The retina is tasked with detecting light from the environment, and transmitting that visual information as an electrical signal to the brain. Visual signals are highly variable, making it difficult or impossible for retinal neurons of a single type to fully represent all of the salient visual features. Instead, the retina splits the visual input into many parallel channels, represented by the 30 or more ganglion cell types that signal to the brain. Each ganglion cell type, or channel, encodes a slightly different feature of the visual scene, such as high or low spatiotemporal frequencies, positive or negative contrast, direction of motion, and edge orientation, among others. Many of these feature selectivities are driven by the activity of complex inhibitory networks. Understanding the computations being made by different ganglion cells requires understanding how these inhibitory circuits are arranged, how light activates them, and how they influence the electrical signals being transmitted to ganglion cells. This dissertation presents experiments designed to address these questions in an inhibitory neuron that reports edge orientation, and in a ganglion cell that preferentially responds to fast image motion. The results are an important demonstration of how seemingly small details about a neuron’s presynaptic circuit are critical for shaping its emergent function and feature selectivity.
School of Medicine
Murphy-Baum, Benjamin, "Inhibitory Circuits Generate Distinct Spatiotemporal Receptive Field Properties in Retinal Neurons" (2017). Scholar Archive. 3882.
Available for download on Friday, March 20, 2020