Year: 2023
Type: dissertation
Abstract: Our understanding of choice and its neural code has greatly advanced over the last two decades. However, these advances have not accounted for our ability to use our past experiences to make choices we have never made before, or construct plans we have previously never thought about. The aim of this thesis is to provide evidence of such a neural code, how it is embedded in a broader prefrontal network during choice, and what mechanisms may drive our ability to piece together past experiences to construct new plans. In the first chapter, I show a grid cell-like encoding of value (i.e. a cognitive map) in a choice task, constructed as subjects make value-based choices. This pattern is observed in local field potential activity and neuronal firing rates of the ventromedial prefrontal cortex. The orientation of the grid cell-like code rotates across different task environments. The cognitive map exhibits distortions which are related to subjects’ choice patterns. In addition to a cognitive map, evidence for sharp wave ripples, a hallmark signal associated with memory retrieval and neural replay, is provided. In the second chapter, I investigate the connectivity structure of the prefrontal cortex during choice, and what the dynamics of this connectivity are. Ventromedial prefrontal cortex has outgoing functional connectivity with the orbitofrontal and anterior cingulate cortex. In contrast, the latter two regions, together with the dorsolateral prefrontal cortex, form a recurrent connectivity motif. This motif is modulated by task demands, and the efficiency of its coupling is related to subjects’ optimality in choice. In the third chapter, I develop a compositional reasoning task where human subjects continuously compose novel plans and provide preliminary neural evidence for the mechanisms allowing subjects to solve this task. These involve abstract representations of mental operations, and their coupling with representations of routes subjects must plan through. Overall, these findings provide a novel perspective for the neural representations and mechanisms supporting flexible behaviour: our ability to infer, generalise, and combine concepts we have never combined before.
Author: Veselic, Sebastijan
Language: English
Topic: Memory and Neural Mechanisms
Subfield: Cognitive Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7065453189