Epilepsy and its component syndromes remains principally a clinical diagnosis with a distinct supplement from the EEG. EEG displays the physiological counterpart to clinical evaluation and is the only laboratory test providing potentially diagnostic information. EEG is the only means by which a diagnosis of epilepsy can be unequivocally established or a seizure accurately classified.
Sensitivity of Epileptiform Phenomena
Epileptiform activity appears in 50% of single awake recordings among patients already known to have epilepsy, the proportion rising to 80% to 85% if sleep is included in the recording. Sleep increases the incidence of benign “rolandic” spikes, anterior temporal spikes, and spike–waves. The slower waves in sleep help unravel questionable spikes of wakefulness. Multiple recordings also increase the yield: two EEGs -80% to 85%; four EEGs -90% (Binnie & Stefan, 1999). Hyperventilation and photic stimulation also increase spike incidence, as discussed and demonstrated further on.
Specificity of Epileptiform Phenomena
These occur in 0.4% of healthy persons and 0.3 to 3% of hospitalized patients without epilepsy (Ajmone Marsan & Zivin, 1970). Some drugs (e.g. clozapine, lithium) may elicit diffuse spikes (Blume, 2006).
Prognosis after a First Seizure
Various percentages have emerged from studies relating the presence or absence of spikes on an EEG after a first seizure and follow-up. In practice, EEG helps to define the epilepsy syndrome, and that is the principal prognostic indicator.
Temporal lobe EEG abnormalities and their associated epilepsies occur so commonly that the EEG reader may overlook significant features elsewhere. Therefore, such extratemporal areas are the first covered.
Occipital lobe seizures, especially if lesion-based, will be associated with a reduced, disrupted, and less reactive alpha activity than on the normal or less affected side. Occipital spikes commonly spread from O1,2 to the posterior–temporal region, T5,T6. These latter overlie the anterior occipital convexity and thus record its activity. Spikes and seizures spread less commonly to parietal regions, P3,4. Interictal spikes appear mostly ipsilateral to occipital seizure onset in 80% to 97% of patients (Blume et al., 2005; Salanova et al., 1992; Williamson et al., 1992).
EEGs of children with benign rolandic epilepsy usually display frequent stereotyped, tangentially orientated, prominent spikes with electropositivity frontally and negativity in the central–parietal area. Extension of the negative field to the midtemporal area occurs commonly. Sleep augments their incidence. These spikes are sufficiently abundant that their identification is requisite for the diagnosis of benign rolandic epilepsy of childhood (Berg et al., 1999). However, such discharges may also appear in patients who do not have epilepsy(Engel, 1984). Usually these spikes arise from normal background activity; however, trains of these may be associated with central delta activity. In contrast, spikes occur sporadically in patients with lesion-related focal motor seizures except those with cortical dysplasia, where rhythmic epileptiform discharges (REDs) are characteristic (Gambardella et al., 1996).
If the frontal lobe convexity is involved in epileptogenesis, spikes and seizures will be readily apparent (Blume et al., 2001). However, mesial–or inferior frontal spikes or seizures may not be detected by scalp electrodes (Blume & Oliver, 1996). The frontal lobe is the one most likely to generate secondary bisynchrony of epileptiform phenomena (Blume & Pillay, 1985). Indirect EEG clues may be regional frontal theta/delta or an ipsilateral reduction of beta activity.
EEG combined with precise clinical analysis is of particular value in patients whose ictal semiology suggests limbic involvement. As such involvement may result from spread as well as origin, one must first look for an extralimbic origin; for example, occipital spikes would suggest such propagation. A lack of temporal spikes with limbic-like semiology may indicate orbital frontal or cingulate limbic origin.
As illustrated first by the Gibbs' (1952) and later by Sadler and Goodwin (1989), the anterior mesial temporal negative spike field extends below scalp 10/20 electrodes involving principally the mandibular notch (M1,2), and ear (A1,2) positions as it creates a tangential field, while the positive component of the dipole involves the ipsilateral and even contralateral parasagittal regions. Neocortical temporal spikes express the negative field over the temporal lobe as a radial field (Ebersole, 2003). These principles help identify and localize temporal spikes, which, when they appear at least 3 times more often over one temporal lobe than the other over several recordings, correlate with seizure origin in 93% to 95% of patients (Blume et al., 1993).
