The Abnormal EEG

Chapter 9 The Abnormal EEG


An EEG is considered abnormal if it has findings known to be associated with a pathologic or disease state. As discussed in Chapter 8, “The Structure and Philosophy of the EEG Report,” this distinction is designed to prevent the interpreter from calling an EEG abnormal simply because it includes a finding that “looks unusual” or is uncommon, because uncommon findings may not be abnormal.


When deciding which types of studies should be labeled abnormal, it is useful to consider the referring physician’s purpose in ordering an EEG. Ideally, the clinician is confronted with a patient with a particular clinical picture and has formed a list of possible diagnoses, also called the “differential diagnosis,” that might explain the patient’s findings. If the result of an EEG could increase or decrease the chances that one or more entities from the differential diagnosis list is the correct diagnosis, then it may be reasonable to obtain the test. However, if the EEG is not likely to have an impact on the probability of any of the diagnoses on the list, then the test is probably not indicated. It is usually not a good idea to obtain an EEG just “to see what it looks like” without a specific question in mind that the EEG could potentially answer. A related concept is that an EEG is indicated if there is some likelihood that the result will affect the patient’s treatment. Simply reporting that the EEG looks “odd” or “unusual” will not likely be helpful to the referring physician. Rather, an EEG should only be considered abnormal if it contains a finding that has some association with a disease or an abnormal state.


The majority of EEG abnormalities do not specifically lead to a single diagnosis. Only a small minority of EEG abnormalities are associated with a short enough list of disease entities that they can be considered “specific” for one diagnosis or another. The majority of EEG findings are diagnostically nonspecific and are associated with a list of disease states that is lengthy and diverse, so much so that specific diagnoses usually cannot be suggested based on the EEG result. An example of a dramatically abnormal but nonspecific EEG abnormality is generalized delta slowing, a finding that is associated with so many types of abnormal states (e.g., coma, post-seizure state, meningitis, anesthesia) that the clinical implications of the finding can only be stated in the broadest terms in the EEG report. The utility of the EEG comes in combining the EEG result with the patient’s history and other findings to narrow the differential diagnosis.


The referring physician is usually best placed to put the EEG findings and the clinical story together to arrive at a clinical conclusion. Without the clinical history, the EEG is a considerably less powerful tool. Even when a clinical history has been submitted with the EEG request, the reader of the EEG should hesitate before suggesting specific clinical diagnoses. The submitted history may be incomplete and often lacks details such as the physical examination or certain laboratory or imaging findings. When reading the submitted history, the electroencephalographer usually cannot know whether that history is complete. For these reasons, the interpreting physician should resist the temptation to make specific diagnoses based on the EEG results. Rather, a list of diagnoses that have been associated with the observed abnormalities, a sort of “EEG differential diagnosis,” should be given at the end of the report, if feasible. The discussion given in the final clinical interpretation section of the report should take into account the clinical history provided, but the EEG differential diagnosis offered should not be limited by that history.


EEG abnormalities can be categorized in a variety of ways. Abnormalities may fall into the categories of 1) abnormal expressions of normally occurring rhythms (e.g., asymmetries of normal rhythms), 2) inherently abnormal rhythms (e.g., “slow” delta and theta rhythms in an adult who is awake), 3) certain repetitive or periodic patterns (e.g., burst-suppression patterns), 4) epileptiform abnormalities (spikes, sharp waves, etc.), and 5) abnormal “super-architecture” (e.g., abnormal sleep state cycling). These various abnormality types are discussed in the following sections. The first two abnormality families, abnormalities of normal rhythms and abnormal rhythms, are discussed by frequency range, starting with slow activity. Basic epileptiform abnormalities are discussed in this chapter. Various epilepsy syndromes and the EEG findings associated with them are discussed in Chapter 10, “The EEG in Epilepsy.”



ABNORMALITIES OF SLOW-WAVE (DELTA AND THETA) ACTIVITY



Descriptive Parameters for Slow Waves


Slow-wave abnormalities can be defined in multiple domains. The most obvious descriptor of a slow wave is its location. A slow wave may occur focally, such as in the left anterior temporal area or the right occipital area. Slowing may occur in broader regions, such as in the right posterior quadrant, in “anterior brain regions,” or over a whole hemisphere (see Figures 9-1 and 9-2). Such regional slowing can be considered a subset of focal slowing. Finally, slowing may also be diffuse or generalized.


image

Figure 9-1 This page of waking EEG shows slowing in the delta range occurring predominantly over the left hemisphere. Compare the left hemisphere channels (shaded) to the right hemisphere (bottom eight channels, unshaded). Waves in both the theta and delta ranges are more plentiful and of higher voltage over the left. The difference is especially prominent comparing the left (top arrow) and right (bottom arrow) parasagittal areas.


image

Figure 9-2 A mild but abnormal increase in slowing is seen in the left posterior quadrant (blue rectangles) compared with the right posterior quadrant (gray rectangles) during drowsiness.


