Computational methods in neuroscience have advanced our understanding of neuronal regulation of motor behavior and locomotion and have been applied to identify encoding of behavioral features in circuits. The cerebellum has an established role in sensorimotor processing during coordinated movements, referred to as the “head ganglion of the proprioceptive system” (Sherrington, 1906). Increasing evidence also highlights its role in the processing of behaviorally meaningful stimuli that have the potential of guiding adaptative movements relevant to the task and priming downstream targets for action. Yet the extent to which these diverse encodings of signals in complex motor tasks are present in the cerebellar nuclei and their influence on behavior remains unknown.
To shed new light on the role of this subcortical region using computational approaches, this thesis begins with an introduction that reviews the circuity of the mammalian cerebellum, highlights its proposed functions in motor behavior, and explores our understanding of its role in locomotion. In the first chapter, I analyze electrophysiological recordings from cerebellar nuclei in a locomotor obstacle avoidance task in mice that involves a rich and diverse set of task relevant features. Given the complexity of and correlations between the behavioral features, statistical modeling is required to attribute the firing rates to the correct combinations. This model enables identifying the encoding of these signals and reporting on the prevalence and degree to which they are present across individual cells in the nuclei. Additionally, this model allows investigation into the encoding of groups of cells that are selective for specific features.
Chapter 2 uses network modeling to generate hypotheses about population level activity in two cortical areas, the primary and supplementary motor areas, and differentiate their computations in monkeys performing a cycling task. Finally, in chapter 3 I concentrate on a specific class of recurrent network models in the balanced state and investigate the linkage between connectivity distribution and firing sparsity, which has the potential to further our understanding on the emergence of feature selectivity in excitatory/inhibitory circuits.
| Identifer | oai:union.ndltd.org:columbia.edu/oai:academiccommons.columbia.edu:10.7916/hrrf-rc77 |
| Date | January 2024 |
| Creators | Khajeh, Ramin |
| Source Sets | Columbia University |
| Language | English |
| Detected Language | English |
| Type | Theses |
Page generated in 0.0019 seconds