Chapter 105 Visual Evoked Potential GARY L. TRICK and BARRY SKARF Table Of Contents |
METHODS VISUAL STIMULATION EVALUATING THE VEP WAVEFORM CLINICAL APPLICATION OPTIC NEUROPATHIES VISUAL DEVELOPMENT AND AGING ALBINISM SUMMARY REFERENCES |
The visual evoked potential (VEP), also known as the visual evoked response (VER) or
the visual evoked cortical potential (VECP), is an electrophysiologic
signal that can be recorded from the human scalp and is
generated by neurons in the brain in response to visual stimulation. The
VEP was initially described by Adrian and Mathews1 in 1934 and has been used in clinical and research laboratories for almost 40 years. For
many years, there were no standard methods for VEP
recording or analysis. Various laboratories recorded the VEP using dissimilar
techniques, and consequently, there were many divergent opinions
concerning the neurophysiologic basis of the components of the response, the
clinical usefulness of the technique, and the abnormalities
associated with specific diseases. This problem was exacerbated in that
many laboratories recorded VEPs without obtaining an accurate ophthalmologic
examination of the patient. As a result, the exact visual status
of the patient and the influence of the patient's visual status
on the test results were unknown. It is now well established that VEPs
should always be performed with the patient's appropriate refractive
correction because uncorrected refractive error will confound
the results. To address these problems, the International Society for Clinical Electrophysiology of Vision published a recommended standard for VEP measurement and reporting.2 People interested in recording or evaluating VEPs, either clinically or experimentally, are urged to read and follow theguidelines. This chapter presents an overview of VEP methodology and its application in evaluating the integrity of the retinocortical pathway and detecting visual disorders in the clinical setting. More detailed discussions can be found elsewhere.3–7 |
METHODS |
Measuring a VEP and using this technique clinically became practical after
the development of the signal averaging computer. This instrument
improves the signal-to-noise ratio of the bioelectrical activity generated
by visually active cortical neurons and recorded from scalp electrodes. The
signal-to-noise ratio is the ratio of the cortical visual response
to the underlying electroencephalograhic (EEG) activity.8,9 The VEP is triggered by, and time-locked to, a precisely defined and repetitive
visual stimulus, whereas the EEG signal is random in relation
to the visual stimulus. The signal averaging computer averages all the
time-locked activity (signal) recorded in a defined interval, whereas
the underlying EEG activity (noise) that is not time-locked is canceled
out. Consequently, signal averaging improves the reliability and
consistency of responses and makes experimental and clinical studies feasible. Nevertheless, it
is important to recognize that artifacts that
are time-locked to the stimulus still can distort the VEP. To record the VEP, one or more standard silver-silver chloride or gold disc EEG electrodes are fixed to the scalp with collodion glue or conducting paste. The recording electrode is placed on the patient's scalp after cleaning the scalp with detergent or ethanol. The impedance of the scalp-electrode connection should be kept below 5 kΩ. Improper skin preparation or drying of the conducting paste will increase impedance and distort the morphology of the VEP waveform. Most laboratories use a single active electrode placed approximately 2 cm above the inion, the ridge that runs just above the joint with the neck at the back of the skull. When recording hemispheric differences, however, multichannel recordings are necessary with additional active electrodes placed laterally over each occipital lobe. In most cases, bipolar recordings are used with a common reference electrode placed on one or both ears, or on the forehead. A ground that is electrically zero cannot be found on the body, but for practical purposes, the forehead or arm provides an acceptable site for the ground electrode. In recording, the VEP signals from the active and reference electrodes are fed into a differential amplifier (Fig. 1) that rejects electrical activity that is common to both electrodes, assuming it is noise that is interfering with the relevant visually driven signals. Additional signal conditioning is achieved with high- and low-bandpass filters. Filter settings must be carefully chosen to avoid selecting too narrow a band and losing a portion of the relevant signal. Filtering also can alter the recorded latency of the response. Faulty electrical connections or improperly placed ground or reference electrodesmay allow 60-cycle activity to distort the waveform morphology. In addition, electroretinogram (ERG) potentials can contaminate signals recorded from VEP electrodes if they are placed too far anteriorly. The EEG alpha activity and myogenic potentials from the face, scalp, and neck that occur in synchrony with the VEP will distort the recorded waveform, whereas artifacts that are not time-locked to the stimulus, such as eye blinks and patient movement, cause less interference. |
VISUAL STIMULATION |
