Secondary Logo

Journal Logo


The Multifocal Visual Evoked Potential

Hood, Donald C. Ph.D.; Odel, Jeffrey G. M.D.; Winn, Bryan J. B.A.

Author Information
Journal of Neuro-Ophthalmology: December 2003 - Volume 23 - Issue 4 - p 279-289
  • Free



The visual evoked potential (VEP) is not new to neuro-ophthalmology. Thirty years ago, Halliday et al. (1,2) reported delayed VEP responses in patients with optic neuritis/multiple sclerosis (ON/MS). Although some neuro-ophthalmologists still obtain a conventional VEP to help in the diagnosis of ON/MS or to rule out non-organic (psychogenic) visual loss, others find little or no use for this technique. The limited use of the VEP can be traced, at least in part, to the size of the stimuli used. The pattern reversal VEP is recorded to a display of at least 15° in diameter (3). Thus, local defects can easily be missed. The bright flash VEP uses full-field illumination and elicits a mass response from the anterior visual pathway. Thus, responses from abnormal regions of the field are summed with those from normal regions. Further, the pattern reversal VEP is dominated by responses from the lower field in most individuals (4–8). Thus, large defects in the upper field can be missed with the conventional VEP (8,9). In general, the lack of spatial information limits the usefulness of the technique.

The multifocal VEP (mfVEP), based on Sutter's multifocal electroretinogram (mfERG) technique (10), was developed by Baseler, Sutter et al. (11) to provide local VEP responses from the visual field. As described below, the technique combines conventional VEP recording techniques with a display that is subdivided into a number of regions, each of which has an independent stimulus controlled by specialized software. From a single, continuous EEG signal, a sophisticated mathematical algorithm extracts the VEP response generated by each of the independent regions. As typically used today, multiple individual VEP responses are generated simultaneously from 60 or so regions of the central 20 to 25° radius of the visual field.

Baseler et al. (11) suggested that the clinical utility of the mfVEP is limited because of the great variation of responses obtained from identical locations in normal individuals. Later, Graham et al. (12) and Hood et al. (13–15) improved the clinical utility of the mfVEP by introducing interocular comparisons to the analysis, as mfVEPs elicited from the right and left eyes of normal individuals are virtually identical. More recently, it has been demonstrated that a strictly monocular test, if properly analyzed, could have clinical utility as well (16–18). Thus the mfVEP is starting to find its role in the clinic, especially in the management of glaucoma (18).

The same equipment used to record mfERGs can record mfVEPs, and hundreds of centers have this equipment. However, only a handful of centers around the world are routinely recording mfVEPs. While the mfVEP can be recorded relatively easily with existing equipment, the software needed to identify local defects is not yet widely available. At the moment, we know of only two groups that have published these techniques, our group (13,17–19) and Graham, Klistorner et al. (12,16,20). However, the mfVEP technique is developing rapidly and advances in commercial hardware and software are expected in the near future.


The Equipment

The mfVEP can be recorded with the same equipment available for mfERG recordings. Although the equipment and VERIS software developed by Sutter (Electro-Diagnostic Imaging (EDI), San Mateo, CA) dominate the market, equipment from other companies (such as Roland Instruments, Germany) can be found outside the United States. Recently, an Australian company has developed a system strictly for recording mfVEPs. This AccuMap system (ObjectiVision Pty, Ltd., Sydney, Australia) is based on the work of Graham, Klistorner et al. (12,16,20). At this time, its availability in the United States is uncertain. Further, this is an evolving technology and developments in hardware and software are certain to appear soon. For example, EDI plans to enhance its analysis of the mfVEP.

The Display

Figure 1A shows the mfVEP display used in the work summarized here. Similar to the one originally described by Baseler et al. (11), it is a standard part of the VERIS software developed by Sutter (10). The Roland and AccuMap systems use a modification of this display. There are 60 sectors, each containing 16 checks—8 black and 8 white. The sectors and the checks are scaled, based on cortical magnification, to be of approximately equal effectiveness for cortical stimulation (11). For example, the central most sectors are about 1° wide, whereas the outermost sectors exceed 7°.

The multifocal visual evoked potential (mfVEP) technique. A. The display. There are 60 scaled sectors. B. The stimulation. A series of frame changes for a sector of the mfVEP. C. The response extraction. D. The signal and noise windows used for quantitative analysis. The sum of the responses to the sectors in the upper and lower fields for 14 normal individuals. C1 and C2 are the initial negative and positive components of the mfVEP. Dashed lines indicate the `signal window' (45 to 150 ms) from which the amplitude of the response is taken and the `noise window' (325 to 430 ms) used in the analysis. The summed responses from the upper and lower field are reversed in polarity.

