Neural mechanisms of memory & prediction, finding structure in experience

Higher cognitive functions by which we learn to understand our environment and behave flexibly within it rely on the construction, from sequential experience, of a coherent representation of our situation (aka an internal ‘model’ or ‘cognitive map’). I aim to understand the neural basis of his process via precise mathematical models that directly link low-level neuronal mechanisms to behaviour in prediction, planning, generalisation and memory. This project involves convergent computational and experimental work in mice and humans at the neuronal, systems and behavioural levels, using methods ranging from multi-photon imaging in mice, through virtual reality and neural-level electrophysiology in mice and epilepsy patients, to functional brain imaging in healthy volunteers. Three streams of work address the integration of sensory information and actions into a common representation, how representations of states structured by transitions allow path integration and prediction for planning, and how such representations interface with encoding and retrieval during memory. Success will provide a quantitative understanding of how neural activity in medial temporal, retrosplenial and medial prefrontal brain areas support these complex cognitive functions, and a starting point for relating cognitive symptoms to dysfunction of this neural system in conditions such as posttraumatic stress disorder, schizophrenia and Alzheimers disease.