Date of Award

Summer 8-15-2021

Author's School

Graduate School of Arts and Sciences

Author's Department

Physics

Degree Name

Doctor of Philosophy (PhD)

Degree Type

Dissertation

Abstract

Animals process high-dimensional sensory information constantly. How does neural activityin sensory cortices represent this information? Recent advances in large-scale recordings allow us to monitor activity of hundreds or thousands of neurons simultaneously across a long period of time. Population recordings showed that cortical neuronal responses to repeated sensory stimulation is highly variable from trial to trial. However, how neurons in neocortex represent sensory information amid high neural variability is not well understood. To answer this question, we used two-photon calcium imaging to record from hundreds of excitatory neurons simultaneously from mouse primary visual cortex. We analyzed neural responses to repeated visual stimulation to reveal the neural representation of external variables and investigated the neural circuit mechanism underlying the observed neural representation by neural network modeling. Our results quantified how single neurons responses to repeated visual stimulation varied from trial and trial and from week to week. Using unsupervised methods, we identified interpretable directions in the neural space, which encode distinct external stimulus-driven or stimulus-independent variables. Together, these results contribute to a better understanding of how high-dimensional neural activity encode distinct variables despite high neural variability.

Language

English (en)

Chair and Committee

Ralf Wessel

Committee Members

Anders Carlsson, Michael Goard, Baranidharan Raman, Mikail Tikhonov,

Share

COinS