Introduction to Reading EEG's


The first encounter with an EEG tracing is somewhat perplexing to the physician and the technician alike as they are likely to be awestruck by the complexity of the record. Often, the initial impression is that the task is too complex to learn. The physician familiar with reading ECGs soon realizes that there is little in common between the analysis of the repetitive complexes of the ECG and the ever changing waveforms of the EEG. At this stage, the prospective electroencephalographer is tempted to exclaim: "This is all Greek and Latin to me" Indeed, learning to read EEGs is not unlike learning to read a foreign language. To read a new language, needless to say, one needs first to learn the alphabet. The alphabet of the EEG consists of the various frequencies and waveforms that comprise the tracing. Just as the letters of the alphabet are combined in different permutations and combinations to form words and then sentences, so the EEG tracings are made up of combinations of waveforms of different frequencies and morphology. To carry the analogy further, it is not enough to be able just to read the words and sentences; one needs to understand quickly the meaning of what is written. In the same way, EEG reading involves analyzing the waveforms and deducing their significance. With experience, one uses a speed reading technique in which a whole page is rapidly scanned for evidences of normal and abnormal phenomena. How successfully this is done depends to a large extent on developing pattern-recognition skills.

Learning to Read:

How does one learn to read EEGs? Like any other branch of medicine this involves a continuous process of learning for many months or sometimes even years. Often the initial learning is accomplished through observing an experienced electroencephalographer read EEGs. The next step involves reading under supervision; having seen how an experienced electroencephalographer interprets a record, and having gathered essential information regarding normal and abnormal patterns, the trainee interprets records in the presence of his or her instructor. Ideally, the instructor should regularly quiz the trainee on the various waveforms and artifacts in the tracings, and the trainee should complement this by seeking answers to the questions. Without at least an elementary knowledge of the basic principles of electricity, neurophysiology, and the technique of recording, it is impossible to learn to read EEGs properly. One needs to know what calibration means, how the various frequency filters work, how various artifacts are identified, and how neurologic disorders produce alterations in electrical activity of the brain. The prospective electroencephalographer also needs to have a thorough working knowledge of the various montages used. All these topics are taken up in considerable detail in this text. Numerous EEGs need to be seen before one becomes familiar with the wide range of normal variations in different age groups and physiological states. The task becomes even more difficult when tracings of neonates and premature infants have to be interpreted. It may take two or three years of experience in reading before one has acquired reasonable expertise.

Terminology:

It is essential to use standard terminology in describing the EEG. The International Federation of Societies for Electroencephalography and Clinical Neurophysiology has proposed definitions for the various terms used in EEG to facilitate communication between different electroencephalographers. In this section, we list the definitions of some of the terms that are most commonly used in EEG reading.

Background Activity:

This term denotes the general setting in which changes in frequency, amplitude, or morphology appear. Although the alpha rhythm may be the background activity in the tracings from the posterior regions, it is important to note that the term background activity is not synonymous with alpha rhythm; thus, over the frontal area, the activity may be mostly in the beta frequency band. The background activity may not always be a normal pattern; the term can also refer to abnormal patterns.

Both the background activity and the changes that appear in the features of the tracing are described in terms of frequency, amplitude, wave shape, symmetry synchrony, location, continuity, and reactivity. It is important to understand the meaning of each of these terms to give a proper description of the EEG.

Frequency:

This term refers to the rate at which a particular waveform repeats; it is usually used in the context of rhythmic activity (repeating with regularity)- Depending on the frequency the activity is classified as delta (less than 4 Hz), theta (4 to 8 Hz), alpha (8 to 13 Hz), or beta (more than 13 Hz) activity Although these terms are ideally restricted to rhythmic activity, they are also often used to describe nonrhythmic or random activity; in this case the frequency of a particular wave will he ascertained by taking the inverse of its duration.

The frequency bands are used to describe the activity irrespective of where it occurs. But the term alpha rhythm is more specifically used to denote the 8 to 13 Hz rhythm occurring during wakefulness over the posterior region of the head; it occurs generally with maximum voltage over the occipital area, is best seen with the patient's eyes closed and under conditions of physical relaxation and relative mental inactivity, and is blocked or attenuated by attention, especially visual attention and mental effort Sometimes the terms fast and slow activity are used to denote a dominant frequency above or below the alpha band. The term monorhythmic is often used when the particular activity shows rhythmic components of a single frequency. When there are multiple frequencies the term polyrhythmic is used. The term periodic applies to EEG waves or complexes recurring at approximately regular intervals; usually the intervals vary from one to several seconds.

