Clinical Neuroanatomy, 27 ed.

CHAPTER 23. Electrodiagnostic Tests

In addition to using a patient’s history, physical examination, and imaging results, the clinician can obtain information about the functional status of various parts of the nervous system by monitoring its electrical activity. This is accomplished via a variety of electrodiagnostic tests, which are described in this chapter.


Electroencephalography provides a noninvasive method for studying the ongoing or spontaneous electrical activity of the brain. The potentials of the brain are recorded in an electroencephalogram (EEG); they appear as periodic waves, with frequencies ranging from 0.5 to 40 cycles per second (cps or hertz [Hz]) and with an amplitude that ranges from five to several hundred microvolts. Because the amplitude of cerebral electrical activity is much smaller than that obtained from the heart in an electrocardiogram (ECG), sensitive (but stable) amplification is necessary to produce an undistorted record of brain activity; this requires proper grounding and electrical shielding.

Clinical Applications

Electroencephalography can provide useful information in patients with structural disease of the brain, especially when seizures occur or are suspected. Electroencephalograms can be very useful in classifying seizure disorders, and because optimal drug therapy varies for different types of seizures, the EEG findings may have important implications for treatment. Electroencephalography is also useful in evaluating cerebral abnormalities in a number of systemic disorders and in performing workups on patients with sleep disorders.

Because computed tomography (CT) scanning and magnetic resonance imaging (MRI) have higher spatial resolution and can localize lesions in three dimensions, these imaging techniques are usually used in preference to EEG for the localization of destructive lesions in the brain. When other tests are not available, an EEG can furnish help in determining the area of cerebral damage. Electroencephalography has its limitations, however, and normal-appearing records can be obtained despite clinical evidence of severe organic brain disease. The use of depth electrography—the localization of a focus by recording from electrodes implanted within the brain—may be advisable in certain cases.


The activity recorded in the EEG originates mainly from the superficial layers of the cerebral cortex. Current is believed to flow between cortical cell dendrites and cell bodies. (The dendrites are oriented perpendicular to the cortical surface.) As a result of the synchronous activation of axodendritic synapses on many neurons, summed electrical currents flow through the extracellular space, creating the waves recorded as the EEG. The pattern of activation of cortical neurons, and thus the EEG, is modulated by inputs from the thalamus and reticular formation (called the reticular activating system by early researchers).


To detect changes in activity that may be of diagnostic importance, simultaneous recordings are obtained, when possible, from multiple areas on both the left and right sides of the brain. Electrodes are ordinarily attached to the scalp over the frontal, parietal, occipital, and temporal areas; they are also attached to the ears (Fig 23–1).


FIGURE 23–1 A single-plane projection of the head, showing all standard positions of electrode placement and the locations of the central sulcus (fissure of Rolando) and the lateral cerebral fissure (fissure of Sylvius). The outer circle is drawn at the level of the nasion and inion; the inner circle represents the temporal line of electrodes. This diagram provides a useful guide for electrode placement in routine recording. A, ear; C, central; Cz, central at zero, or midline; F, frontal; Fp, frontal pole; Fz, frontal at zero, or midline; O, occipital; P, parietal; Pg, nasopharyngeal; Pz, parietal at zero, or midline; T, temporal. (Courtesy of Grass Technologies, An Astro Med, Inc. Produce Group, West Warwick, RI.)

With the patient recumbent or seated in a grounded, wire-shielded cage, a recording at least 20 minutes long is obtained; the eyes should be closed. Hyperventilation, during which the patient takes 40 to 50 deep breaths per minute for 3 minutes, is routinely used during this time because it frequently accentuates abnormal findings (epileptiform attacks) and may disclose latent abnormalities. Rhythmic light-flash stimulation (1-30 Hz), also termed photicstimulation, is performed for 2 minutes or longer as part of the recording routine. In some cases, the EEG is continued after the patient is allowed to spontaneously fall asleep or after sedation with drugs; under these circumstances certain epileptic discharges and other focal abnormalities are more likely to be recorded.