The VEP waveform highly depends on the rate of stimulus presentation. Two
distinct waveform morphologies are associated with low and high temporal
frequency stimulation: transient and steady-state responses, respectively. To
obtain a transient response, visual stimuli are presented
at a temporal frequency low enough to permit the brain activity to decrease
to baseline levels between responses (often two to four reversals
per second for a patterned stimulus or 1 to 2 Hz for a diffuse flash). The
transient VEP has a waveform with discrete negative and positive
deflections, each with a measurable latency from the stimulus onset (Fig. 2). The second type of response, termed the steady-state VEP, is elicited
by presenting stimuli at a temporal frequency above approximately five
reversals per second for a pattern stimulus or 5 Hz for a diffuse flash. The
steady-state VEP has a more sinusoidal waveform (Fig. 2). Although discrete waveform data are lost, latency can be determined
by calculating the shift in phase of the second harmonic of the response.9 In the clinic, the transient VEP is used more frequently than the steady-state
response. The VEP waveform morphology also highly depends on the type of stimulus presentation. Three predominant types of visual stimuli are used to elicit VEPs: diffuse flash, pattern reversal, and pattern onset-off. The responses obtained with each type of stimulus have a distinct characteristic waveform (Fig. 3). THE FLASH VISUAL EVOKED POTENTIAL The earliest VEP studies used photic flashes produced by photostimulators to generate spatially unstructured or diffuse flash stimuli.10,11 Original descriptions of the waveform of the VEP obtained with flash stimuli disagreed concerning many of the waveform characteristics of the response. These discrepancies were probably the result of various factors, including differences in the electrode positions, the flash intensities, and the temporal frequencies of the stimuli used in different laboratories. However, there is now reasonably good agreement that the flash VEP waveform consists of a complex series of waves with positive and negative peaks that are designated in numerical order (Fig. 3). The most prominent and reproducible of these peaks are the second negative peak (N2), with a normal latency of approximately 70 to 90 milliseconds, and the second positive peak (P2), with a normal latency of approximately 100 to 120 milliseconds. It is also now well established that the amplitude of the VEP to diffuse flashes varies with the temporal frequency of stimulus presentation. The largest amplitude is obtained with flashes that are presented at approximately10 Hz. In the clinic, the flash VEP is most often usedwhen a pattern VEP cannot be obtained. Patients with poor visual acuity, media (cornea, lens, vitreous) opacities, poor fixation (including infants), and uncorrected refractive error may have a measurable flash VEP that can be helpful in the assessment. The amplitude of the flash VEP depends on the luminance of the flash stimulus. For the flash VEP, a luminance of 1.5 to 3.0 candelas/m2 is generally recommended. A decrease in the brightness of the stimulus reaching the retina, produced by miosis (accommodative or pharmacologic), neutral density filters, or, rarely, dense cataracts results in latency delays (i.e., increased latency). This is also true for pattern VEPs. A special variant of the flash VEP is the multifocal VEP.12,13 The stimulus for the multifocal VEP is a changing pattern of discrete flashes, usually made of 61 to 241 hexagons covering up to ±25° of the visual field rather than the single full-field flash used in conventional flash VEP. The flash position changes in a pseudorandom fashion according to so-called m-sequences that guarantee that no stimulus sequence is repeated during an examination and that all stimulus sequences appear only once. This is necessary for the multiplexed signal to be separated into signals for every test location. As a result, the multifocal VEP produces a spatially discrete pattern of responses and an underlying mathematical system that derives the true signal from each location, rather than a single stimulus and a single response. THE PATTERN VISUAL EVOKED POTENTIAL The development of techniques that use patterned stimuli to elicit bioelectrical signals from the visual cortex was a methodologic advancement that greatly increased the clinical usefulness of the VEP. This approach takes advantage of the functional organization of the proximal retina and visual cortex, which is preferentially sensitive to spatially patterned stimuli. Two types of patterned stimuli, contrast-reversing or onset-offset patterns, are generally used to record VEPs. Pattern-reversal VEPs are recorded using a uniform repetitive checkerboard or bar grating pattern in which the pattern elements are contrast-reversed in a fixed sequenced. The overall luminance of the display and the contrast between the pattern elements remain constant throughout the entire recording. The waveform of the pattern-reversal VEP (Fig. 3) normally consists of three principal features: an initial negativity with a latency of approximately 70 to 80 msec (N75), a larger positive component with a latency of approximately 90 to 110 msec (P100), and a large negative component with a latency of approximately 