Unfortunately, temporal lobe spikes appear rarely in youth and become more common among adolescents and adults. Temporal interictal rhythmic delta activity (TIRDA) is an additional feature found in some patients with temporal lobe epilepsy (Reiher et al., 1989).
The Oxford dictionary defines “generalized” as “including or affecting or applicable to most parts or cases of things” (Sykes, 1982). As applied to epilepsy, “generalized” indicates conditions that begin without specific warning, whose motor manifestations are usually symmetrically bilateral, and which terminate without focal or lateralized postictal manifestations. Proximal portions of limbs and axial muscles are more involved than distal. Facial and periocular muscles may contract symmetrically.
Hypsarrhythmia appears in only two thirds of first EEGs performed for epileptic spasms. The others may have multifocal spikes and mildly excess slowing (Jeavons & Bower, 1974). In subsequent years, when such patients are transferred to adult neurological care, hypsarrhythmia may have evolved to slow spike–waves and fast rhythmic waves. In others, mild to moderate multifocal or diffuse excess theta with sporadic spikes may appear.
Sequential diffuse spikes or rhythmic waves at 10 to 25 Hz accompany tonic seizures. Interictal EEG patterns correlate with whichever syndrome contains tonic seizures as a component. This most likely would be slow spike–waves and fast rhythmic waves of the Lennox–Gastaut syndrome (Gastaut & Broughton, 1972).
Interictally and ictally, syndromes containing myoclonic seizures with or without an atonic component have bilaterally synchronous spike–waves or polyspike–waves repeating at 3 to 4 Hz.
Bisynchronous 2.5 to 3.5 Hz spike–waves appear on at least 80% of EEGs in patients with absence seizures (Dalby, 1969). As does an absence seizure, spike–waves begin and end abruptly. Lack of spike–waves on two recordings increases the likelihood that staring spells represent another condition, such as excessive daytime sleep or temporal lobe seizures. As spike–wave quantity correlates closely with that of incidence of absence seizures, follow-up EEGs help monitor the efficacy of its therapy (Miller & Blume, 1993).
The waking EEG in patients with only generalized tonic–clonic seizures may be normal or contain bursts of theta. Spike–waves at 3 to 4 Hz, similar to those occurring with myoclonic seizures, may appear spontaneously or upon hyperventilation or photic stimulation. All these features can occur in juvenile myoclonic epilepsy (JME) (Blume, 2002). Generalized tonic–clonic seizures (GTC) are sequentially accompanied by the above-described features of each component. Some secondarily generalized seizures may display a clear focal (as opposed to hemispheric) onset, while others may not. An inferior or mesial onset or ictal muscle artifact may prevent EEG detection of such initial focality.
Although the degree of EEG abnormality among patients with known or suspected dementia correlates somewhat with cognitive ability at the time of recording, these abnormalities principally reflect the pace of intellectual decline.
The following diffuse or regional abnormalities may appear in patients with known or suspected dementia:
Cognition may be slowed in depressed elderly patients, leading occasionally to an erroneous impression of dementia—e.g., “pseudodementia” A normal EEG can suggest that only depression is present (Brenner et al., 1989).
Correlating with its relatively slow progression and limited topographical involvement, the EEG may be normal or display mildly excess frontal or temporal theta.
Two or more of the following features may appear:
The severity of these abnormalities increases somewhat with the degree of cognitive impairment, but it correlates even more closely with the slope of decline (Kaszniak et al., 1978;Rae-Grant et al., 1987). Moreover, characteristic EEG abnormalities may precede neuroimaging evidence of atrophy (Merskey et al., 1980).
