V 193Components of MF-ERG signals and important methods of analysis
M. W. Seeliger
There are topographical changes of multifocal ERG (MF-ERG) waveforms across the retina in humans and monkeys. In the latter, waveform shapes differed considerably between different retinal locations of the same eccentricity, especially prominent between the nasal and temporal side. Such a behavior can be modeled by a set of oscillatory components that are shifted relative to each other, depending on the topographical location. The typical "banana shape" of the peak implicit times of ring averages (slow central response, faster paracentral responses, slower peripheral responses) could also be reproduced in the simulation.
The suggested minimum-standard of data analysis is a set of three graphs, one of the trace array, one of the ring averages (if applicable), and one of the scalar product (3D) plot. The trace array shows the local ERG waveforms plotted at their topographical position. Differences in amplitude are easy to spot, but implicit times can hardly been seen. Ring averages sacrifice much of the resolution, but allow for a good estimate of implicit times. Due to the differences in waveform shape most apparent in the temporal-nasal gradient, it appears not optimal to use ring or even overall averages of MF-ERG data for analysis. The 3D-plot shows the scalar product between the actual data and a reference template, but contains the least information.
Retinal Electrodiagnostics Research Group, Department II, University Eye Hospital, Schleichstraße 12-16, 72076 Tübingen