Advisor(s)
Lawrence Sincich
Committee Member(s)
Allan Dobbins
Julie Quinet
Lynn Dobrunz
Timothy Kraft
Document Type
Dissertation
Date of Award
6-1-2026
Degree Name
Doctor of Philosophy (PhD)
School
Joint Health Sciences (Interdisciplinary)
Department
Neurobiology
Abstract
Neurons in the retina and lateral geniculate nucleus (LGN) facilitate visual experience by detecting and encoding features of luminous stimuli, but fundamental questions regarding their handling of spatial and temporal contrast information remain unanswered. We utilized an adaptive optics scanning laser ophthalmoscope (AOSLO) to address gaps in our understanding of photon encoding by early visual neurons via near diffraction limited imaging and stimulation of macaque retinas in vivo. LGN receptive fields were comprehensively mapped and aligned to the underlying cone mosaic. Receptive field centers within 0.42–4.11° of the fovea were most frequently associated with single cone inputs, physiologically demonstrating parvocellular LGN neurons mediate foveal vision at the limit of cone spacing. Our results were confirmed by biophysical light capture modelling and spatial tuning data and resolve longstanding tensions between anatomical and physiological evidence on the limitation of visual acuity prior to cortical processing. To understand how LGN neurons respond in vivo to spatial and temporal contrast patterns during ongoing stimulus conditions at the photoreceptor level, we developed a novel technique which asks how neural responses are influenced by individual pixels of noise stimulus movies. Simple preferred and non-preferred contrast steps drove symmetric LGN responses, equal in magnitude but opposite in direction, while stimuli incorporating spatiotemporal features resembling an LGN neuron’s receptive field elicited stronger, asymmetric responses weighted towards preferred stimuli. Luminance adaptation and relative LGN receptive field surround strength were also probed by this technique using spatially and temporally correlated stimuli. Finally, we investigated if human psychometric thresholds arising from single cones were shaped by signal encoding in RGCs or LGN neurons. Three schemes of encoding contrast increments or decrements were investigated, and each were fit with three saturating functions to construct neurometric curves. We found greater diversity in neurometric functions between RGC classes than LGN cell classes, suggesting more distinct contrast sensitivity at the retinal level than the thalamic level. Our analysis also shows that both RGCs and LGN neurons play a role in steepening the slope of psychometric performance curves arising from single cones by enhancing the signal-to-noise ratio of cone-originating visual signals.
Keywords
AOSLO;contrast encoding;LGN;macaque;receptive fields
ProQuest ID
Recommended Citation
Ramsey, Keaton Miller, "Photon Encoding In The Early Visual System" (2026). ETDs from 2020-2029. 160.
https://digitalcommons.library.uab.edu/etd-2020s/160