130 to 140 msec (N135). In visually normal adults, the maximal pattern-reversal VEP is evoked with a check size between 10 and 20 minutes.14,15 Smaller or larger checks evoke smaller amplitude pattern VEPs. Larger checks may elicit responses that confound pattern and luminance contributions.15 For the pattern onset-offset VEP, the elements in the pattern are abruptly alternated with a spatially unstructured field of identical space-averaged luminance so that when the pattern appears or disappears there is no change in mean luminance. Generally, the pattern is off (diffuse background) for longer than it is on. Typically, the pattern onset-offset VEP has three prominent components designated C1, C2, and C3 (Fig. 3). C1 is the initial positive component and has a latency of approximately 70 to 80 milliseconds. C2 is a large negative component and has a latency of approximately 100 to 120 milliseconds. C3 is the second major positive component and has a latency of approximately 140 to 150 milliseconds. The pattern onset-offset approach can be useful in distinguishing the pattern and luminance contributions in the pattern VEP.16 The VEP obtained with pattern stimulation, either pattern reversal or pattern onset-offset, is dominated by contributions from the macula.5,17 This is not surprising because the central few degrees of vision have a large representation in the striate area and are located posteriorly in the occipital lobe, closer to the surface electrodes. Checks smaller than 20 minutes evoke a large pattern VEP when the field is limited to the macula field.18 The peripheral retinal contribution to the pattern VEP is the result of pattern and luminance changes rather than a true pattern response.7,18 Large check stimulation of the retina peripheral to the macula may evoke a large amplitude VEP because of luminance components. Checks and gratings are the patterns most frequently used as pattern VEP stimuli. A check is a complex pattern stimulus with spatial frequency power along each diagonal and in the direction of the horizontal and vertical edges. Mathematical analysis of the spatial frequency in each direction of power shows a fundamental spatial frequency with higher harmonics because the sharp borders can be described mathematically only as a square wave.19 In comparison, sinusoidal gratings have gradually appearing and disappearing borders and are described by a single-frequency (fundamental) sinusoidal function with power along only one axis. A sine wave grating of high, medium, or low spatial frequency (fine to coarse) probably excites those neurons in the visual cortex tuned to that spatial frequency20 and another group of neurons that are tuned to the axis or orientation of the grating.21,22 Sine wave grating pattern VEP studies that measure sensitivity to gratings oriented horizontally, vertically, and at 135° and 45° axes using high, medium, and low spatial frequencies may uncover subtle defects in the visual system that are missed with checkerboard pattern VEP. Doing so, however, is impractical in clinical practice. In addition, large amounts of astigmatism, even if corrected with a lens, may depress the pattern VEP at one orientation of the stimulus because of meridional amblyopia.23 When pattern stimulation is used, refractive errors must be corrected so that the pattern is not degraded at the retina.14 Some laboratories attempt to circumvent the problems associated with image degradation by using large pattern elements. However, this does not avoid blurring of the edges of the elements, and therefore, this technique is not optimal. For pattern stimulation, mean luminance should be at least 40 candelas/m2. More importantly, however, the contrast of the check or grating pattern should be 75% or higher, especially for clinical studies.2 The contrast of a pattern is defined as: 100 × (Lmax - Lmin)/Lmax + Lmin), where Lmax = the luminance of the light checks or gratings, and Lmin = the luminance of the dark checks or gratings. For pattern-reversal and pattern onset-offset VEPs, the amplitude of the responses increases as contrast is increased.24–27 The latency increases as contrast is decreased.28,29 |
EVALUATING THE VEP WAVEFORM |
Because the shape of the VEP waveform (i.e., the waveform morphology) can vary significantly within and between patients, most experimental and clinical studies limit consideration to two more reliable features of the waveform: response amplitude and latency (sometimes referred to as implicit time). VEP amplitude is a measure of the size of the response, whereas VEP latency reflects transmission and processing time in the retina-to-occipital cortex pathway after the onset of stimulation. Decreases in VEP amplitude generally result from a reduction in the number or sensitivity of neurons in the retinocortical pathway. Increases in VEP latency may reflect defective retinal transmission, defects in the myelin of the visual pathway from the optic nerve to the occipital cortex, or even synaptic abnormalities in the retina or occipital cortex. VEP amplitude measurements are typically made between the peaks and troughs of the waveform components. Response timing is quantified in terms of the latency of the response. This is a measurement of the time from the onset of the stimulus presentation to the peak of the appropriate waveform component. Amplitude and latency measurements are particularly appropriate for characterizing the transient responses. For steady-state responses, more complex analysis procedures are necessary. The most popular of these is the use of Fourier analysis to quantify the power and phase of the harmonic components of the signal. |
