Project Summary
The primary objective of this proposal is to map human retinal ganglion cell receptive fields in vivo, at
cellular resolution, directly connecting light stimulation of individual retinal photoreceptors to both the
activity of retinal ganglion cells and the conscious percept of retinal ganglion cell (RGC)
activity. Performing this task in vivo with human subjects will have enormous implications for the study of the
neural code leaving the eye.
There are 120 million photoreceptors in the human retina that absorb and respond to light. The human
optic nerve, however, the sole link between the retina and the brain, comprises a mere ~1 million axons. This
bottleneck shapes the neural code of the retina, which has never been completely described1,2. Past the optic
nerve, the transformation of information from the retina into a conscious percept involves neural interactions
that can usually only be studied through relatively coarse methods (such as fMRI). To understand the
organization of data leaving the eye, it will be essential to better connect the functional organization of human
retinal ganglion cells (RGCs) to human visual percepts.
It is difficult to study the neural code leaving the human eye because the optic nerve is inaccessible and
dense. Psychophysics using adaptive-optics-stabilized microstimulation of the photoreceptor mosaic is one
way to probe the information being sent from eye to brain3,4. This technique allows precise control of the input
to the photosensitive layer of the retina.
However, precise control of the inputs to the retina is not enough, since there are multiple parallel
pathways from retina to brain5. A pulse of light falling on a single cone will cause many different RGC signals
to be sent to the brain, most prominently ON and OFF of both midget and parasol, the most common RGC
types6. Disentangling the percepts–what a human subject can report as "seen"—elicited by the different RGC
types is a difficult problem.
In order to probe the perceptual correlates of selectively stimulated groups of RGCs, this
project will leverage perceptual desensitization with adaptive-optics-stabilized microstimuli. First a
“desensitization” stimulus will habituate several RGC pathways with single cone stimulation, and then a “probe”
stimulus will be used to target a single RGC type psychophysically, one cone at a time. In this way, a single
RGC "perceptive field" can be mapped and investigated. Once mapped, RGC perceptive fields can be
subsequently investigated. This work will be performed in Prof. Austin Roorda's adaptive optics psychophysics
lab at the University of California, Berkeley.
Single cone stimuli are unusual in the ex vivo literature. To confirm that desensitization and probe
stimuli are driving the RGCs as expected, this project will also involve testing desensitization and probe stimuli
using electrophysiological techniques in primate retina. Further, the design and evaluation of single cone
pulsing light stimuli that maximally drive some RGC types over others will require ex vivo work. This pursuit
will also be a potentially clinically relevant one, since stimuli that could maximally target RGCs that are
implicated in the early stages of retinal disease, such as OFF RGCs in glaucoma7, could be used to design
future diagnostic tools. The electrophysiology work will be performed in the lab of retinal physiologist Prof.
Teresa Puthussery at UC Berkeley.
If successful, this project could lead to novel clinical diagnostic tools for the treatment of retinal
diseases and inform the understanding of how the neural code from the eye is interpreted by the brain.