Types of Waveforms

The synchronized activity of many of the dendritic units forms the wave pattern associated with alpha rhythm when the patient is awake but at rest with the eyes closed. The alpha rhythm has a periodicity of 8 to 12 Hz. Desynchronization, or replacement of a rhythmic pattern with irregular low-voltage activity, is produced by stimulation of specific projection systems from the spinal cord and brain stem up to the level of the thalamus.

When the eyes are opened, the alpha rhythm is replaced by an alpha block, a fast, irregular, low-voltage activity. Other forms of sensory stimulation or mental concentration can also break up the alpha pattern. Desynchronization is sometimes termed the arousal, or alerting, response, because this breakup of the alpha pattern may be produced by sensory stimulation and is correlated with an aroused or alert state. The beta rhythm is characterized by low amplitude (5-20 μV) waves with a rhythm faster than 12 Hz, most prominent in the frontal regions.

Theta rhythms (4-7 Hz) are normally seen over the temporal lobes bilaterally, particularly in older patients, but can also occur as a result of focal or generalized cerebral dysfunction. Delta activity (1-3 Hz) is never seen in the normal EEG and indicates significant dysfunction of the underlying cortex. Brain tumors, cerebral abscesses, and subdural hematoma are often associated with focal or localized slow-wave activity. CT and MRI, however, can provide more information about the location and structure of the lesion and have largely replaced EEG for the diagnosis of these disorders.

Epilepsy is an expression of various cortical diseases characterized by transient disturbances of brain function manifested by intermittent high-voltage waves. Electroencephalograms from patients with various types of epilepsy are shown in FIGURE 23–2Spikes and sharp waves have characteristic shapes and occur either as part of seizure discharges or interictally in patients with epilepsy. These EEG abnormalities can be diffuse or focal, suggesting a localized abnormality.


FIGURE 23–2 Representative electroencephalograms.

Absence seizures of childhood (petit mal), which are characterized by brief (up to 30 seconds) loss of consciousness without loss of postural tone, are associated with a characteristic three-per-second spike-and-wave abnormality on EEG. Complex partial seizures (which usually have a temporal lobe origin), in contrast, can also be associated with impaired awareness, but the EEG usually shows focal temporal lobe spikes or appears normal because the aberrant and relatively deep temporal lobe discharges cannot be detected with scalp electrodes.

Infectious, toxic, and metabolic disorders affecting the nervous system can be accompanied by characteristic EEG abnormalities. For example, in herpes simplex encephalitis, the EEG displays periodic high-voltage sharp waves over the temporal lobes at regular three-per-second intervals. In Creutzfeldt-Jakob disease (also termed subacute spongiform encephalopathy), the EEG usually shows a pattern of burst suppression characterized by stereotyped, high-voltage slow and sharp wave complexes superimposed on a relatively flat background. In hepatic encephalopathy, bilaterally synchronous triphasic waves are often present.


Whereas the EEG displays ongoing or spontaneous electrical activity, evoked potential recordings permit the measurement of activity in cortical sensory areas and subcortical relay nuclei in response to stimulation of various sensory pathways. Because the electrical signals are small, computerized averaging methods are used to extract the time-locked neural signals evoked by a large number of identical stimuli. The latency, amplitude, and waveform of the evoked potential provide information about impulse conduction along the pathway, or group of neurons, under study, and thus about the functional integrity of the pathway.

Visual Evoked Potentials

Visual evoked potentials (VEPs) are usually elicited by having the patient fixate on a target and flashing a reversing checkerboard pattern on a screen centered around the target. The VEPs recorded in this manner are sometimes called pattern-shift VEPs (PSVEPs). These are recorded using scalp electrodes placed over the left and right occipital poles. This reaction is clinically useful in detecting slight abnormalities in the visual pathways; for example, optic nerve lesions can be recognized by stimulating each eye separately, because the response to stimulation of an affected optic nerve is absent or impaired. With visual pathway lesions behind the optic chiasm, a difference in response of the two cerebral hemispheres may occur, with a normal response in the occipital cortex of the normal cerebral hemisphere and an absent or abnormal response in the affected cerebral hemisphere (see Chapter 15).