Recording the Signal

In general, the VEP signal is recorded with the same electrodes and amplifiers used for conventional VEP recording. The critical differences are in the display, the method of stimulation, and the analysis of the raw records. For the records shown here, a single continuous VEP (EEG) record is obtained with an active electrode placed 4 cm above the inion, a reference electrode placed at the inion, and a ground electrode placed on the forehead. In addition, we record additional channels of VEP activity by placing two active electrodes 1 cm above and 4 cm lateral to the inion (17–19). This method, suggested by Klistorner and Graham (20), yields better responses in some parts of the field, especially along the lower midline (16–18,20). By analyzing the records offline with programs written in MATLAB, the information from the different electrodes can be combined (17–19). In the case of the AccuMap system, the software is built into the system. Technical details can be found elsewhere (16–21).

From a single, continuous VEP (EEG) signal, the software extracts 60 mfVEP responses, each associated with one of the sectors of the display. This is the “magic” of the multifocal technique: 60 responses are obtained from one record. To get a sense of how this magic is produced, the nature of the local stimulation of each sector must be examined.

Extracting the Local Responses from the Signal

To understand the mfVEP technique, it is essential to understand how each of the sectors is varied during the test. Each sector is an independent stimulus. Every 13.3 msec, the frame of the monitor changes and each sector has a 50% chance of reversing contrast or staying the same. Figure 1B shows a series of frame changes in which either the contrast is reversed or no change takes place. Each of the 60 sectors of the display in Figure 1A goes through its own pseudo-random sequence. In fact, the 60 pseudo-random sequences are the same series of `reversals' or `no change', but each of the sectors starts its sequence at a different point in the series. The reason for this, and the nature of the pseudo-random series, are technical details that the reader does not need to understand in depth. It is sufficient to know that these pseudo-random sequences allow the software to rapidly extract the response associated with each of the 60 sectors (10). [Readers interested in learning more about the technical details should consult references 10 and 21.]

How does the software extract 60 responses from a single record? Technically, each response is the result of a serial correlation between the stimulation sequence of a particular sector and the single continuous response. Figure 1C provides a non-technical explanation. If one summed the first 200 msec of all of the records following the point in time at which a particular sector reversed in contrast, the result would look like the response R in Figure 1C. Likewise, if one summed the first 200 msec of all of the records following the point in time at which the same sector did not reverse in contrast, the result would look like NR in Figure 1C. Response R should contain the responses to all the sectors that reversed including the sector in question. NR, on the other hand, will have the responses to all sectors except the sector in question. The difference between R and NR is the response to the sector in question. Whereas the software could calculate the 60 mfVEP responses this way, it does not. The pseudo-random sequence, described above, is chosen in a certain way (an m-sequence) such that, when coupled with a special algorithm, the software can make these calculations very quickly (10). [Because of the patent held by EDI on the m-sequence technique, other manufactures (including Roland and ObjectiVision) of multifocal equipment use different methods for extracting the multifocal responses.]

Displaying the Responses

Figure 2A shows the 60 mfVEP responses for monocular stimulation of the right (blue) and left (red) eyes from a normal subject. Note that these responses are positioned so that they do not overlap. Thus the scaling is arbitrary, as the circles in color indicate. For example, there are 12 responses in the central 2.6° (5.2° in diameter), while each of the responses in the outer ring is produced by regions that are larger than the entire central 2.6°.

Responses in health and disease. A. Normal control. The 60 mfVEP responses for the monocular stimulation of the right (blue) and left (red) eyes of a normal subject. The red, blue and green circles indicate radii of 2.6°, 9.8° and 22.2°, respectively. B. Patient 1: Leber hereditary optic neuropathy. Asterisks indicate four locations where the response OS is much smaller than OD. Plus symbols show two examples of nearly identical responses OU. C. Patient 3: Optic neuritis. D. Patient 4: Ischemic optic neuropathy. The calibration bar (panels A–D) indicates 200nV and 100ms.


Although mfVEPS have been recorded from normal controls (12,13,16,17), these results are not reported in a way that would be of use to other investigators. Further, the currently available versions of the VERIS and Roland software do not have normative values for the mfVEP; the AccuMap system will have them. Even so, for all electrophysiological tests, it is important that each clinic establish its own age-related normative values. This is particularly true for the mfVEP, where the ability to detect subtle defects will depend upon the level of noise, which can vary from one setting to another.