CLINICAL APPLICATION OPTIC NEUROPATHIES |
Optic neuritis and multiple sclerosis (MS) have been studied extensively
with the VEP. Initial studies described a decrease in amplitude and
an increase in latency of the flash VEP in MS.30–32 Halliday and coworkers33–37 described prolonged pattern VEP latency in a high percentage of patients
with optic neuritis. Furthermore, Halliday and coworkers found that
the pattern VEP is frequently abnormal in patients with MS who have no
other clinical evidence of optic nerve disease. Since those seminal
studies, patients with optic neuritis and MS have been examined using
various VEP techniques, and different defects have been observed. Methodologic
differences have contributed to this disagreement, but the diverse
findings probably are also related to the heterogeneous nature and
extent of the pathophysiologic changes in the study groups. Variations
in the location and age of plaques, number of areas affected in the
visual pathway, and specific fiber sizes affected all may have contributed
to this discrepancy. Nevertheless, there is now a general agreement
that testing with pattern stimuli will elicit a significantly delayed
VEP (Fig. 4) in approximately 84% of patients with definitive MS, 65% of patients
with probable MS, and 42% of patients with possible MS.38–40 In patients with MS and a history of optic neuritis, this figure is approximately 92%, and
for patients with visual acuity of 20/20 and no definite
history of optic neuritis, this figure is approximately 40%.27 The changes in response latency that occur in optic neuritis and MS are
generally attributed to conduction delays that increase as transmission
is slowed with increasing body temperature41 and diminish with cooling of the body.42 When pattern-reversal VEPs are obtained with sinusoidal gratings that
stimulate subsets of cortical neurons tuned to specific orientations,22,24 the sensitivity of the VEP for detecting thes defects is greater than
when checks are used. Studies using sinusoidal gratings to elicit VEPs
in patients with MS suggest an increased frequency of delayed responses.43 Changing the grating orientation to evoke VEPs to vertical, horizontal, and
oblique gratings shows orientation-dependent delays.44,45 Amplitude abnormalities also can occur in patients with MS who have no significant vision loss, but are less common than latency increases.46 During the acute stage of an optic neuritis attack, when vision is at the level of light perception, the pattern VEP may be immeasurably small.47 If visual acuity is good, however, the VEP is typically recordable but delayed. Response latency may normalize during the recovery period.38 Psychophysically determined contrast sensitivity losses selective for specific spatial frequencies48,49 are found in patients with MS and no clinically apparent optic neuropathy. Pattern-reversal VEPs obtained with coarse, intermediate, and fine gratings (low to high spatial frequency) appear to be correlated with psychophysical contrast sensitivity function defects in MS.50 VEP contrast threshold elevations show losses with orientation and spatial frequency selectivity in patients with MS.51 Despite recovery to 20/20, a careful neuroophthalmologic examination of a patient with MS may show an afferent pupillary defect, color vision loss, disc pallor, or nerve fiber layer losses in almost all patients with histories of visually significant optic neuritis. If both eyes have been affected, no afferent pupillary defect may be seen. Subtle change in the disc and nerve fiber layer may be difficult to distinguish in MS with no definitive history of optic neuritis. In these cases, the pattern-reversal VEP provides objective information about conduction along the retinocortical pathway. In a patient without a history of optic neuritis and clinically symptomatic disease involving the white matter outside the visual sensory pathway, the VEP is particularly useful in establishing a diagnosis of MS. Unfortunately, other neurodegenerative conditions also can cause a delay of the pattern VEP and lead to confusion in the diagnosis. Therefore, VEP findings in these patients must be interpreted with caution and always in the context of a complete ophthalmologic and neurologic examination. Diseases that inflame the optic nerve or preferentially cause demyelination are often associated with a delayed latency. Optic neuropathy or subclinical anterior visual pathway involvement by sarcoidosis can delay the pattern VEP.52 In Leber's optic neuropathy, there is devastating vision loss, including loss of visual acuity, color vision defects, and visual field abnormalities.53,54 The earliest VEP abnormalities in Leber's optic neuropathy appear to be increases in P100 latency or changes in the waveform morphology (i.e., the development a double positive peak). As the condition progresses, the VEP amplitude decreases to a point where responses become