Brain Stem Auditory Evoked Response

A standard brain stem auditory evoked response (BAER) consists of seven potentials that are recorded from the human scalp within 10 milliseconds of a single appropriate acoustic stimulus. Abnormalities in the response may provide evidence of clinical neurologic disorders involving the brain stem. The test has some clinical value in demonstrating structural brain stem damage caused by various disorders (see Chapter 7).

In a normal human with scalp electrodes placed on the vertex, a click stimulus presented to the ear may evoke typical responses with seven wave components that are believed to come from the region of the auditory nerve (wave I), dorsal cochlear nucleus (wave II), superior olive (wave III), lateral lemniscus (wave IV), and inferior colliculus (wave V). Wave VI may indicate activity of the rostral midbrain or caudal thalamus or thalamocortical projection, and wave VII originates in the auditory cortex (Fig 23–3).


FIGURE 23–3 Far-field recording of brain stem auditory response latencies in humans showing proposed functional-anatomic correlations. Diagram shows normal latencies for vertex-positive brain stem auditory evoked potentials (waves I–IV) evoked by clicks of 60 dBHL (60 dB above normal hearing threshold) at a rate of 10/s. Lesions at different levels of the auditory pathway tend to produce response abnormalities beginning with the indicated components. Intermediate latency (5.8 ms) between latencies of waves IV and V is the mean peak latency of fused wave IV/V when present. Cz+, vertex positivity, represented by an upward pen deflection; Cz-, vertex negativity, represented by a downward pen deflection. (Reproduced, with permission, from Stockard JJ, Stockard JE, Sharbrough FW: Detection and localization of occult lesions with brain stem auditory responses. Mayo Clin Proc1977;52:761.)

Somatosensory Evoked Potentials

To obtain somatosensory evoked potentials (SEPs), repetitive electrical stimuli are applied through electrodes placed over the median, peroneal, and tibial nerves. This usually can be done without causing pain. Recording electrodes are placed over Erb’s point above the clavicle, over the C2 spinous process, over the contralateral somatosensory cortex for stimulation of the upper limb, and over the lumbar and cervical spine and contralateral somatosensory cortex for stimulation of the lower limb. Depending on the pattern of delay, it is possible to localize lesions within peripheral nerve (conduction delay or increased conduction time between stimulation site and Erb’s point or lumbar spine), within spinal roots or dorsal columns (delay between Erb’s point or lumbar spine and C2), or in the medial lemniscus and thalamic radiations (delay recorded at cortical electrode but not at more caudal recording sites).


Methods for noninvasively stimulating the motor cortex and cervical spinal cord in humans have been developed and permit the evaluation of conduction in descending motor pathways. Because the largest neurons have the lowest thresholds, this technique presumably evaluates the integrity of the large upper motor neurons and the most rapidly conducting axons in the corticospinal system. Magnetic stimulation has been found to be effective and reproducible, with few or no adverse side effects. In practice, a stimulation coil is placed over the scalp or cervical spine and is used to excite upper motor neurons or motor axons. Recording electrodes are placed over various muscles, and the amplitude and latency of the response are recorded. Absent, altered, or delayed motor responses are seen when there is damage to the upper motor neuron, to its axon, or to its myelin sheath.


Electromyography is concerned with the study of the electrical activity arising from muscles at rest and those that are actively contracted.

Clinical Applications

Electromyography is a particularly useful aid in diagnosing lower-motor-neuron disease or primary muscle disease and in detecting defects in transmission at the neuromuscular junction. Although it can be very helpful, the test does not usually give a specific clinical diagnosis; information from the electromyogram (EMG) must be integrated with results of other tests, including muscle enzyme levels, muscle biopsy if necessary, and clinical features, to arrive at a final diagnosis.