Creutzfeldt–Jakob Disease (CJ)
The EEG resembles that of a fast-paced Alzheimer's disease with some distinctive additional features. The earliest of these is focal or multifocal spikes, each spike having a brief duration. With progression, these spikes become bisynchronous and repeat in a quasi-periodic fashion (Brenner & Schaul, 1990). Such “periodicity” signifies rapid progression. A close but inexact relationship between periodic spikes and myoclonus exists. Although alerting may attenuate periodic spikes, such may be “paced” by regularly repeating somatosensory afferents.
Spikes rarely become prominent in other dementias, and the periodic component is especially characteristic of CJ. Thus, the following EEG abnormalities appear as the disease progresses: excess delta for age, periodic bisynchronous spikes, decline in delta and background amplitude with preservation of spikes.
In the acute management of stroke, EEG has largely relinquished its role to neuroimaging, but its findings may influence management in some circumstances. The following describes acute changes in ischemic stroke of each of the several arteries.
Regional attenuation of normal background activity at the site of ischemia and arrhythmic regional delta activity in that area and its penumbra are the principal EEG manifestations of acute stroke. Frontal intermittent rhythmic delta activity (FIRDA), diffuse or hemispheric, may appear with large strokes and their accompanying edema. The extent of EEG abnormality depends upon whether occlusion of a major artery or its branch has occurred, the availability of collateral and anastomotic channels, and the speed of occlusion, with gradual occlusion allowing greater collateral compensation.
Frontal–parietal–temporal delta/attenuation indicates stroke of the internal/middle cerebral artery territory. Greater frontal delta (see further) may occur if the anterior communicating artery (ACoA) is small. Unilateral frontal delta and decrease of beta activity may occur with anterior cerebral artery (ACA) occlusion distal to the ACoA. Bifrontal delta suggests ACA occlusion proximal to the ACoA if this ACA stem supplies both ACAs. FIRDA often accompanies ACA infarction, especially if bilateral. Occipital delta involving the parietal and posterior temporal area with attenuation or loss of ipsilateral alpha activity occurs with posterior cerebral artery infarction. Extension of delta to the anterior mesial–temporal area suggests involvement of the artery's temporal branch.
Mild focal delta or excess focal theta for age, if otherwise unexplained, may reflect clinically occult ischemic events or that a “transient ischemic attack” was not truly transient.
Although the incidence of acute or chronic seizure disorders in stroke is about 3%, it climbs to 8% with hemorrhagic stroke and large cortical strokes (Szaflarski et al., 2008). Focal or secondary bilaterally synchronous spikes, periodic lateralized epileptiform discharges (PLEDs) or subclinical
seizures may reflect seizures in strokes with these attributes. Because glia-derived cytokines can promote neuronal death, poststroke seizures may increase the extent of an ischemic lesion (Blume, 2009, review).
Perhaps owing to its sudden onset, embolic stroke causes seizures in about 10% of cases and focal spikes in 50% (Rasheva, 1981).
Because acute central nervous system (CNS) inflammatory disorders are often associated with systemic illness, any EEG abnormalities may reflect electrolytic or other metabolic derangements, especially in children.
If meningitis has a minimal encephalitic component, the EEG changes would be slight. Moreover, they could represent the above-mentioned systemic effects.
Normal background rhythms are replaced by diffuse or regionally accentuated arrhythmic delta activity. Acute-phase EEG abnormalities may add little to clinical evaluation in estimating prognosis. However, in a fully comatose patient, EEG may provide the best window to cortical function, and this aspect may help to predict outcome.
However, in patients whose CNS function is markedly altered, epileptic seizures may produce unusual clinical manifestations or may be clinically undetectable. Continuous EEG–video recording can detect such events and monitor the efficacy of antiepileptic therapy.