immeasurable. Patients with thyroid ophthalmopathy may have a prolonged latency of the pattern VEP before a clinically apparent optic neuropathy.55 Subacute combined degeneration secondary to vitamin B12 deficiency causes demyelination and prolongs the pattern VEP latency,56,57 even with an unremarkable neuroophthalmologic examination.58 Not all causes of optic nerve lesions are associated with delayed conduction and a consequent increase in latency. The VEP has been used extensively to investigate many of these optic nerve disorders.59 Drusen of the optic disc, even when they cause field defects, rarely prolong the P100 of the pattern VEP,60 but in these cases, the early positive waves of the flash VEP may be distorted.61 In anterior ischemic optic neuropathy, the amplitude of the flash or pattern VEP is often depressed,62,63 whereas prolonged latency is observed less frequently and may be limited to patients with the arteritic form of the disease.64 Similar findings are noted with the pattern VEP in patients with an unknown optic atrophy and tobacco or alcohol optic neuropathy.65,66 GLAUCOMA Glaucoma is a slowly progressive anterior optic neuropathy in which irreversible visual dysfunction results from damage to the retinal ganglion cells and their axons. There have been numerous studies of the VEP in patients with glaucoma and glaucoma suspects, but the results have not been particularly impressive. Decreases in pattern VEP amplitude and prolonged VEP latencies are found in many patients with glaucoma and some glaucoma suspects.67–69 Steady-state VEPs appear to be more sensitive for detecting glaucomatous damage thantransient responses.68 VEPs elicited by sinusoidal gratings may be prolonged in patients with increased intraocular pressure without field loss.69 Unfortunately, it remains unclear whether patients with ocular hypertension who later develop glaucoma consistently have VEP abnormalities before developing visual field defects. The pattern ERG is an objective electrophysiologic method that may be more useful in detecting and following glaucoma-related ganglion cell loss.70,71 In addition, several groups have recently begun to examine the use of the multifocal VEP for objective perimetry.72–76 Initial results from these studies suggest that it may be possible to use this technique to detect and monitor visual field defects in patients with glaucoma and glaucoma suspects. CHIASMAL AND RETROCHIASMAL LESIONS Multichannel VEPs to hemifield stimulation can be used to detect dysfunction associated with chiasmal lesions attributable to pituitary tumors, aneurysms, inflammation, demyelination, and trauma. Halliday and coworkers77 described the marked asymmetry in the distribution of scalp potentials that is characteristic of chiasmal lesions. This is evident as the VEPs from one eye being more abnormal over one hemisphere, whereas the VEPs from the opposite eye are more abnormal over the other hemisphere. When large checks are used, the maximal abnormality is located ipsilateral to the visual field defect. When small checks and a small stimulating field are used, the maximal abnormality is located contralateral to the visual field defect. The flash and the pattern VEP have a role in the electrophysiologic examination of patients with unilateral retrochiasmal lesions, particularly in patients who are unable to fixate or concentrate. In unilateral lesions, when abnormalities in the flash VEP are present, they are localized to the affected hemisphere (i.e., contralateral to the visual field defect). Interpretation must be cautious, however, because flash stimulation illuminates the whole retina and evokes responses from both hemispheres. Furthermore, approximately 50% of healthy patients will have hemispheric asymmetries in the flash VEP.78 Consequently, a unilateral retrochiasmatic lesion may be difficult to detect using the flash VEP,79,80 and pattern VEP abnormalities often are more apparent than changes in the flash VEP. The typical pattern VEP finding in unilateral retrochiasmal lesions also is an asymmetry in the distribution of the scalp potentials. With the steady-state pattern VEP, responses measured over the affected hemisphere typically have a lower amplitude.81 A more significant diminution in the amplitude of the responses from electrodes in all locations is found if only the affected hemifield is stimulated. This is true even if there is macular sparing. Hemifield stimulation is also effective when using the transient PVEP, especially when responses to full-field stimulation do not consistently indicate the affected hemisphere.82 The largest amplitude P100 potential to full-field stimulation may be found with electrodes located over the occipital cortex ipsilateral to the field defect.83,84 This can be explained if the vector of the electrical potential of the major positive wave to half-field stimulation, originating in the contralateral cortex, is oriented toward the ipsilateral scalp electrodes.85,86 Inconsistencies in the VEP, even with the use of pattern stimuli, may result from vari-ability in the amount of exposed visual cortex ori-ented toward the surface elctrodes between hemispheres,87 nonuniformity of a macular contribution to the VEP,88 and anatomic variation of the macula location. In addition, a hemifield