|
|
Thomas F.
Collura, Ph.D., P.E. December 7, 1997
Introduction:
The EEG is an electrical waveform that is recorded
from the brain by using electrodes appropriately
placed on the head, then amplifying and displaying
the electrical signal using a computer, or other
suitable instrument. It consists of a wave that
varies in time, much like a sound signal, or a
vibration. As such, it contains frequency components
that can be measured and analyzed, and these
frequency components have interesting and valuable
properties.
A great deal of history is involved in the
definition, naming, and use of these frequency
bands. They are named using Greek letters, a
convention that was begun by Hans Berger, the
discoverer of the EEG in humans. He observed all of
the rhythms known today (except the 40 Hz "gamma"
band), and described many of their basic properties.
Since then, our definitions and understandings of
the rhythms have been refined. However, there still
remains some uncertainty, and controversy, in how to
define and use these bands, for various purposes.
Approaches to understanding:
There are many ways to approach the understanding of
brainwaves. Clinicians view them for diagnostic
purposes, seek to identify patterns that associate
with specific pathologies or conditions.
Psychologists also study them in association with
mental states, mental processing, and to test
concepts of how the brain processes information. We
also know from introspective reports, and structured
experiments, which subjective states tend to be
correlated with a predominance of the various
brainwave components.
Brain rhythms can also be operantly trained, using
biofeedback. By training an individual to learn how
to produce (or reduce) specific frequencies, changes
in the brain can be produced. From a training
standpoint, we can learn what types of mental states
or activities are affected by specific types of
training. Similarly, we can learn which brain/mind
states, qualities, or activities are associated with
a preponderance of, or conversely a lack of, any
particular rhythm or combination of rhythms.
Generally, we cannot tell from the EEG "what I am
thinking" - but we may be able to say "You are
thinking that this is interesting" or "You are
thinking that this is not interesting" We might be
able to say "you cannot relax without drifting off"
which is to read into another's introspective state,
but not in terms of knowing what is the content of
the thoughts.
It is important that we allow the brainwave signals
to tell us what they have to say, and not try to
force their meaning into familiar, predefined terms.
For example, to expect the brainwave, in a primitive
sense, to indicate, for example, "this is the rhythm
for attention." or "this is when you are thinking
'up'", and so on, are ill-conceived. Rather, we need
to study the patterns that emerge during various
behavioral, as well as introspective, states, and
then see what they are defining in terms of a
multidimensional representation of some state-space.
Research that is focused on understanding specific
properties, such as attention, alertness, mental
acuity, etc., has uncovered combinations of rhythms,
and other EEG properties, that are relevant to these
studies. Generally, "derived" properties are found
useful, that involve computer-processing of the EEG,
to produce measurements that are useful for
research, monitoring, etc.
How brain rhythms are generated:
Populations of cells generate rhythms when they
depolarize in synchrony. This activity occurs
primarily in the upper 4 layers (about 1/4 inch
thick) of the outer layers of the cerebral cortex.
The presence of an EEG rhythm indicates that there
is some brain activity occurring in terms of
millions of cells acting together, in a synchronized
fashion. The exact causes of this, and what it means
for the brain and information processing, is an
entire dissertation in itself.
Overall, the observed brainwave frequencies must be
thought of as "epiphenomena," which are the
byproduct of normal brain function, but not a brain
signal in themselves. The brain does not
communicate, or do its business, using the EEG.
Rather, it is a secondary measure, such as the
vibration measured from an engine, or the
temperature of an electronic circuit. Therefore, the
brain does not, for example, produce alpha waves for
any purpose. It produces them as a result of certain
types of brain activity, and we can learn to
recognize them, and take advantage of them, by
learning what they represent, and what happens when
we work with them.
Distribution in Time and Space:
The brain consists of over 100 billion cells,
organized into many different regions, all doing
different things, all acting simultaneously. The
brain is not a computer. It is an assemblage of
millions and billions of computers. Therefore, at
any time and any particular location, the brain may
produce a combination of frequencies. Variations in
time, and in space (observed as different places on
the scalp) are important to understand.
EEG signals are seen to wax and wane, which means to
grow larger and smaller, in time, generally showing
moment-to-moment variation at all times. Alpha is
almost always seen in "spindles" and "bursts,"
almost never seen in a continuous wave. It is the
production of more, or larger, bursts of rhythmic
activity, that is associated with their being a
higher "amount" of that component. Beta, for
example, may occur in very small bursts, of 1/10
second or less, so that it comes and goes very
rapidly. Alpha, on the other hand, generally waxes
and wanes with bursts of from about 1/5 second, up
to 1 or 2 seconds in length.
Spatial distribution can be seen in all components.
Some of these are described below. Since the brain
consists of broadly identifiable areas (frontal
(motor and sensory cortex), parietal, occipital
(visual), temporal (hearing, language), rhythms are
seen to be associated with the particular involved
area. Electrode placement is therefore important
when measuring or training for particular rhythms.
Training at a location will affect the EEG activity
primarily at that location.
The EEG is thus like a symphony, which is a complex
mixture of sounds, changing in time and in space.
The brain is a massively parallel processor that
contains many thousands of cell systems. There may
be a preponderance of one or more rhythms at any
time, and this combination of frequencies, in time
and in space, can help us to understand the
condition, and the activities, of the brain. There
has been work that suggests the existence of a
specific "alpha state," for example. Even though the
brain may be producing a preponderance of alpha
waves at any instant, this does not necessarily
suggest an "alpha state," per se. Brain states may
exist, and they may be correlated with the presence
or absence of various frequencies, in time and
space, rather than just one frequency.
