Purpose: To understand and improve the detection of glaucomatous damage with multifocal visual evoked potentials (mfVEP) obtained from single eyes.
Patients and methods: Monocular mfVEP recordings were obtained from both eyes of 30 individuals with no known visual abnormalities. The 44.5 degrees -diameter display contained 60 sectors. Probability plots, analogous to the total deviation probability plot of the Humphrey Visual Field Analyzer, were created based on tests that compared each eye (monocular test), or the ratio of the responses from the 2 eyes (interocular test), to group norms.
Results: For the monocular test, the number of points exceeding the 5% confidence level was not distributed randomly among individuals or among field locations within an individual. Individuals with small signals (i.e., low signal-to-noise ratios [SNR]) showed too many "abnormal" points, while those with relatively large SNR values showed too few. Reasonably good specificity was obtained by defining an abnormality in terms of a cluster of significant points in the mfVEP probability plot. For the interocular test, the results were close to those expected by chance.
Conclusions: Both monocular and interocular tests will be of value when testing glaucoma patients with the mfVEP technique. The interocular test is a more sensitive indicator of glaucomatous damage when a region when healthy has a large signal (SNR) and damage is largely unilateral, whereas the monocular test will be a more sensitive test when a region when healthy has a small signal (SNR). However, the statistics underlying monocular comparisons of mfVEPs are not simple. To obtain high specificity, criteria based on clusters of points need to be used and norms obtained for every laboratory.