Research
OVERVIEW
Each day we make thousands of decisions about how to behave. Many of these decisions have a particular goal in mind - but how do we decide which goals are more valuable, and select actions appropriate to those goals?
I use a combination of human neuroimaging, invasive electrophysiology and computational modelling to aim to understand the neural basis of choice behaviour. I hope that by combining a description of decision-making at the single-cell and circuit level (via invasive electrophysiology) and at the whole-brain and behavioural level (via neuroimaging), it will be feasible to develop a more holistic understanding of choice that bridges microscopic and macroscopic levels of description. This will ultimately lead to a better understanding of therapeutic targets in neurological and psychiatric disorders in which decision-making goes awry.
MODELLING VALUE-GUIDED CHOICE DYNAMICS
We have investigated network models of interconnected spiking neurons that explain how a cortical region may support a choice via mutually inhibitory neural populations. Specifically, we have made predictions of decision-related activity in the local field potential (LFP) signal derived from the computational model. The LFP is closely related to the magnetoencephalographic (MEG) signal that can be recorded in humans (see video above). Thus, by comparing predictions from the model to human imaging data, we can infer which regions might be subserving similar computational processes, and identify regions supporting choice. I am currently working on several extensions to this, including direct comparison of neural population data with simultaneously recorded LFP data.
INFERRING OTHERS’ INTENTIONS
A key development in reward-guided decision making in the past two decades has been the application of formal computational models to explain variability in neural activity - for instance by looking for fluctuations that correlate with changes in a specific model parameter. We have extended this approach to capture a key process in theory of mind - namely, how do we infer the intentions of another individual with whom we are interacting? This has yielded a new dissociation between the sulcal and gyral portions of anterior cingulate cortex, with the latter appearing to be specialised for social cognition. It has also provided new insights into the computations performed by dorsomedial prefrontal cortex. More recent work has looked at how the use of social information in such settings might be altered in psychopathy.
Hunt et al., eLife, 2015
Behrens, Hunt and Rushworth, Science, 2009
Decisions about Decisions
Decisions are frequently described as a process of evidence accumulation until a categorical choice is made. How do we decide which evidence is worth gathering, and when to stop gathering information and execute a response? Is the process of evidence accumulation biased by which option is currently more preferable based on information already gathered, or by the strategy that the subject adopts? I am currently tackling these and related questions using a combination of behavioural modelling, functional imaging and single neuron electrophysiology.
Hunt et al., Nature Neuroscience, 2014
Video of MEG activity from
Hunt et al., Nature Neuroscience, 2012