Slow waves may be rhythmic or irregular (nonrhythmic). When there is a tendency to rhythmicity but the waves cannot be considered truly rhythmic, the intermediate term “semirhythmic” may be used (see Figure 9-3). Slow waves can be of varying amplitude. If slowing is only observed in a certain sleep stage, such as drowsiness or slow-wave sleep, this fact should be described. Slow waves may occur intermittently (in brief runs or bursts) or continuously (in long, continuous runs with few pauses). These different parameters are worth remembering and should usually be included in the written description of slow-wave activity.


image

Figure 9-3 Although not perfectly rhythmic, the slow waves seen in the left occipital (O1) channels (arrows) are not completely irregular (dots). Some of the unpredictability of the appearance of these slow waves is due to intermixing with other rhythms. Waves of this intermediate degree of rhythmicity can be termed semirhythmic.






















Parameter Examples
Rhythmicity: Irregular versus semirhythmic versus rhythmic
Amplitude: High voltage versus low voltage
Sleep Stage Specificity: Seen in wakefulness versus drowsiness versus sleep
Continuity: Intermittent versus continuous
Localization: Focal versus regional versus generalized

The parameters can be remembered using the mnemonic “RASCL.” Certain combinations of these parameters can define specific, well-described slow-wave abnormalities. For instance, a slow wave that is frontal, occurs intermittently, is rhythmic, and is seen during wakefulness has been given the specific name FIRDA (frontal intermittent rhythmic delta activity), discussed further later in the chapter.



Focal Slowing


Focal slow waves are the classic sign of a lesion in the cerebral hemispheres. Before the era of modern neuroimaging, the electroencephalogram was an important tool for the localization of cerebral tumors. Today the magnetic resonance imaging (MRI) scanner has moved to the forefront in localizing tumors, but the EEG may still play an important role in identifying certain types of focal lesions that may or may not be evident on neuroimaging. The EEG has the advantage over imaging studies of identifying areas of electrical abnormality which may be “nonanatomical,” that is, functional abnormalities of the brain that may not be visible on an MRI or computed tomography (CT) scan. Examples of “nonanatomical” slowing may include postictal slowing (see Figure 9-4), slowing from trauma that has not caused an MRI lesion (see Figure 9-5), or even migraine (discussed subsequently).


image

Figure 9-4 Significant slowing is evident in the top eight EEG channels (left hemisphere) compared with the bottom eight EEG channels (right hemisphere). The fact that this slowing represented a postictal change after a seizure was confirmed by demonstrating clearance of the slowing on a repeat EEG 2 months later. Depending on the nature of the patient and the intensity and duration of the seizure, postictal slowing may last from several seconds to as long as four weeks.


image

Figure 9-5 Slowing and sharp waves are seen over the right hemisphere in this patient following a brain contusion. Note the sharp waves (arrows) over the right hemisphere. Normal alpha range activity with short wavelengths (A scale) is seen on the left compared with slower, theta range activity with correspondingly longer wavelengths (B scale) on the right.


Focal slow waves may mark an area of previous, rather than acute, injury. Brain lesions that cause focal slowing in the absence of epileptiform activity may not necessarily be prone to seizures. Figures 9-6 and 9-7 show a left temporal slow wave abnormality; the perinatally acquired lesion causing the slow wave is shown in Figure 9-8. Figure 9-9 shows a subtle right occipital slow wave brought on by hyperventilation. If asymmetric slowing is only seen during hyperventilation, especially when the asymmetry is relatively mild as seen in this example, it is less likely to be associated with pathology than spontaneously occurring slow-wave asymmetries. Slow-wave asymmetries that alternate sides, whether spontaneous or elicited by hyperventilation, are much less likely to be clinically significant.


image

Figure 9-6 Close comparison of the temporal areas (shaded areas) shows an asymmetry of slow activity with increased theta waves on the left as a result of an old perinatal injury to the left temporal lobe tip in a 17-year-old girl. The arrows indicate individual theta waves in the left temporal area. The patient’s scan is shown in Figure 9-8. The lesion resulted in complex partial seizures.


image

Figure 9-7 Close-up views of the left and right temporal channels taken from the shaded areas of the previous figure are shown. Note the mixture of slow waves with wider bases in the top channels compared with the faster alpha rhythms with narrower bases in the bottom channels.


image

Figure 9-8 This is the T2-weighted magnetic resonance imaging (MRI) scan of the patient whose EEG is shown in the previous two figures. In this MRI sequence, the cerebrospinal fluid appears white. Note the loss of volume in the left temporal pole (arrow), which is responsible for the EEG asymmetry noted in the previous figures.


image

Figure 9-9 A subtle increase in slow-wave activity is seen in the right occipital area compared with the left (compare shaded areas). The posterior rhythm appears against a flat baseline in the T5-O1 channel (upper shaded area), but the posterior rhythm rides up and down on a low voltage wave in the T6-O2 channel (lower shaded area). The same comparison can be made between the P3-O1 and P4-O2 channels.