The ability to track the progression of a disease will depend on various factors, but it is clear that good repeat reliability is essential. Relatively little has been published on repeat reliability, although the existing evidence suggests it is very good (11,12,22). In a recent study of 15 control subjects and 10 patients with glaucoma, the repeat reliability of the mfVEP was better than that of the HVF (22).


The mfVEP response bears a superficial resemblance to the conventional pattern-reversal VEP. Figure 1D contains the sum of all the responses to the sectors in the upper and lower fields of 14 normal individuals. [Note that these responses are shown reversed in polarity as compared with the way they are displayed in the VERIS software because the software reverses the polarity (7,18,21).] In the responses from the lower field, there is an initial negative component (C1) around 65 ms followed by a prominent positive component (C2) around 95 msec, analogous to the N75 and P100 of the conventional pattern reversal VEP (3). In a study designed to compare the conventional VEP and mfVEP under similar conditions, Fortune and Hood (7) concluded that the local mfVEP response is not simply a “little conventional VEP.” The C2 component of the mfVEP is smaller and slightly faster than the P100 of the conventional VEP. Further, whereas the mfVEP from the upper visual field is reversed in polarity as compared with that of the lower field, the conventional VEP generally has the same polarity for upper and lower field stimulation. Fortune and Hood (7) concluded that these differences are due to the fast mfVEP sequence, which they speculate is producing a response with a smaller extrastriate contribution than that of the conventional VEP.


To introduce our method for displaying the results of the mfVEP, consider the following case. Patient 1, a 43-year-old man, presented with a 4-week history of “blurry vision” in the superior temporal field of his OS. He was under stress at work and in the process of a divorce. Humphrey visual fields (HVF) OS showed abnormalities (Fig. 3A), but there were abnormal points in his HVF for the OD as well. Further, his HVF results were suspect because the number of fixation losses was high (13/15 OS and 7/14 OD) and the patient had alcohol on his breath while performing the test. Thus, he was referred for a mfVEP. His records are presented in Figure 2B.

Patient 1: Leber hereditary optic neuropathy. A. Monocular (left and middle panels) and interocular (right panel) HVF probability plots. B. Monocular (left and middle panels) and interocular (right panel) mfVEP probability plots. The colored squares indicate that there is a significant difference at greater than the 5% (desaturated) or 1% (saturated) level. The color denotes whether the right (blue) or left (red) eye had the smaller response. Sample responses are shown for regions that exceed 1% (asterisks) and regions that are normal (plus symbols).

In a number of locations, the mfVEPs from his OS (red traces in Fig. 2B) are clearly smaller than those from his OD (blue traces). The asterisks in Figure 2B indicate four examples. On the other hand, the plus symbols show two locations where the responses from the two eyes are nearly identical. To provide a quantitative measure of these differences, probability plots are derived for the mfVEP analogous to the probability plots for the HVF (Fig. 3A).

The first two panels in Figure 3B are probability plots for the mfVEPs of the left and right eyes compared with the mfVEPs from the left and right eyes of a group of control subjects. For each sector in the field, the amplitude of the mfVEP is determined for the response in a time window from 45 to 150 msec (see “signal window” in Fig. 1D). The amplitude of the response is compared with the mean and standard deviation of the mfVEP amplitudes of a group of control subjects. [Technically, the root-mean-square is taken as the measure of amplitude and the analyses are based on a comparison, the signal-to-noise ratio, of the response in the signal window to the response in the “noise window” (Fig. 1D) (17–19,23).] Each of the squares in the mfVEP probability plots in Figure 3B is located at the center of one of the sectors of the mfVEP display (Fig. 1A). A colored square indicates that the mfVEP was statistically significant at either the 5% (desaturated color) or 1% (saturated color) level, while the color indicates whether it was the left (red) or right (blue) eye that was significantly smaller than normal.

In many patients, an interocular comparison of the mfVEPs is a more sensitive indicator of damage (18,24). In fact, in our experience, the interocular comparisons are more valuable to the neuro-ophthalmologist than are the monocular results. The interocular mfVEP probability plot for this patient is shown as the rightmost panel of Figure 3B. To obtain this plot, the ratio of the amplitudes of the mfVEP of the two eyes is measured for each sector of the display (13–15,18,24). This ratio is then compared with the ratios from a group of controls to establish 5% and 1% significance levels. The result is coded as in the case of the monocular fields. [See Hood and Greenstein (18) for a review of the derivation and use of both monocular and interocular probability plots, and Graham, Klistorner et al. (12,16,20) for a similar approach.]


