The visual evoked potential (VEP), or visual evoked response (VER), is a measurement of the electrical signal recorded at the scalp over the occipital cortex in response to light stimulus. The light-evoked signal, small in amplitude and hidden within the normal electroencephalographic (EEG) signal, is amplified by repetitive stimulation and time-locked, signal-averaging techniques, separating it from the background EEG readings. The precise origin of the VEP signal remains unclear, but it reveals the integrity of the afferent visual pathway; damage anywhere along the path may reduce the signal. The VEP is primarily a function of central visual function, because such a large region of occipital cortex is devoted to macular projections. Thus, peripheral visual loss might be overlooked by VEP testing.
Flash stimulus is useful for patients with very poor vision, in whom the response to pattern-reversal stimulus, which is more subtle, may be limited or absent. If measurable, however, the pattern response provides a more quantifiable and reliable waveform. The pattern may be studied by the number of cycles per second as well as the size of the checkerboard pattern. Smaller sizes allow detection of smaller changes in function. The most commonly studied VEP waveform typically contains an initial negative peak (N1), followed by a positive peak (P1, also known as P100 for its usual location at 100 msec); second negative (N2) and second positive (P2) peaks follow. The latency of onset of a peak after light stimulus and (to a lesser degree) the amplitude of the peak are the most useful features analyzed.
The examiner can compare readings from each eye with standardized normal values, readings from the 2 eyes, and readings from the 2 hemispheres. Peak latencies are relatively consistent, and accurate normative data are available; amplitude data are less consistent and thus less useful. Abnormalities in the waveform result from impairment anywhere along the visual pathways, but unilateral abnormalities may reflect optic neuropathy and thus may help to reveal lesions in the absence of clear-cut fundus abnormalities. Demyelination of the optic nerve results in increased latency of the P100 waveform, without significant effect on amplitude; ischemic, compressive, and toxic damage reduce amplitude primarily, with less effect on latency.
For most clinical situations, the VEP is of limited usefulness. It is subject to numerous factors that may produce abnormal waveforms in the absence of visual pathway damage, including uncorrected refractive error, media opacity, amblyopia, fatigue, and inattention (either intentional or unintentional). In most cases, the VEP is unnecessary for the diagnosis of optic neuropathy and is less accurate for its quantification than perimetry. The 2 scenarios in which VEPs remain clinically useful are (1) evaluation of the visual pathway in infants or inarticulate adults and (2) confirmation of intact visual pathways in patients suspected of nonorganic disease. A consistently abnormal flash response in the infant or inarticulate adult reflects gross impairment. An abnormal pattern response, however, is less useful, as it may indicate damage or may be a false-negative result from inattention or the reasons just cited. Normal responses confirm intact visual pathways.
A new technique being developed is the multifocal VEP (mfVEP); it is designed to detect small abnormalities in optic nerve transmission and provide topographic correlation along the visual pathway. Limited studies to date of the anterior visual pathways correlate visual field abnormalities to the abnormalities confirmed by mfVEP.
, BirchDG, HolderGE, BrigellMG. 2nd ed. Ophthalmology Monograph 2. San Francisco: American Academy of Ophthalmology; 2001.
, OdelJC, WinnBJ. The multifocal visual evoked potential.2003;23(4):279–289.