Active Inference and Optimal Decision-Making (360G-Wellcome-203922_Z_16_A)

Depression is strongly associated with a decline in cognitive function, and is seen in a high number of elderly individuals. It has been suggested that there is a strong epidemiological link between a personal history of depression and an increased risk of developing Alzheimer’s disease in later life. The key aim of this project is to use models of learning and decision-making to better link behavioural characteristics of these two psychiatric illnesses to their brain circuit substrates. We aim to reveal model-based behaviours and brain circuits that are shared with depression and different dementia profiles, including Alzheimer’s disease. We will use an adapted version of a pre-existing video game environment to examine activation of the dopaminergic midbrain, using fMRI. We will look specifically at reward prediction errors, and precision, i.e. the degree of confidence in the action an individual takes. By describing these behaviours using computational models, we hope to provide more accurate descriptions of human behaviour, the causes of such behaviours, and the specific brain regions and neural circuits that these behaviours are linked to. This may lead to earlier diagnosis of such psychiatric illnesses; therefore patients can undergo treatments earlier on in the disease progression.