The results from the mfVEP can be directly compared with those from the HVF. To make this comparison easy, the HVF and mfVEP probability plots are presented on the same scale in Figures 3 to 5. For example, the circles in the leftmost panels of Figures 3A and 3B have radii of 2.6° (red), 9.8° (blue) and 22.25°(green). To obtain an interocular HVF (15,18) (rightmost panel in Fig. 3A), the total deviation scores for the right eye are subtracted from that of the left and coded as in the case of the mfVEP. The significance levels are based on the sensitivities of a group of 100 normal individuals (25). Now the three HVF plots (Fig. 3A) can be compared with the three mfVEP plots, two monocular and one interocular (Fig. 3B).

A. Patient 2: Optic neuritis. The OD HVF (left panel) and interocular mfVEP (right panel) probability plots. Responses in the inset demonstrate delayed signal OD (blue). B. Patient 3: Optic neuritis. The OS HVF (left panel) and interocular mfVEP (right panel) probability plots. In the inset are examples of responses with severely diminished amplitudes (solid dot), responses with somewhat diminished amplitudes and large delays (asterisks), and responses that appear relatively normal (plus symbols). Optic neuritis can have a range of effects on the mfVEP. C. Patient 4: Ischemic optic neuropathy (ION). The OD HVF (left panel) and interocular mfVEP (right panel) probability plots. In the inset are examples of responses with severely diminished amplitudes (solid dot), responses with somewhat diminished amplitudes but normal latencies (asterisks), and responses that appear relatively normal (plus symbols). ION reduces the amplitude but does not affect the latency of the mfVEP. D. Patient 5: Glaucoma suspect. The OS HVF (left panel) and interocular mfVEP (right panel) probability plots obtained on two occasions approximately 4 months apart. On the mfVEP plots, the green box and ellipse show regions with significantly abnormal mfVEPs. The purple ellipse indicates a region with normal mfVEPs. In the inset are responses for these regions. The mfVEP shows good repeatability and detects defects missed on the HVF plots.
Patient 6: glaucoma. An illustration of the spatial resolution of the mfVEP. A. Monocular 24–2 (left and middle panels) and interocular (right panel) HVF probability plots. B. Monocular 10–2 (left and middle panels) and interocular (right panel) HVF probability plots obtained on the same day. C. Monocular (left and middle panels) and interocular (right panel) mfVEP probability plots. The red ellipse indicates an area with decreased mfVEP amplitudes OS (left inset), while blue ellipse indicates an area with decreased mfVEP amplitudes OD (right inset). The purple ellipse indicates an area with normal mfVEPs (middle inset). For the central 10°, the mfVEP has about the same spatial resolution as the 10–2.

In the case of this patient, the mfVEP is confirming a defect in the OS on both the monocular and interocular plots. However, the HVF and mfVEP are not in complete agreement. For example, the HVFs from both eyes show defects in the lower field and these are not seen in the mfVEP plots. There are many reasons for a disagreement between the mfVEP and HVF results (15). In this case, the disagreement has a simple explanation. The patient was not a good field taker and the HVFs were not reliable.


Since October 1998, we have performed mfVEPs on over 200 patients evaluated by two neuro-ophthalmologists, Drs. Myles Behrens and Jeffrey Odel. We summarize our experience by grouping patients into the most common reasons for seeking a mfVEP.

Ruling Out Non-Organic Visual Loss

Like the conventional VEP, the mfVEP can be used to rule out non-organic visual loss. The mfVEP has the advantage in that it provides a topographical representation, which can be compared with the patient's visual fields. The mfVEP in Patient 1, suspected of having non-organic visual loss, clearly indicated an organic problem. The responses from the OS are significantly depressed in a number of locations shown as the red squares in Figure 3B. The responses from four of these locations are labeled with the asterisks in Figure 2B and shown in the inset in Figure 3B (rightmost column). For comparison, the inset in Figure 3B also shows two responses within the normal range (plus symbol in Fig. 2B). As the abnormal region is relatively small, it probably would have been missed by the conventional VEP. In any case, the presence of a clearly abnormal mfVEP in a region that corresponded to the patient's subjective complaint clearly indicated that this was an organic problem. Blood tests revealed Leber hereditary optic neuropathy. He subsequently lost vision in his other eye and developed bilateral optic atrophy.

More typically, patients who are tested to rule out a non-organic cause have normal mfVEPs. We have previously published an example of a patient with a dense inferior bitemporal quadrantanopsia in whom the normal mfVEP confirmed a non-organic diagnosis (26). In addition, the mfVEP can detect a non-organic overlay that can be missed on routine examination. In some of the patients tested, a non-organic cause was not part of the differential diagnosis, as an organic cause for a visual defect was clearly established. However, a mfVEP obtained to confirm the extent of the defect indicated that the HVF results exaggerated the extent and/or depth of the defect (18).

