Introduction to Commonly Used Terms in Electroencephalography


In this chapter, we review the basic terms used in electroencephalography and strategies for communicating EEG findings to others. The creation of a concise description of an EEG tracing is a key part of the art and science of EEG interpretation. The ideal EEG description allows its reader, on the basis of the report alone, to visualize the appearance of the EEG even in the absence of the tracing. The report also includes a section that describes the technique used to record the EEG and a separate clinical interpretation of the results that discusses the potential implications that the findings might have for the patient. The specifics of EEG reporting are discussed in more detail in chapter 8 , the structure and philosophy of the EEG report.

Careful use and understanding of EEG terminology brings specific advantages to both the writers and readers of EEG reports. The most obvious benefit is that strict use of EEG terminology facilitates nonambiguous communication of EEG findings to others. The use of idiosyncratic terms or a personal EEG vocabulary should be avoided because they may not be fully understandable to others. Standard definitions for EEG terms were published in 1974 and more recently in 1999 (see Suggest Readings). Another reason for using a common vocabulary is to aid in the matching of EEG findings (e.g., “spikes” or “intermittent rhythmic slowing”) to the clinical syndromes that may be associated with those findings. This is important in both clinical diagnosis and the conduct of research involving EEG. Finally, using the most precise term possible in a description helps discipline the EEG reader during both the interpretation and report preparation process.

DESCRIPTION OF EEG WAVES

As mentioned, an ideal description of an EEG waveform would allow the reader to visualize or perhaps even produce an accurate drawing of the wave on the basis of the written description alone. If a wave can be drawn two or more fundamentally different ways from a particular written description, the ambiguity in the description provides a clue as to how it could be improved. In general, a complete description of an EEG wave or event includes the following features: location, voltage (amplitude), shape (morphology), frequency, rhythmicity, continuity, and the amount of the wave seen and in which particular clinical states (awake, asleep) it occurs in the record ( Table 3-1 ). For instance, a left temporal theta wave could be described as follows: “Occasional examples of intermittent, medium voltage, sinusoidal 6–7 Hz waves are seen in the left temporal area during drowsiness.”

Table 3-1
Some Wave Parameters with Specific Descriptive Examples of Each
Descriptive Parameter Example
frequency given in cycles per second (cps) or Hertz (Hz) or by frequency range: delta, theta, alpha, or beta
location occipital, frontopolar, generalized, multifocal
morphology spike, sinusoidal waveform
rhythmicity rhythmic delta, irregular theta, semirhythmic slow waves
amplitude high voltage slowing, a 75 μV sharp wave
continuity continuous slowing, intermittent theta, periodic sharp wave

FREQUENCY

Frequencies are quoted either in cycles per second (cps) or in Hertz (Hz). The terms cps and Hz are synonymous. For historical reasons, each frequency band in the EEG is named using one of four Greek letters. By convention, the names of the letters are written out: delta, theta, alpha, and beta; the symbols for these Greek letters are not used. The terms delta, theta, alpha, and beta are used as a shorthand to refer to rhythms in specific frequency bands and are defined as follows:

delta 0 to <4 Hz
theta 4 to <8 Hz
alpha 8 to 13 Hz
beta >13 Hz

These frequency ranges are straightforward, but a few observations are worth noting. Note that according to the formal definition, the alpha range is the only “all-inclusive” range, and includes both boundary frequencies (8 and 13 Hz). The other frequencies include the boundary frequency on one side of the range but not the boundary frequency on the other side of the range. For example, the theta range, for which the boundary frequencies are 4 and 8 Hz, includes 4 Hz but does not include 8 Hz (which belongs to the alpha range), although it does include 7.9 Hz.

Very Fast and Very Slow Activity

Although the list above implies that beta-range frequencies are unbounded on the high side, in reality EEG activity recorded at the scalp with routine EEG instruments rarely exceeds 30 to 40 Hz, and even signals from cortically implanted electrodes rarely exceed 50 Hz. High-frequency filtering techniques and other inherent equipment limitations make it difficult to see faster frequencies in routine recordings, even should they be there to record. The assumption behind using high filters is that little or no cerebral activity exceeds 30 to 40 Hz so that any signals above this range likely represent electrical noise or “artifact” from muscle, electrical interference, or other sources—hence, the strategy to filter out most activity above this range. In the early days of EEG, frequencies above 30 Hz were said to be in the gamma range, but the term (and the concept) was later discouraged. More recently, there has been renewed research interest in “gamma activity,” although such very fast activity has yet to establish its place in the mainstream of conventional EEG interpretation.