stimulus may become a small full-field stimulus if the patient's fixation shifts. Patients with bilateral retrochiasmal lesions secondary to hydrocephalus may also have VEP abnormalities. Patients with hydrocephalus and a swollen optic disc may have a delayed response in the high temporal frequency flash VEP89 or in the transient pattern VEP to sine wave gratings as long as the hydrocephalus is present. Relieving the hydrocephalus normalizes VEP latency in these patients. Patients with bilateral occipital lobe dysfunction can have diminished visual acuity, defective color vision, visual field loss, or even total blindness. Other than an anatomic demonstration of the extent of the lesion by computed tomography, magnetic resonance imaging, or positron emission tomography, few objective findings of the visual deficit may be present. In these patients with severe vision loss, VEP abnormalities are likely. However, the VEP may not provide an objective measure of the extent of the visual loss. In older children and adults, as long as subjective vision is present, even if it is poor, a flash VEP may be measured,90–93 although it is almost always abnormal in children with cortical blindness.94 Nevertheless, the flash VEP, if measurable, can be helpful in predicting the return of vision in infants, even if they have cortical blindness, because the presence of a flash VEP in this setting often suggests the eventual return of behavioral vision after reversal of the underlying disease, such as a ventricular shunt for hydrocephalus.95–97 DEFECTS RELATED TO NEUROTRANSMITTER ALTERATIONS The psychiatric literature contains many reports of VEP abnormalities in numerous psychiatric disorders, often associated with alteration in catechol-amine metabolism.98 Some movement disorders caused by basal ganglia dysfunction have been shown to have specific neurotransmitter deficien-cies. These diseases have no known associated structural defects in the visual system, but the patients often have VEP abnormalities. Huntington's disease, which has a known gamma-aminobutyric acid deficiency, is associated with VEPs that have a small amplitude, abnormal waveform, and normal latency.99,100 Drugs that increase gamma-aminobutyric acid do not alter the VEP and make it unlikely that the VEP abnormalities are simply a reflection of a neurotransmitter abnormality.101,102 A significant delay of the P100 component of the pattern VEP is known to occur in Parkinson's disease, and this increase in latency appears to correlate with the severity of the disease.103 This abnormality may be normalized in patients treated with levodopa or carbidopa.104 Drugs that block dopaminergic receptors, such as phenothiazines, prolong the pattern VEP in patients with schizophrenia.105 The only known site of dopaminergic neurotransmission in the visual system is in the retina. Although no definite structural retinal abnormalities are known in Parkinson's disease,106 the dopamine metabolism in the retina appears abnormal.107 RETINAL DISEASES When used in conjunction with the ERG, and particularly the focal ERG, the VEP can be useful in attributing vision loss to retinopathies or optic neuropathies. Any condition that affects central vision is likely to perturb the VEP, especially when small checks are used. This is particularly true for macular diseases.108–111 Therefore, the VEP can provide an independent and objective means of measuring macular function. Abnormal VEPs have been reported with macular holes and cysts, retinal detachments, and macular degeneration.112 In central serous retinopathy, the pattern VEP is often prolonged but returns to a normal latency when the disease has improved clinically.113–115 Abnormal pattern VEP latencies are observed in patients with diabetes who have normal visual acuity and nodetectable retinopathy,116 but it remains unclear whether this changes reflects subclinical diabetic retinopathy or diabetic optic neuropathy. The VEP is rarely abnormal in patients with retinitis pigmentosa or other retinal degenerative diseases that primarily affect the peripheral retina and spare the macula until later stages of the disease process.117 When used in conjunction with the ERG, and in particular the focal ERG, the VEP is clinically useful in distinguishing retinal and optic nerve sites of dysfunction in cases with subtle visual disturbances and questionable funduscopic defects. OTHER DISEASES Visual evoked potential abnormalities have been reported in various disorders that affect the peripheral or central nervous system without definite visual system abnormalities. Myotonic dystrophy118 and Charcot-Marie-Tooth disease119 may be associated with latency and amplitude abnormalities of the pattern VEP. Delayed flash and pattern VEPs occur in patients with chronic renal failure.120,121 Inhalation anesthetics diminish the amplitude but do not prolong the latency of the pattern VEP.122 Degenerative diseases that affect the spinal cord, cerebellar pathways, or both, such as Friedreich's ataxia, Huntington's disease, neurosyphilis, and AIDS, also can affect the optic nerves and cause visual defects, including a delay of the pattern VEP.123–128 As many as two thirds of patients with Friedreich's ataxia have VEP abnormalities, including small, delayed, or absent responses.123,125,126 