Abstract
Introduction: It has recently been shown, using silent substitution stimuli on a CRT display, that the human flicker electroretinogram (ERG) can reveal the activity of both parvocellular and magnocellular mechanisms at different temporal frequencies, 30Hz showing luminance activity and 12Hz showing chromatic opponent activity. Elsewhere at this meeting (McKeefry et al.) we show the variation in this across the retina, when changes in L-M ratio are only seen at 30Hz. In the current study we employed an LED ganzfeld stimulus (Diagnosys ColorDome) to simultaneously investigate chromatic and achromatic processing in the ERG using a temporal compound stimulus.
Methods: The stimulus comprised a succession of alternate red and green modulations. In odd-numbered cycles, the green LEDs were modulated from black to a given luminance with a raised cosinusoidal profile, while the red LEDs were off. In even-numbered cycles, the red LED was switched off and the green LED was modulated. The stimulus cycled continuously so that chromatic modulation occurred at half the frequency of the achromatic modulation (i.e. period doubling). All frequencies are expressed as nominal frequency, i.e. that of the luminance component. Mean luminance was held constant at 20cd.m‒2 but the red and green peak luminances could be reciprocally adjusted.
Results: ERGs were first recorded to a range of red-green ratios at 36Hz (nominal frequency). The 36Hz component stayed approximately constant throughout the range, with an amplitude of 15-25 uV (depending on the subject), while the 18Hz component reached a minimum, approaching zero in the middle of the range (at each subject's psychophysical isoluminant ratio). Further runs were conducted using this ratio, thus ensuring that luminance intrusion in the chromatic response was mimimised. ERGs were then recorded to a range of temporal frequencies between 12 and 80Hz (nominal frequency). Lower frequencies were avoided to minimise possible rod intrusion. The luminance component showed a band-pass characteristic, peaking at around 40Hz and falling close to zero at 40Hz. The chromatic component was low-pass and approached zero at 40Hz (i.e. 20Hz chromatic modulation).
Conclusion: Using this temporal compound stimulus, period doubling suggests involvement of parvocellular units, whereas responses at the nominal frequency reflect magnocellular activity. Thus it is possible to quickly determine ERG responses which reflect activity in chromatic and luminance pathways when they can be distinguished on the basis of their different temporal responses.