Focal slow-wave abnormalities have generally been associated with deeper lesions located at the level of the deep white matter (as opposed to more superficial gray matter lesions), although exceptions to this rule do occur. As discussed next, more superficial abnormalities of cerebral cortex are classically associated with decreases in beta activity.



Intermittent Rhythmic Delta Activity



Frontal Intermittent Rhythmic Delta Activity and Occipital Intermittent Rhythmic Delta Activity


The term intermittent rhythmic delta activity (IRDA) refers to rhythmic delta activity occurring in brief bursts, usually lasting no longer than a few seconds, typically located either frontally or occipitally (see Figure 9-10). In some examples the bursts can be more generalized (see Figure 9-11), and occasionally they can be asymmetrical. IRDA is a pattern typically seen in wakefulness and is usually associated with processes of mild to moderate severity. Keeping in mind that the patient generally must attain some level of wakefulness to manifest frontal (FIRDA) or occipital (OIRDA) IRDA, these patterns would not be expected in patients whose recordings are restricted to more deeply sedated or comatose states.


image

Figure 9-10 Frontal intermittent rhythmic delta activity (FIRDA) consists of rhythmic runs of delta activity in the frontal areas of varying duration (arrows). Although FIRDA appears frontally, the finding does not necessarily imply a frontal pathology (see text).


image

Figure 9-11 Compared with Figure 9-10, this burst of intermittent rhythmic slow (arrow) has a more diffuse (rather than frontal) distribution. It is seen equally well in the anterior and posterior channels.


One important feature of FIRDA and OIRDA is that these abnormal patterns do not suggest a specific localization. Surprisingly, the tendency for IRDA to occur either frontally or occipitally is not dictated by an anterior or posterior location of the patient’s lesion but rather by the patient’s age. Up to approximately 10 years of age, IRDA tends to occur in the occipital areas. By the early teenage years, IRDA tends to occur frontally. One reason that it is important to correctly identify an example of rhythmic slowing as IRDA is to avoid being trapped into inappropriately using IRDA as a “falsely localizing sign.” As implied earlier, FIRDA is not particularly associated with frontal lesions, and OIRDA is not particularly associated with occipital lesions; either can be associated with anterior or posterior brain abnormalities. Indeed, the location of the IRDA tells us more about the age of the patient than the location of the lesion.


It is interesting to note that the predilection of the hyperventilation response for the occipital area in children and for the frontal area in adults parallels the age dependence of FIRDA and OIRDA. In older patients, the hyperventilation response may mimic the appearance of FIRDA, and in younger patients it mimics OIRDA, which is to say that younger children tend to manifest the intermittent rhythmic slow of hyperventilation in the occipital areas.


FIRDA and OIRDA are etiologically non-specific, and can be caused by a variety of toxic, metabolic, and other processes that affect the central nervous system. Their presence signals some type of cerebral disturbance, focal or diffuse.



Temporal Intermittent Rhythmic Delta Activity


Temporal IRDA (TIRDA) should be considered separately from FIRDA and OIRDA. The presence of intermittent, rhythmic trains of delta activity in either temporal lobe has been associated with temporal lobe epilepsy (see Figure 9-12). Furthermore, the side of the TIRDA, which is usually unilateral, indicates the side of the lesion if one is present. This localizing property of TIRDA distinguishes it from the major types of IRDA (FIRDA and OIRDA). Thus, unlike FIRDA and OIRDA, TIRDA is considered a potentially epileptogenic abnormality and has localizing value.


image

Figure 9-12 Temporal intermittent rhythmic delta activity (TIRDA) is seen in a patient with temporal lobe epilepsy. This rhythmic delta activity is seen best in the left temporal area (arrows), although the left parasagittal area is involved as well. In this example, some of the theta waves are sharply contoured, which is not a necessary feature of TIRDA.



Occipital Intermittent Rhythmic Delta Activity in Childhood Absence Epilepsy


A specific type of occipital intermittent rhythmic delta activity is seen in children with childhood absence epilepsy. Intermittent rhythmic 3-Hz delta activity, similar to OIRDA, can be seen in what can be either brief or prolonged runs in the occipital area in children who have childhood absence epilepsy. This type of rhythmic delta activity distinguishes itself from typical examples of OIRDA in that the runs can be quite prolonged, lasting many seconds.