As mentioned previously, this can also occur in patients with a vitamin B12 deficiency. Prolonged latencies can also be found with hereditary paraparesis130 or spinocerebellar degeneration.131,132 In the group of hereditary progressive diseases collectively known as neuronal ceroid lipofuscinosis, VEP abnormali-ties are frequently encountered.133,134 These include decreases in amplitude, increases in latency, and other distortions of waveform morphology. HYSTERIA AND MALINGERING The VEP can be useful for detecting spurious vision loss in cases of hysteria and malingering that are difficult to assess clinically. In these cases, the VEP will be normal as long as special care is taken to ensure that the patient fixates on the stimulus and remains attentive. The patient must be tested while wearing the optimal refractive correction for the test distance, and the examiner must be acutely aware of the pitfalls in the technique to avoid describing the VEP as abnormal or absent simply because of confounding factors. With the pattern VEP, a delayed latency can occur because of excessive accommodation,135,136 particularly when a small field of stimulation is used. The small pupil associated with accommodation decreases the luminance of the stimulus reaching the retina. Blurring of the pattern with excessive accommodation may also distort the waveform so that the first major positive wave peak is not accurately determined. In testing for spurious vision loss, small check sizes are preferable because the VEP to large checks can be normal even with significantly reduced visual acuity. A normal VEP to a 10- or 15-minute check is inconsistent with a Snellen acuity of less than 20/60. |
VISUAL DEVELOPMENT AND AGING |
The VEP is an objective method for assessing the maturation and functional
integrity of the visual system of infants. In the newborn, the diffuse
flash VEP is prolonged when compared with responses from older children
and adults, and the waveforms show considerable variability among
infants. The morphology of the VEP waveform becomes more uniform as
the latency shortens during the first 6 months of life.137–139 By 1 month of age in healthy infants, the pattern VEP has acquired a distinguishable
major positive peak. The response to large checks approaches
an adultlike configuration and latency by 6 months to 1 year of
age. However, pattern VEPs elicited by small checks still differ from
those obtained in adults, even in children 5 years of age.140 Using the pattern VEP, visual acuity can be estimated in children and infants of any age. Whereas VEP latency and visual acuity are inversely correlated during early development and until 5 years of age, the latency of the pattern VEP shortens in boys from 6 to 11 years old. However, this change is not as consistent in girls in the same age range.141 Generally, estimates of visual acuity are made by measuring VEP amplitudes for a series of different check or grating sizes. Several techniques have been used, such as determining the smallest stripe or check size that can elicit a reliable, reproducible transient VEP response or by extrapolating the functional relationships between VEP amplitude and the spatial frequency, check or grating size of the stimulus.142,143 However, visual acuity estimates are more rapidly and accurately obtained using steady-state responses and the sweep VEP technique.144 With this technique, steady-state responses are recorded as a grating stimulus is rapidly stepped or swept through a series of different spatial frequencies that increase up to and beyond the acuity limit. Using sweep VEPs, Norcia and Tyler145 showed that neonatal visual acuity is approximately 5 cycles per degree and that acuity increases to approximately 20 cycles per degree by 8 months of age. Therefore, the sweep technique estimates visual acuity of 8-month-old infants to be just slightly less than adult levels. This visual acuity estimate lags slightly behind the visual acuity estimated using transient VEP waveforms, which achieves adult levels by 6 to 8 months of age. The sweep technique also has been used to show that infant contrast sensitivity reaches adultlike levels by 6 to 8 months of age.146 The rapid development of visual acuity in infancy occurs concurrently with the development of binocularity and stereopsis. The VEP can be useful for measuring and investigating the development of normal binocularity and of the perturbations that can occur in the sensory interactions between the two eyes that are associated with strabismus, amblyopia, and ocular diseases that can impair visual development. In adults with normal binocular vision, the amplitude of the VEP obtained when the two eyes are stimulated simultaneously by identical patterns exceeds the amplitude of the monocular response generated by stimulating either eye, but it is less than the sum of the two individual monocular responses.147–149 In infancy, studies of the binocular pattern VEP have improved our understanding of the development of binocularity and stereopsis and of screening for visual defects such as amblyopia. Even at 2 months of age, the amplitude of a binocularly generated pattern VEP is larger than that of a monocularly generated pattern VEP.150 In all patients, the amount of binocular summation depends