PREFACE
Electroencephalography (EEG) is a genuine global measure of neocortical
dynamic function with sufficient temporal resolution to follow millisecond
time changes. However, this "window on the mind" is often opaque, revealing
only shadows of mental activity. Progress has been slow because good
experimental design and data analysis depend on prior knowledge of both
passive current flow in the head volume conductor and dynamic properties of
neocortical current sources. But such prior knowledge requires experiments
so substantial trial and error is required. Another obstacle is the highly
interdisciplinary nature of EEG; each scientist can be competent only in a
few of the relevant fields. Thus EEG is currently dominated by experimental
studies having minimal theoretical basis in the underlying dynamics.
This book has two major aims. (1) To facilitate communication between
anatomic, physiologic, behavioral, neuropharmacologic, engineering, physics
and other fields related to EEG. (2) To provide a conceptual framework
based on these disparate data which is sufficiently general to embrace brain
theories applied to different experimental designs, scales, and brain
states. This framework can guide future experimental design and data
analysis, followed by new or modified theories.
Because of the critical need to communicate between these disparate
fields, I avoided the typical edited book with isolated chapters. However,
the subject matter demands experts in different fields. This dilemma was
addressed by choosing chapter authors with unusual backgrounds in both
physical science and neuroscience and experience with EEG data. We were
able to provide many cross chapter references; more connections should
become apparent in the future. We tried to avoid the superficial views of
neuroscience or physical science often adopted by practitioners too firmly
restricted to either camp. However, the broad content forced many
compromises. Historical perspectives, mathematical derivations, and many
details are left to the references.
We assaulted the usual communication barriers associated with lack of
mathematics and physical science training by extensive use of metaphors.
Such metaphors are distinguished from genuine theory based on physiological
data which both suggest and limit metaphors. The mathematics was chosen
selectively to supplement words and figures. Density is 0.5 equations per
page compared, for example, to about 3 equations per page in an advanced
physics text.
While written words often allow different interpretations, mathematics
has more universal meanings. For example, concepts like spatial-temporal
scale, pacemaker, wave, top-down interactions, phase transitions, and self
organization take on deeper meanings when verbal descriptions are
supplemented with matching mathematics. To use a modern metaphor, these
concepts have a fractal character not fully revealed in written text. A
neuroscientist's hostility to mathematics may be appropriate when the
mathematics is unrelated to the task at hand; however such reactions may
also occur when readers are confronted with seemingly old concepts that can
no longer be self-interpreted. In the later case, readers become analogous
to blind persons reading braille translations who complain about the
pictures. But the pictures or equations are needed to provide critical
foundations for much of the verbal text. Non-mathematical readers should be
aware of the importance of mathematics in keeping the presentation honest.
Inclusion makes it far more difficult to fudge agreement between theory and
experiment.
Another communication barrier is spatial or temporal scale chauvinism in
which the inherent limitations of any experimental measure are fully
appreciated only in the context of other person's data. The first three
chapters provide both an introduction and overview of later chapters in
order to counter such parochial views.
Chapter 1 introduces some general characteristics of EEG and
magnetoelectroencephalographic (MEG) data in relation to neocortical current
sources. Issues of spatial and temporal scale are considered in connection
with different measures of brain function. The spatial filtering of fields
by the head volume conductor and new high resolution EEG methods are
discussed. Volume conduction is treated more extensively in P. L. Nunez,
Electric Fields of the Brain: The Neurophysics of EEG, Oxford University
Press, 1981.
Chapter 2 takes a physical science view of neocortex. That is, we
propose a model cortex having interactions between functional units ranging
from neuron to neural masses at columnar scales. Top-down and bottom-up
interactions are discussed. Neocortex is considered a global system in
which the dynamic interactions between different parts is its essential
feature(in a manner analogous to dynamic interactions of persons, nations,
and so on in social systems).
In Chapter 3 a metaphoric link between mind, brain, scientific method
and music, violin, physics is suggested to provide connections between the
issues of scale interactions, neural plasticity, pacemakers, local circuits,
standing waves, external control of conscious states, and EEG frequencies.
New EEG data which exhibit quasi-stable, coherent spatial structure are
provided in support of these ideas. A long time scale(several seconds) is
identified which may be related to the short term memory mechanisms
discussed in Chapter 14.
Chapter 4 by Kenneth Pilgreen provides an overview of clinical aspects
of EEG with emphasis on dynamic properties and external control of neocortex
by various neuromodulators. EEG correlates of altered brain state and new
approaches to scientific integration of EEG are described. Dr. Pilgreen is
a practicing neurologist with a master's degree in physics and experience in
both operational and theoretical aspects of EEG.
Chapter 5 by Fernando Lopes da Silva considers the kinds of EEG rhythmic
behavior exhibited at scales ranging from single neurons to neural
networks. The emergence of novel properties in networks under the control
of modulating systems is described. Methods to estimate chaotic attractors
of EEG are also discussed. Dr. Lopes da Silva holds a medical degree, post
graduate study in engineering and physics, and a Ph.D. in neurophysiology.