Similarly, although the lower bound of the delta frequency range is “zero Hz,” it was initially felt that there was little cerebral activity below 0.5 Hz. Newer techniques have successfully recorded very slow activity, which are termed DC potentials because they resemble shifts of the baseline (direct currents) rather than oscillations. Again, because standard recording techniques usually exclude most such very slow activity, frequencies much below 0.5 Hz are usually not reliably observed in conventional EEG recordings. Despite the possible existence of some very slow potentials in the EEG, in some situations, the great majority of activity below 0.5 Hz usually represents motion or electrical artifact rather than electrocerebral activity. Conventional filtering techniques attempt to remove these large baseline shifts from the recording because they usually represent artifact and may render the EEG tracing unreadable (see Chapter 7 , Filters).

The term slow activity refers to any waves for which the frequency falls below the alpha range (i.e., delta and theta activity—activity below 8 Hz). Likewise, the term fast activity refers to activity for which the frequency lies above the alpha range (i.e., beta activity–activity above 13 Hz).

Frequency, Wavelength, and the Relationship of Frequency to Wavelength

The simplest way to assess the frequency of a wave is to count how many times it cycles within 1 second. If a wave cycles four times (i.e., manifests four “peaks” and four “troughs”) in 1 second, then it is said to be a 4 cps or 4 Hz wave. At times, however, a wave cannot be counted for a full second. Luckily, it is possible to determine the frequency of a wave not just by counting how many times it cycles within a second but simply by measuring its duration or wavelength (measured from peak to peak or from trough to trough). The result of a straightforward calculation shows that if a wave cycles five times per second then each wave’s duration must be one fifth of a second or 0.2 seconds. Similarly, if a wave cycles eight times per second, each wave will last one eighth of a second or 0.125 seconds. This yields the easy to remember relationship below, where λ gives the wavelength and f is the wave"s frequency:


λ = 1 f

The rearrangement of the above relationship is also useful and allows estimation of a wave’s frequency by measuring its duration (wavelength) on the page, even if only a single wave is available for analysis:


f = 1 λ

Both of these simple relationships always hold true. This allows the EEG reader to measure the duration (or wavelength) of any particular wave seen in the EEG (in seconds) and, by taking the reciprocal, to determine its frequency (in cps or Hz). Because standard EEG recordings show pages with vertical divisions of 1 second, the wavelength of a given wave can usually be easily measured or visually estimated by determining what fraction of a second it occupies. On some occasions, it is necessary to measure the wave directly, either with a ruler specifically designed for the purpose or with a digital time cursor in the case of digital recordings. The reciprocal of the wavelength measurement is calculated, yielding the frequency.

A Note on Units

The foregoing examples assume that wavelengths are stated in seconds, but it is common practice to quote wave durations in milliseconds (msec). Rather than stating that a wave lasts 0.2 seconds, it is often said that the wavelength is 200 msec. Using milliseconds is perfectly satisfactory, except when it comes to using the formula cited earlier to calculate frequency. The reciprocal of 200 is 0.005, which clearly is not the correct frequency of a wave of 200 msec duration. This is because inserting a wavelength in milliseconds into the formula yields the result in the undesirable unit of “cycles per millisecond.” To yield the frequency in cycles per second, or Hertz, the wavelength used in the formula must always be stated in seconds (1 cycle/0.2 sec = 5 cps). Examples of how the frequency of alpha, delta, theta, and beta activity are assessed are shown in Figures 3-1 through 3-4 .

Figure 3-1, The frequency of the wave in the blue rectangle is most easily determined by counting the number of wave peaks or troughs seen in 1 second. These waves are rhythmic, fairly sinusoidal in shape, and vary in amplitude. In this example, 10 wave peaks are counted in a 1-second time period, indicating that this is an example of fairly regular 10-Hz alpha activity. This particular tracing represents an example of the posterior rhythm of a 7-year-old boy.

