advertisement

---

Editors Selection IGR 8-2

Examination methods: mfVEP

Comment by Stuart Graham on:

13825 Repeatability of normal multifocal VEP: implications for detecting progression, Fortune B; Demirel S; Zhang X et al., Journal of Glaucoma, 2006; 15: 131-141

Find related abstracts

---

In this paper by Fortune et al. (418) the repeatability of the multifocal multichannel visual evoked potential (mfVEP) is examined, specifically its variability after 12-month period. This is compared with performance on standard Humphrey visual fields. Many previous publications have established that there is a loss of mfVEP amplitude in glaucoma, and that the location of that signal reduction does correspond with subjective field loss. However if the technique is to have application as a form of objective perimetry, it needs to be provide comparable data between tests such that progression analysis can be determined, otherwise its use would be limited to screening or as a confirmatory test when a subjective defect is detected (or not detected). The authors have analysed the limits of agreement (LOA) for the 60 individual points in the field (using the VERIS system) to determine the amount of change that would be required to show a statistically significant progression, and present that as a decibel (dB) so that it can be compared with Humphrey SAP data. On average around a 30% reduction in signal is required to suggest progression (ie equivalent to a change of greater than two standard deviations) with values from 2 to 4.3dB. The variability is greater in the upper field and the signal to noise ratio (SNR) smaller, so this limits the opportunity for detecting change in this region. The LOA values are greater for SAP (2.9 to 8.9dB), but this is offset by the greater dynamic range of subjective perimetry. The results show excellent specificity for the mfVEP when a cluster analysis of 3 points p < 0.05 is used to define a scotoma.

Repeatability of mfVEP was better than of standard automated perimetry

There have been concerns with the reproducibility of signal amplitudes, and this paper addresses some of these concerns. However the main limitation of the mfVEP still remains intra-individual reproducibility and noisy recordings which can lead to false positives. There is still a level of patient co-operation required, together with technician experience to recognize noise such as alpha rhythm (patient losing concentration), muscle noise and other artifacts. New stimulus designs to enhance signal responses, improved amplification and better filtering techniques will eventually provide improved SNRs. This will enhance the reproducibility of the technique, and allow the future development of statistical progression analysis.

---

• ← Previous
• Next →

---