In the clinical context of viral encephalitis, the EEG may help to determine the cause. Periodic broad spikes repeating every 0.5 to 4 s, equivalent to PLEDs, in temporal leads (M1,2; F7,8; T3,4), occurring either unilaterally or bilaterally, strongly suggest herpes simplex virus as etiology. Although these appear only in about 60% of cases in some series (Malhotra et al., 2009), others (Upton & Gumpert, 1970) claim that daily recordings will ultimately reveal their presence. Such PLEDs remain for about 2 weeks and rarely more than 3 weeks. In current practice, acyclovir is given promptly whenever a sporadically occurring viral encephalitis is suspected in adults or children, not waiting for this EEG feature to appear. Moreover, similar periodic epileptiform features occur in a wide variety of acute illnesses.
Subacute Sclerosing Panencephalitis (SSPE)
This disease now occurs very rarely in much of the world. It presents as progressive mental deterioration at age 5 to 10 years, with language and speech impairment, motor incoordination, and periodic myoclonic-like jerks that are actually brief myotonic contractions. The characteristic EEG feature is 300 to 1500 µV (!) periodic (0.5 to 2 s) polyphasic broad or usual-duration spikes appearing regionally in one hemisphere or bisynchronous with an interhemispheric or anterior–posterior lag. In some patients, the interparoxysmal interval may extend to 1 to 5 min (Reiher et al., 1973). These complexes are variably associated with the motor events, but afferent stimuli have no effect. Although such motor events cease in sleep, the EEG complexes persist. Other signs of a deteriorating CNS also appear: background loss, delta, decreased spindles and sleep organization.
EEGs reflect the aggressiveness of this process, with prominent and persistent delta activity, loss of background components, and abundant spikes involving one hemisphere more than the other. Spikes may appear in a “periodic” fashion and often propagate widely throughout the involved hemisphere. Such propagation may underlie the juxtaposition of apparent discordant ictal semiology. The presence of less prominent epileptiform and nonepileptiform abnormalities in the other hemisphere gives rise to the term “regionally accentuated encephalitis.”
This, the most common cause of focal epilepsy worldwide, is characterized by focal or multifocal spikes, with delta and theta paralleling its cortical involvement.
Excess diffuse delta and a lower than normal background frequency for age and state occur in virtually all metabolic disorders. Among patients with depressed levels of awareness, reactivity of EEG patterns to afferent stimuli and the presence of more than one pattern connote mild to moderate encephalopathies, while persistent, unvarying, monomorphic,
non-reactive features attest to a severe encephalopathy (Young et al., 1999). Triphasic waves (TWs) occur commonly in somnolent or stuporous states created by a variety of conditions, particularly hepatic encephalopathy (Ikeda et al., 2003; Niedermeyer, 2005). Although it may be ominous, patients with this feature may recover. The following features of TWs may aid in their distinction from the morphologically similar sequential slow spike–waves of nonconvulsive status epilepticus: longer first phase, lower frequency, longer duration, no extra spike components, and an occasional lag of phase two (Boulanger et al., 2006).
Unlike hepatic conditions, uremia may produce bisynchronous spike–waves and photoparoxysmal responses. However, advances in management have lessened uremic metabolic fluctuations and thus the apparent incidence of such phenomena (Niedermeyer, 2005).
Some electrolytic abnormalities (e.g. hyponatremia) may elicit diffuse delta activity whose quantity depends as much on the rapidity of electrolyte change as upon its degree.
Surprisingly, apparently healthy subjects with intermittent migraine headaches may have mild amounts of excess focal theta activity between attacks. This may increase to regional delta with aphasic migraine or may extend throughout a hemisphere in hemiplegic migraine.
Neuroimaging has reduced the role of EEG in tumor diagnosis and management. However, persistent delta activity, particularly in a patient with intractable, cryptogenic focal epilepsy, may indicate further imaging to seek a neoplasm if an initial CT and/or MRI were negative. Meningiomas produce the least EEG changes, principally focal theta (Bazil et al., 2003). Slowly growing astrocytomas or oligodendrogliomas may display focal delta and moderately frequent focal spikes. Rapidly progressive tumors produce arrhythmic 0.5 Hz delta but few or no spikes. The aforementioned delta and theta replace background activity, particularly with the most rapidly growing neoplasms. Metastatic tumors typically produce prominent abnormalities resembling those of high-grade gliomas.