Figure 3-2, The frequency of the nonrepeating slow wave denoted by the penstroke in the top channel is most accurately assessed by measuring its wavelength. The wave in this example is 0.67 seconds in duration, measured from wave trough to wave trough. The simple relationship of wavelength to frequency described in the text in which frequency is the reciprocal of wavelength (1 cycle / 0.67 seconds = 1.5 cycles per second) tells us that this is a single 1.5-Hz delta wave. Note that the wave’s frequency can be determined even though it stands alone and does not repeat, and also that the shape of this slow wave can be seen despite the superimposed fast activity. If a wave of this size did repeat, it can be visually estimated that 1.5 such waves would fit into 1 second.

Figure 3-3, The frequency of irregular theta waves can be assessed using the same technique of measuring wavelength that was used for the previous figure. In the second channel of this figure, several semirhythmic theta waves can be seen with the durations shown in the figure. Although the waves indicated by the arrows do suggest some rhythmicity, each varies to some extent in wavelength and cadence. By calculating their reciprocals, the measured wavelengths of 0.19, 0.14, and 0.16 seconds correspond to frequencies of 5.3, 7.1, and 6.3 Hz, respectively, all of which are in the theta range (4–8 Hz).

Figure 3-4, The fast activity seen in the shaded area of the first trace is too fast to make an accurate measurement of wavelength and would be cumbersome to count for a full second. Here, in the half-second interval denoted by the shaded rectangle, 20 waves are counted, implying a frequency of 40 Hz. Given the high frequency of these waves and the fact that they were recorded over the frontalis muscle, these may represent an example of electrical muscle artifact.

Waves of Mixed Frequency and the Fourier Theorem

The foregoing discussion deals with descriptions of simple waves of a single frequency. However, after a brief look at actual EEG recordings, it is clear that few, if any, of the waveforms seen on the EEG page resemble the pure sine waves seen in math textbooks. Indeed, as a rule, EEG waves represent a mixture of waves of different frequencies and amplitudes. Part of the reason for this mixture of frequencies is that the waveform recorded at the scalp often consists of a mixture of the products of different wave generators from various locations in the brain. In addition, some wave-generating circuits in the brain produce complex, nonsinusoidal waveforms.

One important skill in EEG interpretation is the ability to “deconstruct” EEG waves visually into their component frequencies. The idea that a repetitive complex wave can always be broken down into simpler, fundamental waves was proved by Joseph Fourier in a mathematical theorem called the Fourier theorem. This theorem states, in simplified terms, that any waveform, no matter how simple or complex, can be exactly described by the sum of a sufficient number of simple sine waves. Although a comprehensive discussion of the Fourier theorem is beyond the scope of this text and not necessary to EEG interpretation, it reminds us that even the most complex EEG waves may be thought of as the sum of some number of fundamental sine waves. For example, the apparent complex wave shown in Figure 3-5 actually represents the combination of a 1-Hz wave, a 2.4-Hz wave, an 11.7-Hz wave, and a 34-Hz wave, each of a different amplitude. It is not necessarily the electroencephalographer’s goal to identify all of the component parts of any particular EEG wave, but it is useful to be able to identify the fundamental wave and the main superimposed waves of an EEG signal. Figures 3-6 through 3-9 illustrate the component waves that make up the complex wave of this example.

Figure 3-5, The complex waveform shown in this figure represents the summation of four separate sine waves with frequencies of 1 Hz, 2.4 Hz, 11.7 Hz, and 34 Hz wave, each of a different amplitude. The four individual contributing components of the wave are shown in the following figures. A 1-second scale is shown.

Figure 3-6, The original 1-Hz component of the mixed wave is superimposed on it. Some of the gradual “up and down” shape of the complex wave can be seen to be explained by this underlying 1-Hz slow-wave component, but the contribution of this slow component is difficult to visualize.