Amplitude:

This is expressed in terms of voltage in microvolts based on a peak-to-peak measurement. One needs to know the sensitivity at which a recording was made to determine this. For this purpose, the reader consults the routine calibration at the beginning or end (paper records only) of the record. The amplitude will vary depending on the technique of recording, the bipolar montages with short interelectrode distances giving a smaller amplitude than the referential montages with larger interelectrode distances. Ideally, amplitude should be described in terms of the actual voltage; however, the terms low, medium, and high amplitude are often used. The term low is used when the amplitude is under 20 uV medium when it falls in the range of 20-50 uV, and high for more than 50 uV. The use of these terms is discouraged owing to lack of uniform criteria. Attenuation and blocking are terms used when there is a reduction in the amplitude of EEG activity, usually in response to some stimulus. The classic example is attenuation of the alpha rhythm in response to eye opening. The term suppression is used when little or no electrocerebral activity can be discerned in a tracing. Paroxysmal activity is a term denoting activity of much higher amplitude than the background that occurs with sudden onset and offset. It need not necessarily denote an abnormal activity.

Wave Shape or Morphology:

Electroencephalographic activity is essentially a mixture of waves of multiple frequencies. The appearance of the waveforms depends on the component frequencies, their relative voltages and phase relationships, and, of course, upon the frequency filters used. The waveforms are also continuously fluctuating in response to stimuli and depend on the state of the patient. Several descriptive terms may be used in this context. A transient is an isolated wave that stands out from the background activity; if it has a sharply pointed peak and the duration is less than 70 ms (less than 2 mm at the paper speed of 30 mm/s), it is called a spike; when the duration is between 70 to 200 ms, it is called a sharp wave. The term complex is used when two or more waves occur together and repeat at consistent intervals; examples are spike and wave complexes and sharp and slow-wave complexes. An activity is described as monomorphic when the morphology of subsequent waveforms is similar whereas the term polymorphic is used when they are of dissimilar morphology (--a mix of frequencies). The description should also include the number of phases. Thus, a wave may be monophasic (positive or negative) or diphasic (positive and negative), triphasic or polyphasic.

Symmetry:

In general, symmetry refers to the occurrence of approximately equal amplitude, frequency, and form of EEG activities over homologous areas on opposite sides of the head.

Synchrony:

This term refers to the simultaneous appearance of morphologically identical waveforms in areas on the same side or opposite sides of the head.

Location:

Several different terms are used. Focal or localized are terms used when a particular activity is confined to one particular region of the head. For example, an activity may be localized to frontal, temporal, parietal, or occipital areas. The term generalized is used when activity is not limited to one region but occurs over a wide area. An activity is said to be lateralized when it is present on one side only.

Continuity:

An activity may be described as continuous or intermittent, depending on the percentage of time it is present. Thus, an activity is called continuous when it occurs without interruption for prolonged periods of time and discontinuous or intermittent when it appears only from time to time.

Reactivity:

The term refers to alterations in the amplitude and waveform of activity in response to a stimulus. An example is the attenuation of alpha activity on eye opening. There are other terms that are used in various special circumstances. These will be taken up in relevant chapters.

Describing the EEG:

An adequate and accurate description of the EEG record is important for several reasons. When a clinician would like to compare the EEGs from different laboratories, or when a recent EEG needs to be compared with the findings of an old record that is no longer available, the written description of the records is essential. It can now be said that if the description is good, the electroencephalographer can picture the actual EEG record in his mind's eye. In making the description as objective and as accurate as possible, it is important to break down the complex tracing in terms of frequency, voltage, reactivity, synchrony, and distribution.

The various activities occurring in different states of consciousness, namely wakefulness, drowsiness, and sleep, should be described clearly. One may start with a description of the background activity, which often tends to be the alpha rhythm in the awake subject. Mention should be made about its frequency, amplitude, location, symmetry, and reactivity to eye opening. Next, the features of other rhythmic activities present should be described in similar terms; for example, beta activity in the frontal or central areas.