Diagnosing and Following Optic Neuritis/Multiple Sclerosis

When it comes to the diagnosis and treatment of optic neuritis (ON), the mfVEP has clear advantages over traditional visual fields and the conventional VEP. It is well known that patients who have recovered from an attack of ON can have reasonably normal visual fields. In some cases, the patients complain of “hazy” or “fuzzy” vision in parts of the visual field, although the HVF can appear normal in these regions. Patient 2, a 39-year-old man, had an acute attack of ON in October of 1999. The diagnosis of MS was confirmed with neurologic, spinal fluid, and MRI examinations. The HVF and mfVEP performed 7 days later confirmed a moderate to marked inferonasal defect OD (14). Within 4 weeks, his vision improved and his HVF approached normal. The patient, however, still had some complaints and we have been following him with the mfVEP. Both mfVEPs and HVFs (Fig. 4A) were obtained 10 months after the onset of ON. His HVF was normal (Fig. 4A, left panel). However, the mfVEP OD was markedly delayed in some regions. Examples are shown by the inset in Figure 4A, right panel. The responses (blue) from the OD are clearly slower than the responses (red) from the OS. Notice that the interocular mfVEP plot (Fig. 4A, right panel) is showing abnormalities while the monocular HVF is essentially normal (Fig. 4A, left panel). [The monocular mfVEP plots were normal.] The abnormalities seen on the mfVEP plot are due, in part, to the small differences in amplitude and, in part, to the difference in implicit time. In any case, the mfVEP is more effective than the HVF for monitoring the state of patients who have recovered from ON.

A given patient can show a range of normal and abnormal mfVEP responses within the recovered HVF field. It is not uncommon to see regions with normal amplitude and timing, regions with normal amplitude but abnormal timing and regions with depressed amplitude and abnormal timing in the same patient (14,27). These findings suggest that the mfVEP is detecting local demyelinization, as the regions with delays can be adjacent to regions with responses with normal amplitude and latency. These findings also explain the range of results reported in the conventional VEP literature. For patients who have recovered from ON, the conventional VEP is often abnormal, but it can appear normal in cases where MS has been documented on MRI and/or neurologic examinations (1,2,28,29). Because the conventional VEP is the unequal sum of the responses from different parts of the field, the result will depend on how the normal and abnormal field regions are weighted. Clearly, the mfVEP has the advantage over the conventional VEP and the HVF in following patients with ON. The mfVEP should prove useful in assessing the efficacy of drugs designed to impede the course of MS.

The mfVEP also can help in the diagnosis of ON/MS. Although the diagnosis of ON can usually be made based upon the patient's history and HVF, a small percentage of patients with ON can present with swollen discs but without pain. In these cases, it is important to distinguish between ON, ischemic optic neuropathy (ION), and a compressive lesion. Patient 3 is a 44-year-old women who noticed blurred vision in the OS 6 days prior to her first test. The HVF showed a central scotoma breaking out above fixation. Two weeks later, both HVFs and mfVEPs were obtained. The HVF showed a paracentral defect OS (Fig. 4B, left panel). There is a much larger region of abnormal responses on the interocular mfVEP plot (Fig. 4B, right panel). In addition, the clear delays in the mfVEP confirm that the problem is ON/MS. Her mfVEPs are shown in Figure 2C, as well as in the inset in Figure 4B. Note that some regions show severely diminished amplitudes (solid dot), some show somewhat diminished amplitudes with large delays (asterisks), and other regions have normal responses (plus symbols).

This case can be contrasted with Patient 4, a 52-year-old man who noted a sudden onset, painless loss of vision in the OD. His HVF revealed an inferior altitudinal defect and he was found to have disc swelling OD and a small cup OS. Visual acuity and color vision were preserved in the affected eye. Both his HVF (Fig. 4C, left panel) and the interocular mfVEP (Fig. 4C, right panel) probability plots show an extensive defect in the inferior field. His mfVEPs are shown in Figure 2D, as well as in the inset in Figure 4C. Unlike the patients with ON, the mfVEPs from patients with ION, when present, are not delayed. As an example, examine the responses marked with asterisks in Figure 4C (right panel). Eight months later, his acuity and visual fields remained stable. Although disc swelling had resolved, he developed segmental optic atrophy. His clinical course and mfVEP are consistent with ION.

