Abstract
Introduction. It is generally understood that the full-field dark-adapted ERG is dominated by the scotopic spectrum of the rod response (Holopigian et al., 1992; Hood & Birch, 1992), although short-wave anomalies have been reported that could be attributed to a melanopic component (e.g., Bierdsdorf & Grand, 1962), and fast melanopic ERG responses have been reported in silent substitution protocols (Fukuda et al., 2010). Moreover, long-wave anomalies in the ERG have long been known in the form of the short-latency x-wave (Adrian, 1946; Armington, 1952). Methods. We examined the spectral sensitivity of the a/b-wave complex in human ERGs recorded with skin electrodes across the spectrum over three log units of intensity. Results. The long-latency component at all test wavelengths and intensities conformed to a single function of scotopic intensity (rod quantum catch), which we term the "main sequence" for ERG times-to-peak as a function of intensity. However, responses to red lights showed an extra short-latency peak (x-wave) at high intensity attributable to cone responses. Nevertheless, rescaling the data as a function of photopic intensity did not result in a clean alignment of the a-waves across the spectrum, as would be predicted by the cone attribution. The a-wave latencies, however, became aligned by scaling the data as a function of equal-energy radiance. Conclusion. The results suggest a reconceptualization of the spectral sensitivity of the dark-adapted cone a-wave (and the subsequent x-wave) as neither photopic nor scotopic, but conforming to the radical hypothesis that the a-wave responses throughout the spectrum derive from a pre-luminance sensitivity to the radiant energy impinging on each class of photoreceptor. This radiant energy baseline provides a firm basis for assessing the melanopsin contribution to the ERG response.
Meeting abstract presented at VSS 2016