Intermittent activity should be described in similar terms, including the location and synchrony. The presence of various phenomena such as V (vertex) waves and sleep spindles should be clearly described. If sharp waves, spikes, or other intermittent activity is present, it should be described in terms of location, polarity; and amplitude; how the activity is affected by changes in state should also be noted.

The effect of activation procedures such as hyperventilation and photic stimulation should be described. In the case of hyperventilation, the way in which this procedure influences the background activity and whether it induces other changes should be mentioned. For photic stimulation, the reader should mention whether there is a driving response and, if so, whether it is symmetrical and at what frequency or frequencies it occurs. If there are any specific responses like photoparoxysmal or photomyogenic responses, they should all be called attention to in the description.

Interpreting the EEG:

The basic question to ask after completing a visual analysis of the EEG is whether the findings are consistent with the accepted norms for the age and state of the patient. This means that the reader should have a thorough knowledge of the normal variations of EEG patterns in relation to age and state of the patient. Such a judgment is possible only after the reader has seen numerous EEGs and has formed an impression in his or her own mind about normal patterns.

If the EEG is normal, the interpretation ends with a statement to that effect. If an abnormality is present, the next phase of interpretation involves categorization of the abnormality in more specific terms Categorization should be precise Some examples are: focal epileptiform abnormality, generalized epileptiform abnormality, focal slowing, generalized slowing, intermittent rhythmic delta activity polymorphic delta activity, asymmetric alpha rhythm, asymmetric photic driving, asymmetry of sleep spindles or vertex waves. If an abnormality is found to be localized, one needs to specify what area of the brain underlies the abnormality.

The next step in the interpretation is to suggest what kind of changes may be happening in the brain that could account for, or be compatible with, the abnormal EEG pattern. This entails a clear knowledge of the relationship between the various EEG abnormalities and the various disorders that affect the brain. One of the major problems in this aspect of interpretation is that many different types of disorders affecting the brain can give rise to the same type of EEG abnormality so that very often only general comments can be made. These comments may be like the following: "This pattern is suggestive of a diffuse encephalopathy," or "the finding is compatible with a focal structural lesion," or "this finding is suggestive of a focal seizure disorder' or "the abnormality is compatible with a seizure disorder of the generalized type" etc Following the technical interpretation, it is always useful to provide a clinical correlation on the basis of the patient's clinical history. Often it may be of value to say whether the EEG abnormality seen is consistent with the clinical diagnosis. Table 13.1 gives a brief synopsis of EEG reading.

More on Artifacts -Physiological Artifacts:

In general, four different varieties of artifacts are encountered in EEG work: environmental, instrumental, electrode and physiological. The EEG technician and the physician reading EEGs both need to be familiar with all of them. Needless to say, artifacts in the EEG should he eliminated whenever possible or kept to a minimum. To achieve this goal requires close collaboration between the technician and the physician reading the EEGs.

Environmental, instrumental, and electrode artifacts have been discussed in various earlier chapters, principally the chapters on recording electrodes and troubleshooting. In this section we consider the topic of physiological artifacts.

There are four major sources of physiological artifacts, namely, the heart, the muscles of the head and neck, the eyes, and the skin. The person reading the EEG must be able to recognize these artifacts if they occur in the recording. He or she needs to learn to "read through" the artifacts whenever possible. In this respect, the reader acts like a filter in much the same way as the frequency filters function on the EEG machine. We will take up each of these artifacts in turn. Appendix 6 Shows some EEG tracings containing these artifacts.

ECG Artifacts:

Because of their regularity and distinctive morphology, ECC artifacts are usually the easiest to recognize in the EEG tracings. Any uncertainty about whether or not an artifact is an ECG is easily resolved by actually recording an ECG on one channel along with the EEGs. This, of course, needs to be done at the time that the routine EEG is taken. Obviously, a close, harmonious working relationship between the EEG technician and the physician reading the EEGs is essential in resolving such problems.

EMG Artifacts:

These are the most common artifacts seen in EEG recordings. Muscle potentials originate mostly from the muscles of the head: the frontalis muscle the masseters, sternomastoids, and temporal muscles are common sources. Electromyographic artifacts are recognized mainly by their distinctive morphology; the spikes themselves are very sharp and of short duration. In this regard, however, particular attention needs to he given to the settings of the high-frequency filters on the EEG machine. Settings lower than 70 Hz will result in the spikes being rounded off so that they may easily be confused with brain electrical activity in the beta band, or brain-origin spikes.

