Research

Research study — Synaptogenesis induced by GABA in the Developing Mouse Cortex

 

 

Link to — Consciousness Subject Outline

Further discussion — Covington Theory of Consciousness