Abstract
Purpose :
The overarching goal of this research is to understand how the retina behaves under intermediate lighting conditions (e.g. dawn and dusk) when both rods and cones are active, and to relate the circuit-level retinal processing of rod and cone signals to human perception.
Previous human perceptual experiments have revealed interactions between flickering rod and cone stimuli that are thought to occur in the retina. Here we explore the neural basis of rod-cone flicker interference in On and Off ganglion cells that project to the primate magnocellular visual pathways.
Methods :
Electrophysiology
Spike recordings were taken from On and Off parasol retinal ganglion cells in an in vitro non-human primate retina preparation. Light from long and short wavelength LEDs was projected onto the isolated retina to preferentially activate L-cones and rods, respectively. The temporally-modulated, spatially-uniform stimulus was 2° on the retinal surface.
Psychophysics
After a dark adaptation period, human subjects focused on a small fixation cross while a spot of light (2°) flickered at 10° eccentricity. After each stimulus presentation (2 s) subjects were asked if they detected flicker. Answers were entered into a keyboard and the stimulus was adjusted accordingly until a perceptual flicker threshold was achieved. Similar to electrophysiology experiments, light from the red and blue monitor phosphors were used to to preferentially activate L-cones and rods, respectively.
Results :
Recordings from On and Off parasol ganglion cells revealed suppressive interactions when long and short wavelength LEDs were co-modulated at 8 Hz. Suppressive interactions became additive when a phase delay of 180° was introduced to one of the two stimuli or when the stimulus frequency was reduced to <6 Hz. Similarly, human subjects reported a boost in flicker threshold when stimuli were co-modulated at 8 Hz, but not at 4 Hz. Introduction of a 180° phase delay between the stimuli produced a dramatic reduction in flicker threshold at 8 Hz.
Conclusions :
This destructive interference between rod and cone signals appears to reflect a linear combination of kinetically-distinct rod and cone signals upstream of the ganglion cell synaptic inputs. Using our experimental data as a foundation, we construct a mathematical model that captures known rod-cone interactions and accurately predicts retinal output in response to arbitrary time-varying rod and cone stimuli.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.