As is the case with ECG artifacts, the EEG technician plays an essential role in the detection of EMG artifacts. While taking the EEG, he/she closely observes the patient, notes various movements, and records their occurrence directly on the EEG record. These notations are especially helpful in recognizing EMG artifacts when the tracings are read and interpreted. Moreover, the technician's efforts in establishing rapport with the patient, and in helping him or her to relax, go a long way toward reducing many EMG artifacts.

Eye-Movement Artifacts:

Electrically, the eyes behave very much like batteries rotating in their sockets. This means that electrodes located in the anterior regions of the head can very readily pick tip changes in voltage that are correlated with eye movements. The artifacts are quite distinctive and are easily recognized after some experience. To assist recognition, it is helpful in the course of training to record samples of eye movements -up, down, right, and left-using some of the common montages. In this way, technician and reader alike can become more familiar with the spatial distribution of these artifacts.

If there is any question concerning whether or not activity recorded in the EEG is due to eve movement, an electrode should be attached to the patient's cheek directly below the eye. This electrode is connected to grid I of one channel; grid 2 of the same channel is connected to the ipsilateral earlobe. The recording from this derivation is compared with the recording from an anteriorly placed derivation in a longitudinal bipolar chain, eg., Fp,-F7 or Fps-F8. Since the eye falls between these two derivations, eye movements will appear in the two channels as out-of-phase deflections or mirror-image signals. If the deflections in the two channels are in phase, they are not eye movements but may be of cerebral origin.

Galvanic-Skin Artifacts:

Like all living tissue, the skin is electrically active. Changes in electrical activity of the skin are usually associated with sweating, although some changes in voltage may be observed between two points on the skin in response to stimulation (the so-called Tarchanoff effect). Generally speaking, changes in skin potential associated with sweating or stimulation are very slow changes. They occur mostly at the very low end of the delta frequency band, making them easy to recognize in the EEG tracings. Sometimes these artifacts can become quite large When this happens, they usually can be reduced considerably by adjusting the low frequency filters from the standard 0.3-Hz setting to the 5.0-Hz cutoff point.

Writing the EEG Report:

A major complaint sometimes made about the EEG report by some referring physicians (other than neurologists) is that it makes little sense and often does not help in the diagnosis or management of their patients. To some, the report may even seem misleading. For this reason it is important that the report be constructed in two parts: one part deals with the actual description of the EEG findings and their interpretation; the other part contains a clinical correlation that renders the report meaningful to the referring physician.

It is important to begin the report with a brief history and the clinical findings to date (usually available from the physician requesting the EEG and/or from the technologist's worksheet). It is also helpful to mention what the referring physician hopes to find out from the EEG, if it is explicit. The next paragraph should provide descriptive details regarding the testing situation. These should include whether the test was done at the bedside or in the intensive care unit and whether any modifications were made in the electrode connections, as, for example, using a reduced array in a neonate The use of special electrodes like nasopharyngeal or sphenoidal leads should be noted here. Also mention whether the patient was sleep deprived, and whether any medication was given before or during the test and, if so, what kind and how much.

Next comes the section that describes the EEG and the state of the patient. This portion of the report should be purely descriptive and should not contain any interpretive statements such as normal or abnormal. Following this is the paragraph setting forth the impression. Is the EEG normal or abnormal, and if abnormal, what kind of abnormality was seen?

The last paragraph should attempt to correlate the EEG findings with the clinical picture. Thus, for example, in the case of a seizure disorder in which the EEG is normal, it should be mentioned that the EEC does not support the diagnosis of seizure disorder. But it also may be pointed out that a normal EEG does not necessarily rule out a seizure disorder. In this context, one may suggest further studies such as a sleep-deprived EEG or a repeat EEG using special electrode. If the EEG in a patient suspected of having a metabolic encephalopathy shows diffuse slowing or frontal intermittent rhythmic delta activity (FIRDA), one may state that this finding is consistent with the clinical diagnosis.

Michael Young --- April 2004
with special thanks to Frank Duffy and Ernst Neidermeyer
- this work is a derivative from their published material -