EEG alterations from medications vary according to dose, rate of metabolism, and volume of distribution. Systemic effect such as hyponatremia and hyperammonemia will also perturb the EEG.
Excess beta activity and a mild theta increase most commonly characterize medication effects. High doses of drugs such as clozapine, lithium and phenothiazines may elicit spikes or polyspikes, as can acute withdrawal from alcohol and barbiturates. Triphasic waves may reflect valproate-associated hyperammonemia or intoxication by baclofen, lithium, and others. Phenytoin may slow background rhythms, and carbamazepine and its epoxide metabolite may augment theta and produce diffuse delta (Blume 2006, review).
Ajmone Marsan C, Zivin LS. Factors relating to the occurrence of typical paroxysmal abnormalities in the EEG records of epileptic patients. Epilepsia. 1970;11: 361–381.
Bazil CW, Herman ST, Pedley TA. Focal electroencephalographic abnormalities. In: Ebersole JS, Pedley TA, eds. Current Practice of Clinical Electroencephalography. Philadelphia: Lippincott Williams & Wilkins; 2003:303–307.
Berg AT, Shinnar S, Levy SR, et al. Newly diagnosed epilepsy in children: Presentation at diagnosis. Epilepsia. 1999;40(4):445–452.
Binnie CD, Stefan H. Modern electroencephalography: Its role in epilepsy management. Clin Neurophysiol. 1999;110:1671–1697.
Blume WT. Clinical and basic neurophysiology of generalized epilepsies. Can J Neurol Sci. 2002;29:6–18.
Blume WT. Cytokines and strokes of ill fortune. Epilepsy Currents. 2009;9(2):42–43.
Blume WT. Drug effects on EEG. J Clin Neurophysiol. 2006;23(4):306–311.
Blume WT, Borghesi JL, Lemieux JF. Interictal indices of temporal seizure origin. Ann Neurol. 1993;34:703–709.
Blume WT, Ociepa D, Kander V. Frontal lobe seizure propagation: Scalp and subdural EEG studies. Epilepsia. 2001;42:491–503.
Blume WT, Oliver LM. Non-invasive electroencephalography in supplementary motor area epilepsy. In: Luders HO, ed. Advances in Neurology. Philadelphia: Lippincott-Raven; 1996:309–317.
Blume WT, Pillay N. Electrographic and clinical correlates of secondary bilateral synchrony. Epilepsia. 1985;26(6):636–641.
Blume WT, Wiebe S, Tapsell LM. Occipital epilepsy: Lateral versus mesial. Brain. 2005;128:1209–1225.
Boulanger J-M, Deacon C, Lecuyer D, et al. Triphasic waves versus nonconvulsive status epilepticus: EEG distinction. Can J Neurol Sci. 2006;33:175–180.
Brenner RP, Reynolds CF, Ulrich RF. EEG findings in depressive pseudoseizures and dementia with secondary depression. Electroencephalogr Clin Neurophysiol. 1989;72:298–304.
Brenner RP, Schaul N. Periodic EEG patterns. J Clin Neurophysiol. 1990;7:249–267.
Dalby MA. Epilepsy and 3 per second spike and wave rhythms. Acta Neurologica Scand Suppl. 1969;45:1–180.
Ebersole JS. Cortical generators and EEG voltage fields. In: Ebersole JS, Pedley TA, eds. Current Practice of Clinical Electroencephalography, 3rd ed. Philadelphia: Lippincott Williams & Wilkins; 2003:12–31.
Engel J Jr. A practical guide for routine EEG studies in epilepsy. J Clin Neurophysiol. 1984;1(2):109–142.
Gambardella A, Palmini A, Andermann F, et al. Usefulness of focal rhythmic discharges on scalp EEG of patients with focal cortical dysplasia and intractable epilepsy.Electroencephalogr Clin Neurophysiol. 1996;98(4):243–249.