Human striated muscle is composed functionally of motor units in which the axons of single motor cells in the anterior horn innervate many muscle fibers. (Although the size of motor units varies from muscle to muscle, in the largest motor units hundreds of muscle fibers may be innervated by a single axon.) All the fibers innervated by a single motor unit respond immediately to stimulation in an all-or-none pattern, and the interaction of many motor units can produce relatively smooth motor performance. Increased motor power results from the repeated activation of a given number of motor units or the single activation of a greater number of such units.

The action potential of a muscle consists of the sum of the action potentials of many motor units; in normal muscle fibers, it originates at the motor end-plates and is triggered by an incoming nerve impulse at the myoneural junction. Clinical studies indicate that normal muscle at rest shows no action potential. In simple movements, the contracting muscle gives rise to action potentials, whereas its antagonist relaxes and exhibits no potentials. During contraction, different portions of the same muscle may discharge at different rates, and parts may appear to be transiently inactive. In strong contractions, many motor units are simultaneously active, producing numerous action potentials.


Stimulation is usually applied over the course of the nerve or at the motor point of the muscle being tested. Muscles should always be tested at the motor point, which is normally the most excitable point of a muscle in that it represents the greatest concentration of nerve endings. The motor point is located on the skin over the muscle and corresponds approximately to the level at which the nerve enters the muscle belly.

A concentric (coaxial) needle or a monopolar solid-steel needle is inserted at the motor point of a muscle and advanced by steps to several depths. Variations in electrical potential between the needle tip and a reference electrode (a metal plate) on the skin surface are amplified and displayed. The electrical activity can be displayed on a computer screen. Observations are made in each area of the electrical activity evoked in the muscle by insertion and movement of the needle: the electrical activity of the resting muscle with the needle undisturbed, and the electrical activity of the motor units during voluntary contraction (Fig 23–4). Because various muscle fibers may respond differently, several insertions of the needle into different parts of a muscle may be necessary for adequate analysis.


FIGURE 23–4 Action potentials in electromyography. A: Nerve potential from normal muscle; B: fibrillation potential and C: positive wave from denervated muscle; D: high-frequency discharge in myotonia; E: bizarre high-frequency discharge; F: fasciculation potential, single discharge; G: fasciculation potential, repetitive or grouped discharge; H: synchronized repetitive discharge in muscle cramp; I:diphasic, J: triphasic, and K: polyphasic motor unit action potentials from normal muscle; L: short-duration motor unit action potentials in progressive muscular dystrophy; M: large motor unit action potentials in progressive muscular dystrophy; N: highly polyphasic motor unit action potential and short-duration motor unit action potential during reinnervation. Calibration scale (vertical) in microvolts. The horizontal scale shows 1000-Hz waveforms. An upward deflection indicates a change of potential in the negative direction at the needle electrode. (Reproduced, with permission, from Clinical Examinations in Neurology, 3rd ed. Members of the Section of Neurology and Section of Physiology, Mayo Clinic and Mayo Foundation for Medical Education and Research, Graduate School, University of Minnesota, Rochester, MN. WB Saunders, 1971.)

Types of Activity

Insertional activity refers to the burst of action potentials that is usually observed when the EMG needle is inserted into the muscle. In normal muscle, insertional activity is short lived, and there is usually electrical silence after the initial burst of insertional activity. Increased insertional activity is observed in denervated muscles and in many forms of muscle disease.

Motor unit potentials (MUPs) are also examined by EMG and provide important information about innervation (or denervation) of the muscle fibers within a muscle. The MUP in any given muscle has a characteristic size and duration. If lower motor neurons, roots, or nerves are injured so that motor axons are severed and muscle fibers are denervated, the number of MUPs appearing during contraction is decreased. Nevertheless, the configurations of the remaining MUPs are usually normal. The decreased number of MUPs reflects denervation of some of the muscle fibers. Later, there may be reinnervation of the previously denervated muscles as a result of sprouting of new motor axon branches from undamaged axons, whose motor units increase in size. As a result, the MUPs increase in amplitude and duration and in some cases become polyphasic. These polyphasic MUPs provide evidence of reinnervation (and thus implies prior denervation) and can have diagnostic value, providing evidence of disease involving motor neurons or their axons in the ventral roots or peripheral nerves.