Training of EEG rhythms
Biofeedback techniques can be used to train EEG
rhythms. Training systems can use visual feedback,
auditory feedback (sounds), or use a personal
trainer to provide verbal feedback, thus making the
trainee aware of which brain rhythms are present.
Displays can be of many types, and computer displays
are capable of producing a wide variety of useful
displays. These can include "thermometers", video
games, and other graphic displays. Systems can be
set up to train to reinforce, or to reduce, any
rhythm or combination of rhythms, or for more
complex situations such as training different
locations to be synchronized, or desynchronized, or
to train different locations to produce (or inhibit)
different frequencies.
Early EEG training emphasized the production of a
particular frequency, for example, alpha-wave
training. More recently, the emphasis has been on
training flexibility, or appropriateness, of brain
rhythms. That is, the brain needs to produce the
desired rhythms at the proper times, and in the
proper locations. The development of these complex
protocols is an important area of current research,
and clinical development.
We can also train more complex, derived properties,
such as brainwave synchrony, coherence, or
relationships between brain rhythms recorded from
different sites. This has been found particularly
useful in training concentration and relaxation, for
peak-performance training, and for athletics,
golfers, etc. Certain EEG properties have been found
conducive to being "in the zone," which is a highly
efficient and responsive state, useful for improving
performance in many applications.
It is important to realize that, although rhythms
can be trained, to produce desired results, the
production (or reduction) of the specific rhythm is
not an end in itself, and the change in the EEG may
not signify that the desired change has occurred.
Rather, the desired brain/mind changes are a
byproduct of the training, independent of changes in
the EEG itself. The brain is a self-regulating
system, and may behave much like a thermostat, that
tries to keep the system stable. To use an analogy,
if a window is left open in a house in the winter,
the house may not be cold, but the furnace will be
working hard, and the heating bills will be high. If
the window is closed, representing a return to
normal operation, the temperature may not rise
significantly, but the furnace will work less. Thus,
the brain may achieve a desired state, even if the
measured variable, the brain rhythms, do not change
significantly, in and of themselves. Nonetheless,
changes in the brain have occurred, and their
benefits may be forthcoming, even in the absence of
large changes in the EEG signal.
Summary of EEG Frequency Bands:
The basic EEG rhythms are summarized briefly as
follows, with regard to their typical distribution
on the scalp, subject states, tasks, physiological
correlates, and the effects of training. This
summary should be taken as a general roadmap, not as
fixed and hard rules.
* Delta (0.1-3 Hz):
Distribution: generally broad or diffused, may be
bilateral, widespread
Subjective feeling states: deep, dreamless sleep,
non-REM sleep, trance, unconscious
Associated tasks & behaviors: lethargic, not moving,
not attentive
Physiological correlates: not moving, low-level of
arousal
Effects of Training: can induce drowsiness, trance,
deeply relaxed states
* Theta (4-7 Hz):
Distribution: usually regional, may involve many
lobes, can be lateralized or diffuse;
Subjective feeling states: intuitive, creative,
recall, fantasy, imagery, creative, dreamlike,
switching thoughts, drowsy; "oneness", "knowing"
Associated tasks & behaviors: creative, intuitive;
but may also be distracted, unfocused
Physiological correlates: healing, integration of
mind/body
Effects of Training: if enhanced, can induce
drifting, trancelike state if suppressed, can
improve concentration, ability to focus attention
* Alpha (8-12 Hz):
Distribution: regional, usually involves entire
lobe; strong occipital w/eyes closed
Subjective feeling states: relaxed, not agitated,
but not drowsy; tranquil, conscious
Associated tasks & behaviors: meditation, no action
Physiological correlates: relaxed, healing
Effects of Training: can produce relaxation
Sub band low alpha: 8-10: inner-awareness of self,
mind/body integration, balance
Sub band high alpha: 10-12: centering, healing,
mind/body connection
* Beta (above 12 Hz)
The beta band has a relatively large range, and has
been defined as anything above the alpha band.
* Low Beta (12-15 Hz), formerly "SMR":
Distribution: localized by side and by lobe
(frontal, occipital, etc.)
Subjective feeling states: relaxed yet focused,
integrated
Associated tasks & behaviors: low SMR can reflect
"ADD", lack of focused attention
Physiological correlates: is inhibited by motion;
restraining body may increase SMR
Effects of Training: increasing SMR can produce
relaxed focus, improved attentive abilities, may
remediate Attention Disorders.
* Midrange Beta (15-18 Hz)
Distribution: localized, over various areas. May be
focused on one electrode.
Subjective feeling states: thinking, aware of self &
surroundings
Associated tasks & behaviors: mental activity
Physiological correlates: alert, active, but not
agitated
Effects of Training: can increase mental ability,
focus, alertness, IQ
* High Beta (above 18 Hz):
Distribution: localized, may be very focused.
Subjective feeling states: alertness, agitation
Associated tasks & behaviors: mental activity, e.g.
math, planning, etc.
Physiological correlates: general activation of mind
& body functions.
Effects of Training: can induce alertness, but may
also produce agitation, etc.
* Gamma (40 Hz):
Distribution: very localized
Subjective feeling states: thinking; integrated
thought
Associated tasks & behaviors: high-level information
processing, "binding
Physiological correlates: associated with
information-rich task processing
Effects of Training: not known
Measuring frequencies
Frequencies may be measured in several ways. One is
to use a Fast Fourier Transform (FFT) to estimate
the amount of energy for all frequencies, in a
defined interval of time, usually about 1 second.
This is accurate, but lacks fast response, if
training is a primary goal. Filtering is also used,
which provides a faster response, but is limited to
specific bands. Digital filters are implemented
using computer software, and are a preferred method.
Copyright (c) 1997, 1998, 1999, 2000, 2001, 2002,
2003, 2004, 2005 Thomas F. Collura, Ph.D. |
