Tanni, Sander; (2020) Neural correlates of navigation in large-scale space. Doctoral thesis (Ph.D), UCL (University College London).

Preview

Text Sander Tanni Thesis - copyright figures redacted.pdf Download (8MB) | Preview

Abstract

Navigation and self-localisation are fundamental to spatial cognition. The cognitive map supporting these abilities is implemented in the hippocampal formation. Place cells in the hippocampus fire when the animal is at a specific location – a place field. They are thought to be involved in navigation and self-localisation but usually studied in constrained environments, limiting observable states. In this thesis, I present two experiments studying place cells in large open field environments, a novel auditory cue-triggered navigational task, and a technical solution for conducting large scale automated experiments. Place cells are frequently reactivated during immobility, rapidly replaying trajectories through environments. These replay events are thought to be involved in navigational planning. Using a novel automated cue-triggered navigational task in a large open field environment, I show that replay is not associated with navigation to the goal. Instead, it occurs reliably at the end of successful trials, when an associated reward is received, but not during consumption of scattered pellets. The trajectories in these events are predictive of the animal’s movement after, but not before, the reward. The number of place fields per cell, their size and other properties have not been fully characterised. Using multiple large open field environments of different size, I show that place field size, shape and density changes systematically with distance from walls. However, through a homeostatic mechanism, the mean firing rate and proportion of co-active units in the population remains constant throughout environments, as does the accuracy of their spatial representation. Multiple place field properties are conserved by cells across environments, including the number of fields, which is quantified relative to environment size using a gamma-Poisson model. Place cell population models suggest two sub-populations, with uniform and boundary dependent field distributions. These results provide a comprehensive account of place cell population statistics in different size environments.

Download activity - last month

Export as CSVJSONXML

Download activity - last 12 months

Export as JSONXMLCSV

Downloads by country - last 12 months

Export as XMLJSONCSV

Archive Staff Only

View Item