Profile
Project Title:
Exploring Mechanistic Correlates of Cognition via Neuroimaging and Information Theory
Project Summary:
The brain is primarily an information processing organ. However, traditional brain-scanning approaches, not directly underpinned by an informational framework, are limited in providing mechanistic neural explanations for cognitive processes. This project aims to correct this flaw by applying cutting-edge techniques from information theory to neuroimaging.
A recent advance in information theory, Partial Information Decomposition, provides mathematical tools which not only quantify the amount of information processing in the brain, but also decompose it into different types, which may map closely onto different kinds of cognition. Initial application of this method to neuroimaging data has shown that ‘basic’ somatosensory areas of the brain are associated with Redundant information, information duplicated across brain regions, while brain areas associated with ‘higher’ cognition are also associated with Synergistic information, information only available when regions are integrated. (Luppi et al 2022, Nat-Neuro)
However, the full potential of this method has not yet been explored. It is possible to decompose information into many more types, for instance information transfer or emergence, but this has not yet been explored in neuroimaging data. This project will do this; by using pre-existing large neuroimaging datasets in rest and task we will explore the full taxonomy of information processing dynamics and investigate how these distinct forms of information processing are employed across different brain regions map to perform complex cognitive processes
Supervisors: