Chapter 5 – EEG Components and Their Properties

Chapter 5 – EEG Components and Their Properties

The EEG typically consists of a complex waveform that include a mixture of frequencies. However, two considerations motivate us to identify specific components. The first is that often a particular type of wave dominates, and is visually prominent. When this occurs, we say that the EEG is in that particular state, such as “in delta,” or “in theta.” The second consideration is that when frequency transforms or filters are used, it is possible to isolate a component band, even in the presence of other components. Therefore, regardless of which rhythms are dominant, if any, we can always identify component bands using computer processing.

I prefer to refer to these EEG components as “component bands” rather than “frequencies” or “bandwidths.” That is because they are distinguished more properly by when and where they occur in the brain, their visual appearance, and their physiological meaning. The use of frequencies to distinguish them is rather artificial, and can lead to some ambiguities. A particular component may appear outside of its customary frequency range, and just because an EEG component is in a particular range does not mean that it necessarily conforms to that band’s usual definition. For example, alpha may be less than 8 or more than 12, while still meeting the definition of alpha. It should also be emphasized that components are often not truly “sinusoidal” in appearance, and have distinctive morphology. Frequency analysis such as the Fourier Transform assumes that waves are pure sinewaves. Any deviation from a pure sinewave leads to the appearance of higher harmonics, thus complicating the mathematical analysis. Therefore, a visual inspection of the EEG is always important, to avoid these concerns.


Typical EEG Component Band Ranges:

• 
  Delta 1 – 3 Hz
  
• 
  Theta 4 – 7 Hz
  
• 
  Alpha 8 – 12 Hz
  
• 
  Lo Beta 12 – 15 Hz (SMR)
  
• 
  Beta 15 – 20 Hz
  
• 
  High Beta 20 – 35 Hz (may contain EMG)
  
• 
  Gamma 40 Hz
  

It should be emphasized at the outset that the common EEG components are defined not by their exact frequency bands, but by their properties in terms of distribution in time and space, relationship to physiological states, and other properties. The primary EEG complements have been identified through clinical and research experience, and the associated frequency ranges have come along afterward. The frequency bands therefore describe they complements, but do not define them. EEG complements can and do exist outside of their defined ranges, and it is also true that more than one complement may exist in a given frequency band range. Therefore, it is important not to arbitrarily identify any EEG rhythm based solely on its apparent frequency. The location and behavior of the complement are also important, as is the state of the client, possible medications, drowsiness, and other factors.

Delta is the slowest of the rhythmic EEG complements and is generally considered to be between one and three or 4 Hz. Upon visual inspection, Delta is rarely if ever sinusoidal rather it tends to have a distinctive wandering pattern and its shape parenthesis morphology) is important in assigning its origins. A small amount of Delta is normal. However excessive Delta can appear either focal he or globally. And generally reflects injury or dysfunction in neural feedback, access focal Delta is often associated with lack of function in the affected area is. In such cases, down training Delta is often the indicated option and generally has the effect of re-activating the affected areas.

Insert figure 5-1

Figure 5-1. Examples of various forms of delta waves

Theta is a rhythm that is mediated by subthalamic mechanisms, and, like Delta, tends to have a distinctive non-sinusoidal appearance. A certain amount of Theta is normal, particularly in the frontal areas, where it can be associated with a volition and movement. However, access Theta is among the most common deviations associated with brain dysregulation. Focal Theta is often seen in regions that are quote off-line unquote and Theta down training is among the most common options when this is evident.

Insert Figure 5-2

Figure 5-2. Example of Theta waves

Theta (typ. 4-7 Hz)

• 
  Low-frequency rhythm associated with internalized thoughts
  
• 
  Mediated by subthalamic mechanisms
  
• 
  Associated with memory consolidation
  
• 
  Generally non-sinusoidal, irregular
  
• 
  Seen during hypnogogic reverie
  
• 
  Seen as precursor, and sequel to sleep
  
• 
  Edison’s “creativity” state
  
• 
  Distribution: regional, many lobes, laterlized or diffuse
  
• 
  Subjective states: intuitive, creative, recall, fantasy, imagery, dreamlike
  
• 
  Tasks & behavior: creative, but may be distracted, unfocussed
  
• 
  Physiological correlates: healing, integration of mind and body
  
• 
  Effects: enhanced, drifting, trance like, suppressed, concentration, focus
  

The alpha wave is sometimes defined as the “8-12 Hz” rhythm. However, this factor is only incidental in what distinguishes the alpha rhythm. The key aspects of the alpha rhythm is that it is a resting rhythm of the visual system, that it is maximum posteriorly, that it increases when the eyes close, and that it is symmetrical with a characteristic waxing and waning. All of these factors stem from the fact that it is a thalamocortical reverberation involving the visual pathways and the primary visual cortex, and that it represents the visual system relaxing, and also performing some types of background memory scanning. An individual is also typically aware, but relaxed, during periods of alpha.

The actual frequency of alpha can vary outside the 8 – 12 Hz range, and other components can also show up in this range. Therefore, any signal that is in the 8 – 12 Hz range is not necessarily alpha. What is certain is that, if one sees a signal that is sinusoidal and symmetrical, has a characteristic waxing and waning, is maximum posteriorly, and increases when the eyes closed, then it is an alpha wave.

Insert Figure 5-3

Figure 5-3. Alpha waves
Alpha (typ. 8-12 Hz)

• 
  Resting rhythm of the visual system
  
• 
  Increases when eyes are closed
  
• 
  Largest occipital – O1, O2
  
• 
  Characteristic waxing and waning
  
• 
  Generally sinusoidal, hemispheric symmetrical
  
• 
  Indicates relaxation
  
• 
  Role in background memory scanning
  
• 
  Round trip thalamus-cortex-thalamus ~ 100 ms
  
• 
  Typically 8 – 12 Hz, but may be 4 – 20 Hz
  
• 
  Distribution: regional, evolves entire lobes, strong occipital with closed eyes
  
• 
  Subjective states: relaxed, not drowsy
  
• 
  Tasks & behavior: meditation, no action
  
• 
  Physiological correlates: relaxed, healing
  
• 
  Effects: relaxation
  

Insert Figure 5-4.

Figure 5-4. Typical range of alpha peak frequencies in a normal population. (data courtesy of David Kaiser)

Figure 5-4 shows the typical range of alpha peak frequencies in a normal population. It should be noted that a significant percentage of the population will have a peak alpha frequency that is different from the typical value of 10 Hz.

Another rhythm that can occupy the alpha band but which is not alpha, is the “mu” rhythm. This wave has a visually distinct “wicket” appearance, and is clearly nonsinusoidal. It also does not have a characteristic waxing and waning, and is more often maximum centrally, not occipitally. Its meaning is by no means as clear as that of alpha, and some controversy remains whether it is abnormal or normal, and what clinical determinations can be made in its presence.

An interesting aspect of alpha biofeedback was revealed in a study done in the 1960’s. In this study, a controlled experiment was devised in which half of the participants took part in alpha enhancement training, and the other half took part in alpha reduction training. The expectation was that the alpha enhancement group would show improvement, while the alpha reduction group would not. What was discovered, however, is that both groups indicated improvements including increased self-awareness, and ability to relax. These investigators concluded that the alpha training experience was a placebo, because of the similarity in results in both groups. However, a different interpretation is more likely. Whereas the alpha enhancement group indeed learned an alpha state, the other group was actually undergoing a form of activation training, or “squash” training. Interpreting this study as showing placebo effects is similar to having half of a group do pushups, and the other do pullups, and concluding that exercise is a placebo. Naively, these are opposite tasks, but in reality, both are exercise. Similarly, both alpha enhancement and alpha reduction are beneficial neurofeedback tasks, and the presence of benefits in both groups does not imply that there is a placebo effect.

An important distinction must be made between alpha activity and brain activity in a brain region. Alpha activity is associated with reduce brain activation; when alpha waves are present, that location is in an idling state hence is less active. Therefore, when we see increased alpha waves the brain is actually less active. An activated brain state is associated with high-frequency, low-amplitude EEG activity. This is the basis of the “squash” protocols which will be discussed with training protocols. In essence, a training protocol that encourages the lowering of EEG amplitude, particularly alpha or theta amplitude, will result in an activation of the related brain areas.

Asymmetry of alpha, particularly frontal alpha, is important to mood. In normal individuals, left frontal alpha is typically 10% to 15% lower than right frontal alpha. This asymmetry is important for normal mood control. It has been reported (Davidson, . Rosenfeld & Baehr, ) that depression is associated with higher left frontal alpha, and that operant training to restore the asymmetry in the direction of lower left frontal alpha results in mood improvement (Baehr). The rationale behind this approach is the observation that the left frontal area is responsible for positive judgments, and associated approach behavior, while the right frontal area mediates negative judgments, and associated withdrawal behavior. For normal mood, the negative area (right side) should be somewhat less active than the positive area (left side). Therefore, a slightly lower alpha on the left corresponds with greater activation of this positive judgment area. Baehr and Baehr treated depressed clients with alpha lowering training on the left. Hammond (REF) demonstrated improvements in clients treated with a protocol that increased beta in the left frontal region.

Another important subtlety with regard to alpha is the presence of two basically different alpha ranges. The fast alpha range, between 10 and 12 Hz, is the typical occipital resting rhythm. It reflects background memory processing, and an idle yet not inattentive state. The slow alpha, typically taken as 8-10 Hz, appears more frontally, and is more associated with emotional processing. When EEG spectral displays are seen in real time, the independent waxing and waning of these two alphas is clearly evident. When performing assessments and neurofeedback training, it is becoming more common to distinguish between these two alpha bands, and to treat them individually. It is interesting in this regard to note that alpha may slow down with age. While significant alpha slowing is associated with degraded mental processing, a general slowing down with age may indicate that more processing is occurring at the endpoints of the thalamocortical cycling, thus delaying the cycle time and reducing the frequency. As seen in experienced meditators, for example, in this context, slowed alpha may indeed be an indicator of acquired knowledge, possibly even wisdom.

Low beta is one of the more interesting brain rhythms, partially because it comprises the sensorimotor rhythm, when it is observed from the sensorimotor cortex. Low beta is actually and alpha wave, if one considers that it is thalamo-cortical reverberation, and that it represents an idle state. Barry Sterman conditioned pass at him are in cats and discovered that cats who learned to produce SMR were significantly more resistant to the effects of toxins.

Sterman identified SMR as a resting rhythm of the motor system, and described it as indicating the intention to remain still. This is one of the more philosophically profound aspects of neural feedback, that it can indicate and train something as seemingly obscure as intention. Sterman initially began to operably train the SMR rhythm in tax after observing that they produced this rhythm when they were still. By using a reward consisting of broth and milk, he successfully trained some of the cats to increase their SMR rhythm. Later, fate intervened when these cats were randomly assigned to trials of withstanding toxic doses of Mono methyl hydrazine fuel. This later research was intended to assess the toxicity of this aviation fuel. However, anomalous results caused Sterman to re-investigate the data. He then discovered that the operably trained cats were significantly more resistant to the toxic affects of the fuel. In essence, the SMR training had rendered the tax more resistant to the toxic challenge. New paragraph Sterman further observed that the operably trained cats were significantly more resistant to seizures. This led to investigations of SMR training as a means to reduce seizures in children. Overall, SMR training has emerged as a significant benefit in a wide range of situations, particularly related to pediatric attention and seizures, as well as helping in cases of insomnia. It appears that SMR training is a core mechanism associated with brain and body stability and resistance to stress.

Insert Figure 5-5.

Figure 5-5. Examples of Low Beta or Sensorimotor Rhythm (SMR) at 13 Hz.
In the broad context, given that we have the ability to operatively condition a rhythm associated with the intention to remain still, as well as brain and body resilience, we can understand why SMR training has the clinical importance that has been seen over the last several decades. Rather than addressing behavioral issues, or using approaches that focus on symptoms, neural feedback of SMR, when applied to hyperactivity, gives clinicians a way to fundamentally alter a clients ability to be comfortable with stillness, in contrast to continually seeking stimulation and action.

Low beta

• 
  Distribution: localized by side and lobe
  
• 
  Subjective states: relaxed, focused, integrated
  
• 
  Tasks & behavior: relaxed, attentive
  
• 
  Physiological correlates: inhibited motion (when at sensorimotor cortex)
  
• 
  Effects: relaxed focus, improved attentive ability
  

Beta waves are those most generally associated with conscious, deliberate thought. When present, they indicate brain activation, and cortico-cortical communication. Because the cortico-cortical connections mediating beta tend to be between nearby sites (“shortrange connections”), beta tends to be more localized than lower frequency rhythms.

Beta is one of the more common complements trained in neurofeedback, and it is used as a way to train activation of specific areas. In particular, when beta deficits are evident, and clinical signs include those associated with under activation, beta training can be an effective way to activate the affected regions, and achieve more normalized thoughts and behavior. Beta is also used in peak performance training, in which clients without clinical complaints seek improved mental sharpness or acuity.

Insert Figure 5-6.

s

There is evidence that the gamma rhythm is connected with low-frequency rhythms such as theta, and that there may be a “gating” mechanism. The gamma bursts can appear as individual wavelets that occur at a low frequency rate, and are relatively short. Collura, Don, and Warren (1985, 2004) identified spectral signatures that suggested that bursts of 40 Hz activity at a rate of 6-7 per second in a clairvoyant subject, during successful trials, as distinguished from unsuccessful trials. Freeman, W.J., Burke, B.C. and M.D. Holmes (2003) demonstrated phase resetting of the gamma rhythms, at rates of 7 to 9 per second. Lutz et al (2004) also demonstrated enhanced gamma in experienced meditators.

Figure 5-7. Correlation between infra-slow oscillations and conventional component bands.

Figure 5-7 shows the correlation between infra-slow (.002 - .05) oscillations and conventional component bands. It is evident that, in addition to the expected correlation with delta and theta, that there I a correlation with the variations in the gamma band as well.
Note that the shortness of the gamma bursts accounts for the fact that a wide filter bandwidth is required to track these events. The sidebands created by the modulation of the gamma bursts produces energy as far as 5 or more Hz away from the 40 Hz “carrier” band, so that the actual energy in the signal extends from 35 to 45 Hz, at a minimum.

DC and Slow Cortical Potentials

DC and Slow Cortical Potentials have been discussed in chapter 2 in relation to neurofeedback mechanisms and training. It is sufficient to note here that the physiological mechanisms that generate the EEG operate at frequencies extending all the way down to the 0 Hz or “standing” potential.


DC (0-0.1 Hz)

• 
  Standing potential, 0.0 – 1 Hz
  
• 
  Reflects glial, other mechanisms
  
• 
  Includes sensor offset and drift
  
• 
  May include “injury” potential
  
• 
  Difficult to record, may be unstable
  
• 
  Requires Ag/AgCl sensors
  
• 
  SCP is more useful clinically
  

Slow Cortical Potentials consist of the signals that vary with long time-constants, moving above and below the baseline over periods of seconds. An SCP shift may occur over a period of 1 to 5 seconds, and reflects a change in cortical excitability.

The emerging use of feedback using very low frequencies has been referred to as “Infra-Low Frequency” or “Infra-Slow Fluctuation” potential work. Girten, Benson, and Kimiya (1973) observed such slow oscillations more than 40 years ago, but their use for neurofeedback has been relatively recent. The low-frequency cutoff can be set as low as 0.001 Hz, and frequency adjustments can be made with a resolution as low as 0.0001 Hz. This does not mean that the potentials oscillate with a period of 1/0.001 = 1000 seconds, which would be on the order of 16 minutes. Rather, the filters serve to block out all but the largest transient shifts in the DC baseline. Therefore, if filters are set, for example, with a range of from 0.001 to 0.0015 Hz, the low cutoff serves to bring the signal back to zero baseline with a very long time-constant, and the high cutoff serves to ensure that only significant shifts in the baseline will be passed through the filters.

ILF (or ISF) work is generally done with the addition of inhibits on most if not all of the conventional EEG frequency bands. This changes the nature of the feedback so that there is general reinforcement when the EEG quiets in general, and additional rewards are provided when the DC level of the EEG shifts by a sufficient amount. It is possible to reward either positive or negative shifts separately, or to reward any shift.

Development of ILF/ISF neurofeedback has followed a primarily empirical path, with clinical experience and subjective reporting becoming a driving factor (Othmer 2010; Othmer & Othmer, 2008). As a result the scientific underpinnings and origins of these signals remain somewhat unclear and not without controversy. The factors that can potentially contribute to these fluctuations include both brain and non-brain sources, and influences including variations in skin and sensor properties cannot be discounted.