Figure 3-7, The same wave as Figure 3-6 with the 2.4-Hz component isolated and superimposed. Of all the component waves (with the possible exception of the fastest 34-Hz wave), this 2.4-Hz component frequency can be most easily identified in the mixed wave. Along with the 1-Hz wave, the 2.4-Hz wave lends some of the “roller-coaster” shape to the final wave.

Figure 3-8, The 11.7-Hz wave component is shown above the mixed wave for the sake of clarity. This specific component is somewhat difficult for the eye to sort out from the mixture of frequencies in the complex wave, but it can be identified.

Figure 3-9, The 34-Hz component of the initial wave is probably the easiest to identify. It is seen to “ride” atop the lower frequency waves.

For those interested in the basic mathematics that describe these waves, a review of the formula for a simple sine wave function helps us to recall some of the features of a wave that can be described thus:


f(x) = A sin (bx + ).

In this formula, varying the coefficient A higher and lower will change the height or amplitude of the resulting wave. Varying b will change the frequency of the wave. Finally, changing the value of will shift the wave to the left or right on the x axis, which is also referred to as shifting the phase of the wave. It is not necessary to know this formula to interpret EEGs, but it is useful to keep the formula and its coefficients in mind when learning to separate out and describe the component parts of wave mixtures.

Although addition of one sine wave to another can be described with mathematical formulas, it is more useful to become accustomed to the appearance of how sine waves add visually as opposed to mathematically.

Figure 3-10 shows the result of superimposing or adding a lower voltage 10-Hz wave onto a higher voltage 1-Hz wave. Note that the lower voltage fast activity appears to “ride” the hills and valleys of the slower wave. The key is that both individual waves can still be visually appreciated even though they are mixed together.

Figure 3-10, The result of the addition of a 1-Hz wave to a lower voltage 10-Hz wave is shown. The top trace shows a 1-Hz sine wave at an amplitude of 100 μV, and the middle trace shows a 10-Hz wave at 20 μV. When the 1-Hz and 10-Hz waves are added together, the bottom wave results, showing the 10-Hz wave “riding” on the fundamental 1-Hz frequency.

When two waves are mixed and one has a frequency that is a multiple of the other, the higher frequency multiple is called a harmonic, and the lower frequency wave may be called the fundamental frequency. (Occasionally, the lower frequency wave may be called the subharmonic .) For instance, if there is a 5-Hz fundamental frequency, it is not uncommon to see a superimposed 10-Hz harmonic. When a wave is mixed together with its harmonic, the resulting wave often has a particular appearance of regular notching.

Figure 3-11 shows an example of adding a 2-Hz wave to its 4-Hz harmonic. Note how the shoulder of the slower wave is regularly notched by the faster wave in this idealized example. The notching may appear on the upslope, downslope, peak, or trough of the slower wave, depending on how the phase of one wave is lined up with the phase of the other when the two are added together. This pattern is important to recognize because the notching patterns created by these harmonics can occasionally be mistaken for spike-wave discharges.

Figure 3-11, This figure illustrates the addition of a fundamental wave to its harmonic. The top trace shows the fundamental 2-Hz wave at 25 μV, and the middle trace shows a 4-Hz sine wave of the same amplitude. When a wave is added to another wave that is a multiple of the first wave’s frequency (i.e., a harmonic frequency), the distinctive tracing seen in the bottom trace results. Note the regular “notching” pattern of the smaller wavelet riding on the larger wave. Because the two waves partially reinforce, the summation wave has an amplitude of 40 μV in this example. The degree to which the two waves reinforce and the particular shape that results is partly a function of how the two waves are shifted in the left–right axis with regard to one another (phase shift).

Figure 3-12 shows how the appearance of the superimposition of the same 2-Hz and 4-Hz waveforms can differ depending on how the phase of one is shifted to the right or left in comparison to the other.

Figure 3-12, The two waves shown here represent the summation of the exact same two component waves (top and middle waves) from the previous figure, except that the phase of the second wave is shifted to the left or right a slightly different amount compared with the first wave before the two are added together. Such combinations of fundamental waves with their harmonics do occur in cortical circuits and should not be mistaken for “spike wave.”