Confirming Unreliable or Questionable Fields

Some patients who have difficulty with the HVF are easy to spot as the reliability indices (percent of fixation losses, false positives, and false negatives) are abnormal. Although occasionally poor visual field takers are also poor mfVEP subjects (18), the mfVEP provides a viable alternative for most of these patients. Patient 1, discussed above (Fig. 3), provides an example. He had 87% (OS) and 50% (OD) fixation losses, far exceeding the criterion for an unreliable field taker.

The neuro-ophthalmologist is often faced with an apparent defect on the HVF, but with insufficient evidence to feel comfortable making a diagnosis. In such cases, the mfVEP provides another topographical map that can be used to confirm or refute the defect detected on the HVF. In our experience, this is one of the most important uses of the mfVEP in neuro-ophthalmology.

Patient 5 is a 70-year-old man who was a glaucoma suspect. His HVF OD appeared normal, but seven HVFs OS obtained over a 19-month period occasionally showed abnormalities. The HVF obtained in October 2001 (Fig. 4D, left panel) showed defects that were more subtle than those seen on some of the earlier tests. The field obtained a few months later (Fig. 4E, left panel) suggested a defect in the lower field (green ellipse). The mfVEPs (right panels of Fig. 4D and E) were obtained on the same days on which the fields were obtained. Notice that defects are seen on the mfVEP plots in both the upper and lower fields on both test days. Some of the responses from these regions (green ellipses) are shown in the insets of Figure 4D and E along with responses from regions with normal mfVEPs (purple ellipses). The mfVEP confirms glaucomatous damage.

Patient 6 is a 62-year-old man with peculiar bitemporal scotomas and with a strong family history of glaucoma. Since March of 2001 he was aware of a problem just off fixation, and just below the midline, in the left visual field of his OS and the right visual field of his OD. He first noticed that he was unable to see the “O” in a vertical HOTEL sign in contralateral and corresponding locations of both eyes. His 24-2 HVFs (Fig. 5A) show a small defect OD and the 10-2 HVFs (Fig. 5B) show small defects OU. Whereas glaucomatous damage was suspected, a chiasmal lesion or demyelinating event could not be ruled out. Bilateral defects appear on the mfVEP (Fig. 5C), closely matching the topography of the defects on the HVF. The responses show no sign of delays as might be expected with MS.

Following Disease Progression

As mentioned above, the mfVEP shows good repeat reliability (18,20,22,30). The records in Figure 4D and E provide an illustration. These mfVEP responses were obtained 3 months apart. This reliability allows the mfVEP to be used as a means to follow disease course, a particularly important resource when the HVFs cannot be trusted.

Combining mfVEP with mfERG

Multifocal VEP may be combined with multifocal electroretinography (mfERG) in diagnosis (31). In the work-up of obscure visual loss, patients will have a mfVEP first and then be dilated to have a mfERG on the same visit. If the mfVEP and the HVF are both abnormal, a mfERG is used to exclude outer retinal damage. This combination of tests has frequently been helpful in distinguishing between branch arterial occlusions and ION, both of which can cause altitudinal visual field defects. In effect, the mfVEP establishes the organic origin of the HVF defect, confirms the extent of the field loss, and may suggest its cause (with delayed responses in the case of ON), while the mfERG rules out damage to the outer retina.


Like the HVF, the mfVEP is not without its limitations. Some are similar to those of the HVF. For example, as in the case of the HVF, lids can occlude the field of view, care needs to be taken to correct for refractive error, and eye movements should be monitored. Three problems with the mfVEP are of special concern.

Spatial Resolution

As the case in Figure 5 illustrates, the spatial resolution of the mfVEP can rival the 10-2 HVF. In the periphery, however, the spatial resolution can be relatively poor. The sectors in the peripheral ring are over 7° in width. Considering the fact that one would like to see at least two, if not three, contiguous abnormal points to confirm a defect (16,18,19), it is possible to overlook reasonably large, focal defects if they are restricted to the outer ring (beyond about 15°). For example, it is difficult to detect the blind spot in most recordings (18).

Poor mfVEP Subjects

Just as there are patients who are unreliable HVF takers, there are patients who can not be tested with the mfVEP. Patients who refuse to cooperate or who go to sleep may be difficult to test on either the HVF or the mfVEP. In our experience, however, most patients who are poor HVF takers are able to perform the mfVEP test (18). On the other hand, there are some good HVF takers who do not produce usable mfVEP recordings. Relatively few patients have mfVEPs too poor to be of clinical use. In these cases, the responses are difficult to discern because of a high noise level secondary to a large alpha contribution or muscle tension.

