Spatiotemporal characterization of value judgments and reward processing in the human brain

The main goal of this project is to provide insights into how people make everyday value and preference judgments. For instance, how do we choose among different goods at the supermarket or how do we decide which mobile phone to purchase? How do we weigh the pros and cons of the various options to make our choices as rewarding as possible? How do we make use of our prior experience with the various alternatives to help us make better choices and obtain bigger rewards in the future? How do we learn, through trial and error, to update our future predictions in order to adapt and keep up with the competitive and ever-changing world around us? Importantly, we do not want to merely measure how people behave in some of these scenarios but we would also like to understand how the human brain processes the relevant information and how the outcome of that process ultimately cause one to express or externalize (i.e., by acting out) their preferred choice.

In recent years, modern neuroimaging methods have made it possible for researchers to "take a peak" into the human brain. Human electroencephalography (EEG) is a non-invasive method whereby electrodes placed on the surface of the scalp measure the brain's electrical activity. This technique provides precise information on when the brain processes information in time but cannot provide a clear picture of where in the brain this information is coming from. In contrast, functional magnetic resonance imaging (fMRI) measures (non-invasively in an MRI scanner) changes in the local blood flow in the brain as we process information. This in turn allows one to identify where in the brain this processing takes place but at the expense of poor precision in terms of when processing occurs in time. It is therefore easy to see that if these techniques were applied simultaneously, they could clearly compliment each other and help provide answers as to both when and where the brain processes information.

To perform concurrent EEG and fMRI experiments, however, one needs to overcome a series of technical challenges and devise an analysis methodology that will integrate the two datasets in a manner that circumvents the disparate space and time scales on which they are acquired. In collaboration between the Schools of Psychology and Physics at the University of Nottingham as well as our international partners (i.e., Department of Biomedical Engineering at Columbia University in New York), we have started to use some cutting-edge technology and analysis tools to overcome these challenges and make these simultaneous measurements possible. In this project we will recruit volunteers who will be asked to perform simple value and preference judgments while we simultaneously record EEG and fMRI data from them. We will then use the methodology we have recently devised to look at the data in an attempt to provide answers to the questions outlined above.

Crucially, understanding which brain areas are involved in value judgments, how they interact with one another and at which times can potentially lead to practical applications. For instance knowing how to tab into and monitor the various brain processes involved in value-based decision making can have applications in domains as diverse as economic and public policy analysis (e.g., when deciding on savings strategies and health behaviours) and marketing (e.g., when optimizing advertisement strategies and product design). In addition, improved understanding of the neurobiology of decision-making can potentially have applications in identifying prognostic indicators of normal and abnormal ageing or in defining precursors of disorders known to compromise ones decision-making faculties (e.g., autism, personality disorders, etc).

In recent years the study of the neurobiological and computational basis of value-based decision making has received considerable attention and it provided the foundation upon which the field of neuroeconomics was built. Despite recent progress in understanding the neural correlates of value-based decisions, key questions pertaining to how value is computed at a mechanistic level remain. Specifically, it is unclear whether sensory regions are modulated by the value (or probabilistic evidence) conferred by different decision alternatives and how and where this information is combined to generate the value signal needed to make a decision. Another essential element of valuation pertains to how we process rewards to update future value expectations. Accurate reward representations associated with potential choices are critical for adaptive decision making. These representations can be acquired with reinforcement learning mechanisms, which use the prediction error (PE), the difference between expected and actual rewards, as a learning signal to update expectations. The spatiotemporal dynamics of the network associated with PE processing and the mechanisms by which PE valence and magnitude are used to update expectations and guide future choices remain inconclusive. In this project we will use state-of-the art neuroimaging to couple single-trial analysis of the EEG with simultaneously acquired fMRI to infer, with both high temporal and high spatial resolution information, the cascade of constituent cortical processes involved in value judgments and reward processing in humans. Our working hypothesis is that trial-to-trial variability in electrophysiologicaly-derived measures has information content that is meaningful in representing the dynamics of latent brain states that are unobservable via stimulus or behaviourally derived measures, which in conjunction with fMRI can be used to expose the cortical networks involved in value-based decision making.

The potential impact of the developments supported by this application could address several of BBSRC's new strategic priorities (e.g. economic, social and public policy impact, cognitive ageing, systems approach to biological research).

Economic, social and public policy impact: In our everyday lives we constantly engage in value and preference judgments to make choices that are more likely to lead to desirable outcomes. Knowing how to tab into and monitor the various brain processes involved in value-based decision making can have applications in domains as diverse as economic and public policy analysis (e.g. to inform decisions on health behaviours and savings strategies) and marketing (e.g. to help optimize advertisement strategies and product design).

Imagine for instance having to make a choice between a high-calorie strawberry smoothie and a healthy alternative such as a fruit salad. Obesity is a growing problem in the western world and can have direct and severe consequences on people's health as well as indirect costs on a nation's health care system. A more thorough characterization of the cortical networks underlying preference judgments and reinforcement-guided decision making can potentially lead to ways of enhancing the value of less-preferred choices (in this example low-caloric foods), possibly through feedback protocols that make use of incentive motivation to reinforce and promote optimum (here healthier) behaviour.

Another translational aspect of this work relates to the newly developed field of neuromarketing. For example, how does one develop an effective and influential advertisement campaign (e.g. to promote a new product or service and raise public awareness for key health and socioeconomic issues such the one outlined above)? Traditional marketing techniques rely heavily on market research through cumbersome approaches (e.g. through surveys and questionnaires) that can suffer experimenter biases and that are often hard to quantify. In contrast, neural signatures of value can provide a true quantitative measure of internal preference and as such can be used to triage through various advertisement pieces and product designs.

Cognitive ageing and disorders: In normal and abnormal cognitive ageing even simple decision making is effected, as evidenced by changes in behavioural measures such as accuracy and response time. A number of disorders known to compromise ones decision-making faculties (e.g. autism, personality disorders) can also lead to similar behavioural changes. The imaging tools and methods proposed here could potentially be used to characterize the neural changes that are causal to these behavioural changes. As such this work can lead to the development of new methods for identifying diagnostic and prognostic indicators of cognitive deficits, as well as more directed behavioural and pharmacological treatments.

Systems approaches to biological research: This work also addresses BBSRC's priority on "exploiting new ways of working". It is practically axiomatic that new tools, particularly in neuroimaging, lead to new observations and more fundamental understanding of the processes under observation. For decision making some of the neural processes underlying it can only be observed via trial-to-trial electrophysiological variability. All fMRI studies of decision making to date, more-or-less, look at mean changes and do not consider trial-to-trial changes (other than those that can be measured behaviourally). Here, we will develop a new methodology that exploits trial-to-trial changes in EEG (via multivariate analysis techniques) to drive the analysis of simultaneously acquired fMRI data in order to identify latent brain states associated with value-based decisions. Here, this new approach will be used to address issues in some of BBSRC's priority areas (as outlined above) but at the same time has the potential to become a powerful new tool in cognitive neuroscience in general