He has long experience with EEG studies in animals and humans in his
laboratory at the University of Amsterdam.
Chapter 6 by Richard Silberstein describes relations between cognitive
processes and steady-state visual evoked potentials. The visual stimulus is
used during both task and control states to allow study of spatial-temporal
patterns in narrow frequency bands which are artifact-free. The data also
suggest that evoked potentials are partly a mixture of traveling and
standing neocortical waves. Dr. Silberstein has an honors degree in physics
and a Ph.D. in physiology. The EEG data discussed in Chapters 3,6,12, and
13 were recorded at the Swinburne Centre for Applied Neurosciences, headed
by Dr. Silberstein.
Chapter 7 by Alan Gevins and Brian Cutillo describes relations between
cognitive processes and transient evoked potentials. The chapter provides a
review of traditional evoked potentials and the modern studies of
task-related shifting patterns of statistical interdependency between
electrode sites carried out at EEG Systems Laboratory(The famous "shadows of
thought" which also provide experimental support for the regional and local
circuits discussed in Chapter 13). Dr. Gevins has a mixture of engineering
and psychology backgrounds and founded EEG Systems Lab in 1977. Dr. Cutillo
is a psychophysiologist also having long EEG experience.
Chapter 8 reviews discrete linear systems in which physical or
physiological variables exhibit time dependence. This chapter provides
background for local neural circuit models and spectral and coherence
measures of EEG dynamics.
Chapter 9 reviews continuous linear systems in which variables are
functions of both time and location in physical or neural media. Topics
include physical and neurologic wave propagation and membrane diffusion.
This chapter provides background required for more advanced study of complex
global systems.
Chapter 10 introduces nonlinear systems using simple examples of
electric circuits, mechanical oscillators, and ecologic systems. Jump
phenomena, limit cycles, and temporal chaos are included. More advanced
topics associated with continuous nonlinear systems include stable spatial
structures with temporal chaos and statistical measures of dynamic
properties. A critique of correlation dimension estimates of EEG data is
provided.
Chapter 11 suggests that global dynamic behavior is expected in a
cortex-like system. In certain limiting cases coherent spatial modes
governed by linear or quasi-linear differential equations are predicted. It
is argued that local circuit and global dynamic behavior coexist naturally
and changes from more local to more global brain states occur due to
selective changes in local and global control parameters. Part of the
motivation for this theory is to explain aspects of EEG in terms of the
underlying physiology. However, a more important result may be its
description of more general local and global dynamic behavior. The
mathematical developments are presented in the Appendix.
Chapter 12 provides data on the dynamic properties of EEG which appear
to support the theory of Chapter 12 for states in which global properties
are dominant. Included are multichannel data on alpha rhythm, steady-state
evoked potentials, and EEG under halothane anesthesia. These EEG exhibit
coherent, stable spatial structure which can apparently coexist with complex
temporal behavior including limit cycle and chaotic modes. Apparent nodal
lines of standing waves appear consistently in the data.
Chapter 13 by Richard Silberstein considers the selective action of
several neuromodulators in different cortical layers. Since different
layers are more associated with intracortical or corticocortical fibers,
neuromodulators may selectively alter local and global control parameters to
change brain state. Local resonances (typically in the 40 Hz range) are
contrasted with global and regional resonances (below 20 Hz). Brain
diseases including Parkinson's disease, Tourette Syndrom, and schizophrenia
are viewed as pathologically hypercoupled or hypocoupled states resulting
from faulty control parameters. These ideas can be tested with new EEG
experiments. If even partly correct, they should have a major influence on
neuropsychiatry.
Chapter 14 reviews the statistical mechanics of neocortical interactions
at multiple scales as developed over the past 20 years by Lester Ingber.
The theory describes macroscopic neocortical variables such as EEG with a
chain of arguments involving overlapping microscopic and mesoscopic scales.
The general methods originate in quantum and gravitational field theory.
This approach of nonlinear nonequilibrium statistical mechanics has a
complementary relationship to the strictly macroscopic theory of Chapter 11,
similar to the relationship between classical kinetic theory of gasses and
fluid mechanics. An important prediction of the statistical theory is that
the number of neural firing patterns that can simultaneously persist for
several seconds is in the range of 5 to 10. Theses patterns may store
short-term memories which are known to be limited to 7 + 2 items. This work
may also provide quantitative theoretical foundation for the local and
regional circuits suggested earlier, which may produce spatial-temporal
patterns of EEG associated with particular brain states. Dr. Ingber holds a
Ph.D. in theoretical physics and has developed theoretical approaches to
nuclear physics, social systems, and financial markets, although his primary
concern is neocortical dynamic function.
----------------------------------------------------------------------------