Eccentric Fixation

As with HVF testing, it is important to monitor the patient's eye position. An eye camera or direct visualization of the eye should be used to assure that fixation is steady. Unsteady fixation can cause diminished responses in the center of the field (18,32,33). Monitoring the eye, however, will not assure that the fixation is accurate. Some patients with central visual problems can have eccentric fixation. Figure 6 shows the effects of a 3° fixation error. A control subject was instructed to maintain steady fixation down and to the left by 3° for the OD while the OS was tested with central fixation. Compared with the control condition (Fig. 6A,B), the eccentric fixation condition (Fig. 6C,D) showed apparent defects in both eyes on the interocular probability plot. It is relatively easy to attribute these “defects” to eccentric fixation. The probability plot shows smaller responses in diagonally opposite parts of the field. That these symmetrical defects are due to eccentric fixation is confirmed by examining the responses from near the midline. Some of these responses (see inset in Fig. 6D) show a polarity reversal between the two eyes. In sum, it is important to monitor eye position to avoid false positives due to unsteady fixation and to scrutinize the mfVEP plot and responses to avoid false positives due to eccentric fixation.

The problem of eccentric fixation. A. The interocular mfVEP probability plot for a control subject fixating at the center of the stimulus when testing OU. B. The 60 mfVEP responses corresponding to the probability plot in A. Responses in the inset are of the same polarity and appear normal. C. The interocular mfVEP probability plot for the same subject instructed to fixate down and to the left by 3° when testing OD and fixating in the center when testing OS. D. The 60 mfVEP responses corresponding to the probability plot in C. Responses in the inset show clear polarity reversals and amplitude differences between the two eyes. Eccentric fixation can give the appearance of an abnormality in an otherwise normal eye.<


The clinical value of the mfVEP lies in its ability to detect small abnormalities in visual signal transmission from centric and eccentric field and to provide a topographical display of these deficits. Together with the mfERG, it can provide objective evidence of visual pathway pathology. As with the mfERG, the value of the mfVEP is optimized if it is compared with HVFs obtained concurrently.

The multifocal technology is still in its infancy. Better methods of signal analysis, more effective stimuli, and improved recording conditions will soon be developed (18). It is not yet a widely accessible device. Although recording high quality mfVEPs is not difficult, it s requires a technician trained in electrophysiological recordings. Commercial software for adequately analyzing the mfVEP is not yet available in most countries, including the United States. Until the analysis is standardized, the interpretation of the mfVEP requires experience and is best performed by an experienced electrophysiologist familiar with the mfVEP test in concert with a neuro-ophthalmologist.


The authors gratefully acknowledge the support of Dr. Myles Behrens. We also would like to thank Candice Chen, Jennifer Hong and Annemarie Gallagher for their help throughout this work. A special thanks is extended to Xian Zhang for his contributions to the analysis tools essential to this study.


