Frontal cortical mechanisms and interactions during learning and decision making

The aim of the current proposal is to look at the neural mechanisms of learning and decision making. Learning, the acquisition of new behaviours and information, and decision making, the ability to act on that information, are two of the most fundamental cognitive operations that are called into play throughout our lives. Impairments in learning and decision making are often central to psychiatric conditions; maladaptive choices are not just made in the first place but they are also not adjusted as a result of experience. It is important to emphasize that not only do we investigate basic aspects of learning and decision making but, for example, we also attempt to understand such processes in the social domain.
Learning and decision making do not depend on a single brain region but on the interactions that occur between many regions. Our aim is to investigate these interactions in five key circuits. The focus will be on the parts of the five circuits that are located in the frontal lobes of the brain and the aim is to test whether and how the frontal components of these circuits regulate activity in the rest of the circuits of which they are parts.
The interactions within brain circuits will be investigated in two principal ways. First, we will attempt to manipulate and alter activity in one component of each circuit and then record the consequences for behaviour and for activity recorded in other parts of the circuits. This can be done by making a selective lesion in a brain area and for this reason it is necessary to use animal models. In the current case the brain areas are only present in a few species and so the macaque is the model species. The lesion?s effect on other parts of the network can be investigated by using a non-invasive brain scanner technique, functional magnetic resonance imaging (fMRI), to record the blood oxygen level dependent (BOLD) signal that provides an index of brain activity. The second way that we can look at brain circuits is by using the brain scanner to acquire a different type of MRI data, diffusion weighted MRI (DW-MRI). DW-MRI scans provide information about the white matter pathways that run between different component parts of brain circuits. We plan to look at how the pathways linking brain circuits change during learning to make decisions in the most effective manner possible.

Prefrontal and anterior cingulate cortex (PFC and ACC) play fundamental roles in voluntary or goal-guided behaviour. They are essential for decision making and learning. Changes and biases in such cognitive processes are central to the behavioural alterations seen in psychiatric illnesses.


The aim of the current proposal is twofold. First, we intend to uncover the mechanisms of operation of PFC and ACC by using computational approaches to describe the effects of selective brain lesions and patterns of brain activity recorded with functional magnetic resonance imaging (fMRI). Second, we intend to uncover the manner in which these brain areas are able to perform such functions through their interactions with other brain regions. Combined lesion-fMRI studies will address this second aim. The work will be carried out using a macaque model and will employ the Oxford MRC-funded macaque MRI facilities.
The focus is on five brain regions.


It is hypothesized that ventromedial frontal cortex is essential making value-guided decisions because it rapidly focuses on key choice options; after a lesion it is hypothesized that activity in other brain regions will no longer be correlated with the values of the key choice options that might be taken. It is hypothesized that lateral orbitofrontal cortex and dopamine make distinctive contributions to the learning of stimulus-reward associations; while the former learns precise associations the latter reinforces choices in proportion to their proximity to reward regardless of the exact nature of any contingency. Third, it is hypothesized that frontopolar cortex has a special contribution to make to learning because it mediates counterfactual learning; the ability to learn not from experience but from observation of consequences of choices not taken. Fourth, it is suggested that the anterior cingulate sulcus has a role in assigning value to the component stages of a course of action that determines how primates engage in, start, and stop courses of actions. Fifth, it is suggested that the anterior cingulate gyrus has a role in the valuation of social information that is related not just to the manner in which the social status of conspecifics is perceived but to the manner in which an individual expresses social status. Finally it is hoped that provision of the resources requested will make it possible to continue a programme of work using diffusion weighted MRI to compare anatomical structure in humans and macaques and investigating individual differences in learning induced changes in anatomical connexions.