Two types of spontaneous or ongoing activity observed by EMG have particular significance. The term fibrillation is reserved for spontaneous independent contractions of individual muscle fibers so minute that they cannot be observed through the intact skin. Denervated muscle may show electromyographic evidence of fibrillations that are most pronounced for 1-3 weeks and that can persist for months after losing its nerve supply. Fasciculations, or twitches, in contrast, can be seen and palpated, and they can be heard with the aid of a stethoscope; they represent contractions of all (or most) of the muscle fibers of a motor unit. Spontaneous fasciculations can vary because of the length and number of muscle fibers involved; they usually result from disorders of the lower motor neuron or its axon. Benign fasciculations, such as those from exposure to cold or temporary ischemia (eg, caused by crossed legs), are unassociated with other clinical or electrical signs of denervation (see Fig 23–4).

In a complete nerve lesion, all of the motor axons are severed, so that fibrillation potentials occur without MUPs; partial nerve lesions show both fibrillation and motor unit activity from voluntary muscle contraction. Diminution or cessation of fibrillation potentials and the appearance of small, disintegrated motor unit action potentials occur with nerve regeneration. Fibrillations in a paretic muscle are increased by warmth, activity, and neostigmine; they are decreased by cold or immobilization.

After complete section of a nerve, denervation fibrillation potentials are evident (after about 18 days) in all areas of the muscles supplied by a peripheral nerve. Some motor unit discharges persist in partial nerve injuries despite the clinical appearance of complete paralysis. Mapping the areas of denervation fibrillation potentials aids in the diagnosis of single nerve root disorders and spinal nerve root compression.

Repetitive Stimulation

In the absence of pathologic conditions, axons can conduct impulses at a high frequency, and the neuromuscular junction can faithfully follow these high-frequency impulses, producing a surface muscle action potential that retains its amplitude with rates of stimulation up to 20-30 Hz for up to 1 minute. In contrast, in myasthenia gravis, the response is decremental; the MUP decreases in amplitude after several stimuli at rates as low as 3 or 4 Hz. The Lambert-Eaton myasthenic syndrome exhibits a different pattern; in this disorder, there is a defect of neuromuscular transmission characterized by incrementalresponses, which increase in amplitude with repetitive stimulation. These distinct patterns of response to repetitive stimulation are of considerable diagnostic value.

Single-Fiber EMG

Single-fiber EMG (SFEMG) permits the recording of action potentials from single muscle fibers using very fine electrodes. This technique permits the measurement of muscle fiber density within a given motor unit and thus can be of significant value in the diagnosis of muscle disorders. Jitter (variability in the timing of action potentials for single muscle fibers comprising a given motor unit) can also be studied with this technique. Jitter appears to result from abnormalities of the preterminal part of the axons close to the neuromuscular junction. Single-fiber electromyography may be especially useful for the diagnosis of disorders involving motor neurons (eg, amyotrophic lateral sclerosis) and the neuromuscular junction.


As noted in Chapter 3, myelination increases the conduction velocity (the speed of action potential transmission) along axons. Conversely, damage to the myelin (demyelination) results in a decrease in conduction velocity. Damage to the axon or axonal degeneration, on the other hand, results in loss of ability of the axon to conduct impulses. These physiological changes can be measured in nerve conductions studies.

By stimulating peripheral nerves with electrodes placed on the skin and recording muscle and sensory nerve action potentials, it is possible for one to examine conduction velocities, distal latencies, and amplitudes of responses, which provide important information about the functional status of the myelinated axons within a peripheral nerve. These studies can be helpful in determining whether peripheral nerves have been affected and, if so, help to determine the pathologic process involved (eg, demyelination vs. axonal injury).

For these studies, surface electrodes are placed on the skin for stimulation of accessible peripheral nerves, and the resulting compound action potential is recorded elsewhere over the nerve or over a muscle that is innervated by the nerve being studied. Two stimulation sites are usually used so that conduction velocity can be ascertained (by dividing the distance between the two stimulation sites by the difference in conduction times). These whole-nerve conduction velocities measure the properties of the fastest conducting (and largest) axons within the nerve, that is, the myelinated axons and have normal values of more than 40 m/s in adults.