on the stimulus characteristics, the component of the response that is being monitored, and the position of the recording electrode. For example, the effect is seen best using stimuli reduced in contrast.151 Binocular cortical interactions can also be inferred by showing a decrease (inhibition) in the amplitude of the monocular pattern VEP after a period of monocular preadaptation of the other eye with the identical stimulus. This effect is indicative of an interaction in the visual cortex during the preadaptation that acts to inhibit the VEP.152 Binocular and monocular pattern VEP contrast sensitivity thresholds do not differ, but there are differences in the way the two types of pattern VEPs change with alternation of the temporal and spatial stimulus frequencies.153 Binocular facilitation is greatest with medium and low temporal frequencies and with a low spatial frequency grating.154 The binocular response is smaller than a monocular response if each eye is stimulated with a pattern reversing at different rates.155 Binocular interaction may be best studied using random red and green dot patterns in which the pattern of one colored stimulus appearance alternates with the other in a correlated phase. The red dots become green and the green become red. The two channels stimulate each eye individually; a red filter is placed before one eye and a green filter is placed before the other. Large-amplitude VEPs to random dot stimuli are found in adultsand in infants at 2 months of age, but not in neo-nates.156–158 Shifting fixation alters this VEP.159 Disturbances in binocular vision can be monitored with the pattern VEP. Results vary with stimulus parameters, such as the screen size, amount of contrast, and size of pattern.5 VEP amplitude is reduced and the waveform is distorted when large checks are used to stimulate eyes with anisometropic or strabismic amblyopia.160 However, other reports show that when amblyopic eyes are stimulated by large checks, the amplitude of the response is larger than for normal eyes.4 In contrast, pattern VEPs to checks smaller than 30 minutes are slightly prolonged in latency in these eyes.161,162 In many patients with amblyopia, occlusion therapy improves all pattern VEP parameters of amblyopic eyes, in parallel with improvement in subjective visual function. Conversely, prolonged occlusion in a child can disturb the VEP of the nonamblyopic eye.163 Patients with amblyopia also fail to show the normal binocular facilitation of the pattern VEP amplitude.164,165 The VEP can give information, unobtainable by other methods, about the visual system in patients with congenital disorders. In infants with congenital ocular motor apraxia who appear clinically blind because of failure to fixate or follow, normal pattern VEPs are valuable in showing an intact visual pathway.166 Many aspects of normal visual functioning change as a function of age. There is now a consensus that the VEP changes with age, although there is some debate concerning the specific changes that occur.38,167–172 There is general agreement that VEP latency increases with age.38,168,169 It was originally suggested that the latency of the pattern VEP was unaffected by age until approximately 60 years of age, at which point it increased gradually.168 Later reports indicated a slow but steady increase in latency beginning in adolescence and continuing throughout life.169 However, the rate of change depends on check size, reversal rate, and luminance. The age-related change in VEP amplitude is less well established. Although some evidence suggests an age-related decrease,170,171 other studies have found no significant decrease.169,172 What is clear is that the patient's age should be carefully considered in any clinical or experimental studies in which VEP latency or amplitude is measured. |
ALBINISM |
Albinism is a hereditary disorder of the melanin pigment system in which there is abnormal development of the retina and the retinocortical pathway. In all species of albino mammals, including humans, there is a visual pathway anomaly in which nerve fibers that originate in the temporal retina are misrouted at the optic chiasm. This misrouting results in an anomalous temporal nerve fiber decussation and an abnormal projection to the occipital cortex. It also disrupts the retinotopic organization of the cortical projections and produces a definite VEP asymmetry that is evident in all human albinos older than 5 years of age.173,174 This asymmetry is characteristic of human albinism and can be used to detect questionable cases. |
SUMMARY |
The VEP provides a measure of visually driven cortical activity that can be used to assess normal visual functioning as well as disorders of the visual system. Standards for VEP measurement and reporting have been developed2 that facilitate studies of the neurophysiologic basis of the response and its clinical application. When appropriately performed the VEP also offers an objective assessment of visual performance and permits quantification of deficits in acuity, contrast, and adaptation when one measures more than the latency. Consequently the technique can be very valuable for defining, detecting and monitoring abnormalities associated with specific diseases as well as normal and abnormal visual development. |