Multifocal VEP and ganglion cell damage: applications and limitations for the study of glaucoma

With the multifocal technique, visual evoked potentials (VEPs) can be recorded simultaneously from many regions of the visual field in a matter of minutes. Recently, the multifocal visual evoked potential technique (mfVEP) has generated considerable interest, especially among those seeking objective measures of glaucomatous damage. It is well accepted that significant ganglion cell damage can occur before functional deficits are detected with static automated achromatic perimetry, the "gold standard" for detecting and monitoring glaucomatous damage. In this article, we ask the following questions: What are the potential applications of the mfVEP technique? What are its limitations? To what extent will it replace or augment static automated achromatic perimetry? To answer these questions requires an understanding of the mfVEP technique, as well as techniques needed to relate its results to those of automated perimetry. describes how the mfVEP is elicited, recorded, derived and displayed. If both eyes of an individual are normal, then mfVEPs recorded for monocular stimulation of each eye are essentially identical. However, the amplitude and waveform of the mfVEP responses vary across individuals, as well as across the visual field within an individual. These variations in the normal mfVEPs are described in Section 3. In, these variations are related to cortical anatomy, and to the cortical sources contributing to the mfVEP. The mfVEP is predominantly generated in V1. Although there are undoubtedly extrastriate contributions, these contributions are probably smaller for the mfVEP than for the conventional VEP. The mfVEP is not a small version of the conventional VEP. To detect ganglion cell damage with the mfVEP requires methods for analyzing the responses and for displaying the results. In, a method for detecting ganglion cell damage is described. This method compares the monocular responses from the two eyes of an individual and produces a map of the defects. This map is in the form of a probability plot similar to the one used to display visual field defects measured with automated perimetry. Procedures are described for directly comparing these mfVEP probability plots to the probability plots for Humphrey visual fields (HVFs). The interocular mfVEP test described in will not be sensitive to bilateral damage. describes a test based upon monocular mfVEPs. The statistical basis of the monocular mfVEP test is relatively complex (see ). In any case, under many conditions the interocular test will be more sensitive and this is discussed in. summarizes a number of clinical applications of the mfVEP and concludes that the mfVEP has a place in the clinical management of glaucoma. To understand the limitations of the mfVEP, a signal-to-noise ratio (SNR) approach is described in. Using the techniques described in, the relationship between the amplitude of the mfVEP and the sensitivity loss of the HVF is discussed in. The evidence supports a simple model in which the amplitude of the signal portion, but not the noise portion, of the mfVEP response is proportional to HVF loss where HVF loss is expressed in linear, not dB, units. It is hypothesized that both the signal in the mfVEP, and the sensitivity of the HVF, are linearly related to ganglion cell loss. A theoretical approach, developed in, allows a direct comparison of the efficacy of the mfVEP and HVF in detecting glaucomatous damage. In short, when the mfVEP has a large SNR it will often be superior to the HVF in detecting damage. On the other hand, when the mfVEP has a small SNR, the HVF will probably be superior. summarizes the relative advantages of the HVF and the mfVEP. In summary, the mfVEP does have a place in the clinical management of glaucoma, although it is not likely to replace static automated achromatic perimetry in the near future. However, this is an evolving technology and the future will undoubtedly see major improvements in the mfVEP technique.