Gastaut H, Broughton R. Epileptic Seizures: Clinical and Electrographic Features. Diagnosis and Treatment. Springfield, IL: Charles C. Thomas; 1972:37–47.
Gibbs F, Gibbs E. Psychomotor epilepsy. In: Gibbs FA, Gibbs EL, eds. Atlas of Electroencephalography Reading, MA: Addison Wesley; 1952:162–209.
Ikeda A, Klem GH, Luders HO. Metabolic, infectious and hereditary encephalopathies. In: Ebersole JS, Pedley TA, eds. Current Practice of Clinical Electroencephalography. Philadelphia: Lippincott Williams & Wilkins; 2003: 348–377.
Jeavons PM, Bower BD. Infantile spasms. In: Vinken PJ, Bruyn GW, eds. Handbook of Clinical Neurology. Vol 15: The Epilepsies. New York: Elsevier, 1974;219–234.
Kaszniak AW, Fox J, Gandell DL, et al. Predictors of mortality in presenile and senile dementia. Ann Neurol. 1978;3:246–252.
Malhotra A, Bell WE, Henderson FW. Infections of the central nervous system. In: David RB, ed. Clinical Pediatric Neurology. New York: Demos Medical; 2009:211–236.
Merskey H, Ball MJ, Blume WT, et al. Relationships between psychological measurements and cerebral organic changes in Alzheimer's disease. Can J Neurol Sci. 1980;7:45–49.
Miller H, Blume WT. Primary generalized seizure disorder: Correlation of epileptiform discharges with seizure frequency. Epilepsia. 1993;34(1):128–132.
Muller HF, Kral VA. The electroencephalogram in advanced senile dementia. J Am Geriatr Soc. 1967;15:415–426.
Niedermeyer E. Metabolic central nervous system disorders. In: Niedermeyer E, Lopes Da Silva F, eds. Electroencephalography, 5th ed. Philadelphia: Lippincott Williams & Wilkins; 2005:439–454.
Rae-Grant A, Blume W, Lau C, et al. The electroencephalogram in Alzheimer-type dementia: A sequential study correlating the electroencephalogram with psychometric and quantitative pathological data. Arch Neurol. 1987;44: 50–54.
Rasheva M. Epileptic seizures in the acute stage of embolic stroke. Electroencephalogr Clin Neurophysiol. 1981;52:78P (abstract).
Reiher J, Beaudry M, Leduc CP. Temporal intermittent rhythmic delta activity (TIRDA) in the diagnosis of complex partial epilepsy: Sensitivity, specificity and predictive value.Can J Neurol Sci. 1989;16:398–401.
Reiher J, Lapointe LR, Lessard L. Prolonged and variable intervals between EEG complexes in subacute inclusion body encephalitis. Can Med Assoc J. 1973;108: 729–732.
Sadler RM, Goodwin J. Multiple electrodes for detecting spikes in partial complex seizures. Can J Neurol Sci. 1989;16:326–329.
Salanova V, Andermann F, Olivier A, et al. Occipital lobe epilepsy: Electroclinical manifestations, electrocorticography, cortical stimulation, and outcome in 42 patients treated between 1930 and 1991: Surgery of occipital lobe epilepsy. Brain. 1992;115:1655–1680.
Sykes JB. The Concise Oxford Dictionary. Oxford: Clarendon; 1982:411.
Szaflarski JP, Rackley AY, Kleindorfer DO, et al. Incidence of seizures in the acute phase of stroke: a population-based study. Epilepsia. 2008;49(6):974–981.
Upton A, Gumpert J. Electroencephalography in the diagnosis of herpes simplex encephalitis. Lancet. 1970;1:650–652.
Williamson PD, Thadani VM, Darcy TM, et al. Occipital lobe epilepsy: Clinical characteristics, seizure spread patterns, and results of surgery. Ann Neurol. 1992;31:3–13.
Young GB, Kreeft JH, McLachlan RS, et al. EEG and clinical associations with mortality in comatose patients in a general intensive care unit. J Clin Neurophysiol. 1999;16(4):354–360.