1. Halliday AM, McDonald WI, Mushin, J. Delayed visual evoked response in optic neuritis. Lancet. 1972; 1:982–985.
2. Halliday AM, McDonald WI, Mushin, J. Visual evoked response in diagnosis of multiple sclerosis. Br Med J. 1973; 1:661–664.
3. Harding GF, Odom JV, Spileers W, et al. Standard for visual evoked potentials 1995. The international society for clinical electrophysiology of vision. Vision Res. 1996; 36:3567–3572.
4. Halliday AM, Michael WF. Changes in pattern-evoked responses in man associated with the vertical and horizontal meridians of the visual field. J Physiol. 1970; 208:499–513.
5. Michael WF, Halliday AM. Differences between the occipital distribution of upper and lower field pattern-evoked responses in man. Brain Res. 1971; 32:311–324.
6. Ossenblok P, Spekreijse H. The extrastriate generators of the EP to checkerboard onset. A source localization approach. Electroencephalogr Clin Neurophysiol. 1991; 80:181–193.
7. Fortune B, Hood DC. Conventional pattern-reversal VEPs are not equivalent to summed multifocal VEPs. Invest Ophthalmol Vis Sci. 2003; 44:1364–1375.
8. Fortune B, Hood DC, Johnson CA. Comparison of conventional and multifocal VEPs. Invest Ophthalmol Vis Sci. 2002; 43:Abstract nr 2126.
9. Howe JW, Mitchell KW. Visual evoked cortical potential to paracentral retinal stimulation in chronic glaucoma, ocular hypertension, and an age-matched group of normals. Doc Ophthalmol. 1986; 63:37–44.
10. Sutter EE. The fast m-transform: a fast computation of cross-correlations with binary m-sequences. Soc Ind Appl Math. 1991; 20:686–694.
11. Baseler HA, Sutter EE, Klein SA, et al. The topography of visual evoked response properties across the visual field. Electroencephalogr Clin Neurophysiol. 1994; 90:65–81.
12. Graham SL, Klistorner AI, Grigg JR, et al. Objective VEP perimetry in glaucoma: asymmetry analysis to identify early deficits. J Glaucoma. 2000; 9:10–19.
13. Hood DC, Zhang X, Greenstein VC, et al. An interocular comparison of the multifocal VEP: a possible technique for detecting local damage to the optic nerve. Invest Ophthalmol Vis Sci. 2000; 41:1580–1587.
14. Hood DC, Odel JG, Zhang X. Tracking the recovery of local optic nerve function after optic neuritis: a multifocal VEP study. Invest Ophthalmol Vis Sci. 2000; 41:4032–4038.
15. Hood DC, Zhang X. Multifocal ERG and VEP responses and visual fields: comparing disease-related changes. Doc Ophthalmol. 2000; 100:115–137.
16. Goldberg I, Graham SL, Klistorner AI. Multifocal objective perimetry in the detection of glaucomatous field loss. Am J Ophthalmol. 2002; 133:29–39.
17. Hood DC, Zhang X, Winn BJ. Detecting glaucomatous damage with the mfVEP: how can a monocular test work? J Glaucoma. 2003; 12:3–15.
18. Hood DC, Greenstein, VC. The multifocal VEP and ganglion cell damage: applications and limitations for the study of glaucoma. Prog Ret Eye Res. 2002; 22:201–251.
19. Hood DC, Zhang X, Hong JE, et al. Quantifying the benefits of additional channels of multifocal VEP recording. Doc Ophthalmol. 2002; 104:303–320.
20. Klistorner AI, Graham SL. Objective perimetry in glaucoma. Ophthalmology. 2000; 107:2283–2299.
21. Sutter EE. Imaging visual function with the multifocal m-sequence technique. Vision Res. 2001; 41:1241–1255.
22. Chen CS, Hood DC, Zhang X, et al. Repeat reliability of the multifocal visual evoked potential in normal and glaucomatous eyes. J Glaucoma. 2003; 12:399–408.
23. Zhang X, Hood DC, Chen CS, et al. A signal-to-noise analysis of multifocal VEP responses: an objective definition for poor records. Doc Ophthalmol. 2002; 104:287–302.
24. Hood DC, Greenstein VC, Odel JG, et al. Visual field defects and multifocal visual evoked potentials: evidence of a linear relationship. Arch Ophthal. 2002; 120:1672–1681.
25. Johnson CA, Spry PGD. Normal age-related sensitivity loss for perimetry tests that evaluate a variety of different visual functions [abstract]. Invest Ophthalmol Vis Sci. 1999; 40:S67.
26. Miele DL, Odel JG, Behrens M, et al. Functional bitemporal quadrantopia and the multifocal visual evoked potential. J Neuroophthalmol. 2000; 20:159–162.
27. Kardon RH, Givre SJ, Wall M, Hood D. Comparison of threshold and multifocal-VEP perimetry in recovered optic neuritis. In: Wall M, RP Mills, eds. Perimetry Update 2000: Proceedings of the XVIIth International Perimetric Society Meeting Sept. 6–9, 2000. New York: Kugler, 2001:19–28.
28. Holder GE. Multiple sclerosis. In: Heckenlively JR, Arden GB, eds. Principles and Practices of Clinical Electrophysiology of Vision. St. Louis: Mosby, 1991:797–805.
29. Frederiksen JL, Petrera J. Serial visual evoked potentials in 90 untreated patients with acute optic neuritis. Surv Ophthalmol. 1999; 44:S54–62.
30. Klistorner, AL, Graham SL. Multifocal pattern VEP perimetry: analysis of sectoral waveforms. Doc Ophthalmol. 1999; 98:183–96.
31. Hood DC. Assessing retinal function with the multifocal technique. Prog Ret Eye Res. 2000; 19:607–646.
32. Menz MD, Poloschek CM, Menz MK, et al. The effect of fixation error on the multifocal VEP. Annual Meeting Abstract and Program Plannner [on CD-ROM]. Association for Research in Vision and Ophthalmology. 2002; abstract nr. 4740.
33. Winn BJ, Shin EH, Greenstein VC, et al. Interpreting the multifocal visual evoked potential (mfVEP): The influence of cataracts, refraction and fixation. Annual Meeting Abstract and Program Plannner [on CD-ROM]. Association for Research in Vision and Ophthalmology. 2003; abstract nr. 32.
© 2003 Lippincott Williams & Wilkins, Inc.