Figure 3-13 shows the result of adding a fundamental 2-Hz wave (top trace) to a 7-Hz wave of slightly varying amplitude (middle trace). In reality, the important skill is not necessarily to be able to predict what the result will be of adding two particular waves together but, rather, the reverse procedure: to be able to look at the bottom trace in Figure 3-13 and to be able to visually reconstruct the distinct 2-Hz and 7-Hz rhythms contained within the more complex waveform. Although the examples shown in these figures are of idealized sine waves, the same visual skills apply to the “deconstruction” of actual EEG waveforms. Of course, the idealized waves seen in these examples do not occur in such pure forms in complex biological systems such as the brain, which are more complex and subject to more variation.

Figure 3-13, The top trace shows a 2-Hz wave of constant amplitude, and the middle trace shows a 7-Hz wave of slightly varying amplitude. The bottom wave shows the result of adding the top wave to the middle wave. In this example, the 7-Hz activity is most readily appreciated in the combined wave, whereas the 2-Hz component is less obvious, although it makes a definite contribution to the final wave’s shape.

Figures 3-14 and 3-15 show examples of other ways that pure sine waves can vary, such as in frequency and amplitude, and represent somewhat more realistic approximations of real-life EEG waves.

Figure 3-14, The sine wave shown has a constant amplitude (height) but a 15% variation in frequency. Note that the width of each wave increases and decreases perceptibly. Such slight variations in frequency are common in physiologic EEG waves.

Figure 3-15, This sine wave has the same average frequency as the previous wave, but in this example, the frequency is held constant, and there is a 10% variation in the wave’s amplitude.

Figure 3-16 shows a close-up of a recorded EEG wave that contains mixed frequencies.

Figure 3-16, An EEG signal acquired from a patient shows a complex, “real-world” mixture of delta, theta, and faster rhythms. Although some sinusoidal elements can be appreciated, many of the waveforms are irregular.

LOCATION

The location of an EEG event is usually best described in terms of the electrode(s) it involves. The proper placement of the EEG electrodes is the job of the EEG technologist who produces the recordings. The schema used for electrode naming and placement is called the International 10-20 system, formalized by in 1958. Two 10-20 electrode systems are in use today, the system using the original electrode nomenclature and a modified 10-20 system in which some of the electrode positions have been renamed. Luckily, the great majority of the electrodes have the same names in both systems. Unfortunately, to date, neither system has established primacy over the other, and use of both systems is still widespread. At the risk of some confusion, examples of both systems are used in this text so that readers will gain some familiarity with each. Some feel the newer system is more logical, whereas others do not feel it offers enough advantages to overturn decades of tradition in nomenclature. Because no single electrode name is used to indicate different positions in the two systems, the two conventions can coexist unambiguously side by side. Although it may be tempting for new electroencephalographers to memorize only the newer system because it may seem to represent the “wave of the future,” it is probably best first to memorize the system in use in one’s own laboratory but to have at least a passing acquaintance with the other if called on to read records or reports that come from other laboratories.

The term 10-20 is based on the general strategy of measuring the distance between two fixed anatomical points, such as the nasion (the point where the bridge of the nose meets the forehead) and the inion (the prominent point on the occiput), and then placing electrodes at 10% or 20% intervals along that line. This 10-20 system represented an improvement over previous electrode placement systems, some of which relied on absolute measurements (instead of percentages) but failed miserably when heads of different sizes were studied—imagine the effect of using the same system of absolute measurements on the heads of adults or patients with hydrocephalus and the smaller heads of newborns or even premature infants. The general plan for the placement of the 21 primary electrodes of the original 10-20 system is shown in Figure 3-17 . The electrode nomenclature for the modified 10-20 system are shown in Figure 3-18 . The four electrode positions for which names have changed are shown shaded.

Figure 3-17, The electrode nomenclature for the original 10-20 electrode system is shown. This original naming system is still in use in many EEG laboratories.

Figure 3-18, Primary electrode names for the newer “modified” 10-20 electrode system are shown. Note that the majority of electrode names are unchanged; the shaded electrodes represent the name modifications proposed by this new system. In addition to those electrodes shown, this system also establishes names for electrodes in intermediate positions that are not routinely used in standard EEG recordings (see Figure 3-21 ). This modified 10-20 system has not yet been universally adopted, however.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here