Nerve conduction studies, as carried out in clinical settings, do not assess the function of slow-conducting, nonmyelinated axons, and thus cannot detect damage to these small axons, as occurs in the small fiber neuropathies. These can be diagnosed on the basis of the clinical picture of small fiber dysfunction (pain and autonomic dysfunction) and by confirmation of damage to small axons as seen in skin biopsy (which permits visualization of the distal tips of small nerve fibers in the epidermis).

Decreased conduction velocities are seen in peripheral neuropathies characterized by demyelination (eg, Guillain–Barré syndrome, chronic inflammatory demyelinating polyneuropathy, and Charcot-Marie-Tooth disease). Slowed conduction velocities are also seen at sites of focal compression.

Measurements of amplitude, of either the muscle action potential elicited by motor axon stimulation or the sensory nerve action potential, can also provide useful information. Reduction of amplitude is especially pronounced in disorders characterized by loss of axons (eg, uremic and alcoholicnutritional neuropathies). The presence, absence, or reduction of innervation can be determined by electrical stimulation of peripheral nerves, and the location of a nerve block can be shown. Anomalies of innervation can be detected by noting which muscles respond to nerve stimulation, and abnormal fatigability after repeated stimulation of the nerve can be noted.

In the presence of paralysis, a normal response of innervated muscles to stimulation of the peripheral nerve shows that the cause of paralysis is proximal to the stimulated point. Alternatively, an absent or weak response suggests further testing to detect the site and nature of the defect, which probably includes pathology distal to the stimulation site.

H-Reflexes and F-Wave

Nerve conduction studies provide information about the status of distal segments of peripheral nerves in the limbs but not about conduction within proximal parts of the nerve or spinal roots. The H-reflex and F-wave involve conduction through spinal roots and proximal parts of peripheral nerve and thus provide important diagnostic information about disorders that involve these areas. To elicit the H-reflex, submaximal stimuli are applied to mixed (motor-sensory) nerves at an intensity too low to produce a direct motor response. These stimuli evoke a muscle contraction (H-wave) with a relatively long latency because of activation of Ia spindle afferent fibers, which travel via the dorsal roots to the spinal gray matter, where they synapse with lower motor neurons, whose action potentials then travel through the ventral roots and then to the muscle. Absence of the H-reflex suggests pathologic conditions along this pathway and is often a result of radiculopathies (disorders involving peripheral nerves) or polyneuropathies involving spinal roots or proximal parts of the peripheral nerves (eg, Guillain-Barré syndrome).

The F-wave is a long-latency response, following the direct muscle potential, that is evoked by supramaximal stimulation of motor-sensory nerves. It is produced by antidromic (retrograde) stimulation of motor axons, which results in invasion of action potentials into their cell bodies in the spinal cord and evokes a second (reflected) action potential that travels along the motor axon to muscle. As with the H-reflex, absence of the F-wave implies pathologic conditions of spinal roots or proximal parts of peripheral nerves.


American EEG Society: Guidelines in EEG and evoked potentials. J Clin Neurophysiol 1986;3(Supp 1):1

Aminoff MJ: Electromyography in Clinical Practice, 2nd ed. Churchill Livingstone, 1987.

Chiappa KH: Evoked Potentials in Clinical Medicine, 3rd ed. Lippincott-Raven, 1997.

Engel J, Pedley TA: The Epilepsies. Lippincott-Raven, 1997.

Kimura J: Electrodiagnosis in Disease of Nerve and Muscle, 2nd ed. FA Davis, 1989.

Niedermeyer E, daSilva FL: Electroencephalography, 3rd ed. Williams & Wilkins, 1994.

Oh SJ: Clinical Electromyography and Nerve Conduction Studies, 2nd ed. Williams & Wilkins, 1997.

If you find an error or have any questions, please email us at Thank you!