Generalized or Diffuse Slowing


In general, theta and delta rhythms are not expected in the waking adult EEG. In younger patients; however, theta waves may be seen in the normal waking EEG. The range of normal posterior rhythms in childhood may serve as a useful reminder that theta frequencies are commonplace in the EEG of the awake child (see Table 2-2 in Chapter 2, “Visual Analysis of the EEG).” For instance, in individuals young enough that a posterior rhythm of 7 Hz is considered normal, other 7-Hz rhythms may also be seen elsewhere in the normal waking EEG.


The determination that an adult EEG is abnormal simply because theta waves are present is made more difficult by the fact that it is normal for theta waves to appear with drowsiness. Therefore, in one way or another, the electroencephalographer must establish that a patient is not simply drowsy at the time that theta waves are seen to label them abnormal. Interpreting an EEG as abnormal on the basis of the presence of theta waves that are, in reality, related to drowsiness would constitute a significant error in interpretation. Indications that a patient is probably awake would include the presence of a posterior rhythm, temporalis or frontalis muscle artifact, eyeblink artifact, or conversation. The technologist can document (or force) alertness by having the patient count or perform some other task to exclude drowsiness as an explanation for observed slow-wave activity.


The list of pathologic states that can cause diffuse slow-wave activity is long and includes almost any abnormal state that can cause a diffuse cerebral disturbance. The most common causes are postictal (postseizure) states, postanoxic states, diffuse traumatic injuries, infectious or inflammatory processes of the nervous system, toxic states related to drugs or other metabolic derangements (hepatic, renal, etc.), and a large number of other processes that have the potential to affect large portions of the cerebrum. Most of the many possible causes of coma are also on the list of possible causes of diffuse slowing in the EEG. For this reason, it is not feasible for the electroencephalographer to give a specific differential diagnosis for this pattern in the report. Often the interpretation will summarize the possible etiologies as a “diffuse cerebral disturbance” or a “diffuse encephalopathy.” The interpretation may also specifically discuss any clinical entities that have been questioned in the clinical history. EEG patterns in coma, many of which consist of slow-wave patterns, are discussed in further detail in Chapter 12, “Electroencephalographic Patterns in Stupor and Coma.”



Migraine


Because it is such a common disorder, the role of the EEG in migraine merits a separate discussion. A large number of abnormalities have been described in patients with common forms of migraine (“common” or “classic” migraine) between migraine attacks, so-called interictal EEG abnormalities in migraine. Many of these claims have been challenged because of the lack of well-controlled studies. The main problem is that the high frequency of migraine in the general population increases the chance that an individual who has any particular EEG abnormality may also incidentally have migraine. Some authors have claimed that focal slowing, and even epileptiform activity, are seen with increased frequency interictally in migraine patients. The “high-frequency photic response” (the ability to maintain a photic driving response at particularly high flash stimulation frequencies) has also been described in migraine patients. None of these phenomena, however, has been clearly proved to have an incidence in migraine patients above that seen in the general population. Because there are no EEG abnormalities that have proved to be useful in helping to diagnose uncomplicated migraine patients between attacks, EEG is not recommended as part of the routine evaluation of migraine patients. The EEG may be indicated, however, in the rare cases in which there is a question that a headache symptom may represent an epileptic seizure phenomenon.



Complicated Migraine


Focal slowing, either rhythmic or irregular, may be seen during the course of complicated migraine attacks. Figure 9-13 shows an example of high-voltage irregular slowing over the left hemisphere during an attack of hemiplegic migraine (a rare type of complicated migraine) in a 12-year-old girl. Confusional migraine, another rare type of complicated migraine that occurs predominantly in children, may cause bilateral slowing in the EEG. In some cases, migraine with visual aura can cause slowing during an attack, particularly in the posterior quadrants, although in many cases the EEG is not particularly remarkable during attacks.


image

Figure 9-13 A 12-year-old girl recorded during an episode of hemiplegic migraine. Note the high voltage slowing over the left hemisphere (top eight channels) compared with the relatively more normal-appearing activity over the right hemisphere (bottom eight channels).

Related posts:

Default ThumbnailThe Structure and Philosophy of the EEG Report Introduction EEG Patterns in Stupor and Coma Electroencephalographic Electrodes, Channels, and Montages and How They Are Chosen Filters in the Electroencephalogram Introduction to Commonly Used Terms in Electroencephalography

Stay updated, free articles. Join our Telegram channel
Tags:
Mar 12, 2017 | Posted by in NEUROLOGY | Comments Off on The Abnormal EEG

Full access? Get Clinical Tree
Get Clinical Tree app for offline access