Frequency-domain analysis of the human corneal electroretinogram

The corneal electroretinogram (ERG) was measured in the dark-adapted state from 12 volunteers with normal vision. 2. Time-domain analysis included measurements of amplitudes and implicit times of the ERG components; a- and b-wave 3. For analysis in the frequency-domain, the power density spectra of the ERG signals were calculated and the dominant frequency and peak power density were derived. 4. The dominant frequency and peak power density of a given ERG response depended upon the duration of the segments specified for analysis; the shorter the segment the lower was the peak power density and the higher the dominant frequency. 5. The dominant frequencies of the ERG power density spectra exhibited large inter-subject variability and could not be used as a diagnostic index for small amplitude rod-dominated ERG responses. 6. The oscillatory potentials, isolated by an analog bandpass (100-1000 Hz) filter (Grass P511 preamplifier) were contaminated by contributions from the a- and b-waves. 7. The digital filter (Butterworth filter of order 5) was considerably better in isolating relatively pure oscillatory potentials. The power density spectra of these potentials, obtained from the 12 subjects, consisted of one major band with mean dominant frequency of 142.3 ± 9.8 (SD) Hz.