Understanding decision making by linking brain and behaviour

Project Information

BrainBehaviourModel

Grant agreement ID: 795866

DOI

10.3030/795866

Project closed

EC signature date 13 April 2018

Start date 1 June 2018

End date 31 May 2020

Funded under

• EXCELLENT SCIENCE - Marie Skłodowska-Curie Actions

Total cost

€ 165 598,80

EU contribution

€ 165 598,80

Coordinated by

UNIVERSITEIT VAN AMSTERDAM
Netherlands

Periodic Reporting for period 1 - BrainBehaviourModel (Understanding decision making by linking brain and behaviour)

Reporting period: 2018-06-01 to 2020-05-31

Summary of the context and overall objectives of the project

Decision-making is an integral part of our daily lives. We regularly face decisions in simple and more complex situations: “Is it safe to cross the road?” or “Should I buy this house?”. Even if an optimal choice might exist objectively, people vary in decision-making behaviour due to differences in the information that they pick up, personal goals and preferences, and traits such as impulsivity or reasoning ability. One of the challenges for researchers in the field of cognitive neuroscience is that we can only observe the outcome of someone’s decision. We cannot directly measure the (unconscious) thought process that led to it. Over the last few decades, insightful mathematical models have been developed that allow us to estimate hidden cognitive parameters that influence our behaviour, such as the rate of evidence accumulation and the decision threshold. Individual values of these parameters may serve as markers for working memory capacity and IQ, as well as for clinical symptoms such as ADHD, high-anxiety and depression. Yet, we are only starting to grasp how cognitive parameters are encoded in terms of brain activity. While we know that individual values of cognitive parameters correlate with metabolic changes (fMRI-BOLD responses) and with the amplitude of electrophysiological signals (EEG) from specific brain regions, we do not understand how brain regions work together to control our decision-making behaviour. The main objective of the research project is to shed light on the neurobiological processes underlying cognitive parameters. It combines mathematical models of decision-making behaviour with computational models of brain activity that can explain how measured EEG activity may arise from synaptic interactions between brain regions. Furthermore, the project places special emphasis on the role of the basal ganglia, a network of nuclei that are located deeply inside the brain that are considered to be pivotal for action selection. Understanding the exchange of information between basal ganglia nuclei and the cortex is important for optimization of deep brain stimulation treatment in neurological and psychiatric disorders such as Parkinson’s disease and obsessive-compulsive disorder, where disturbances in decision-making behaviour may strongly affect the quality of life.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

Two new data sets were collected with EEG recordings from healthy participants performing simple decision-making tasks. The first data set was used to obtain a proof-of-principle for the novel modelling approach that combines mathematical models of behaviour with computational models of EEG activity. This revealed that both bottom-up and top-down information flow between cortical regions influences the rate at which evidence is accumulated during the decision process. Results of this study are currently being finalized and prepared for submission to a scientific journal.
The second data set was obtained to further apply the novel modelling approach to other aspects of decision-making, in particular the effect of urgency by imposing a strict deadline on the decision time. This revealed a more rapid build-up of the centroparietal EEG-ERP component when participants were required to respond before an imposed deadline compared to when they could decide freely. The application of the modelling approach to this data is currently still ongoing and expected to be finalized and prepared for submission to a scientific journal within the next year.
A biologically-detailed computational model of cortico-basal ganglia networks was implemented in a commonly used open-source software program. Using this model, we demonstrated that the effect of dopaminergic medication on abnormal neural oscillations in Parkinson’s disease can be explained by wide-spread alterations in coupling strength between regions of the network. This study was published in a peer-reviewed scientific journal.
Using simultaneous recordings from deep brain stimulation electrodes and MEG, we demonstrated that separate functional networks between the cortex and subthalamic nucleus project to different parts of the subthalamic nucleus but with large spatial overlap. This suggests that it might be very difficult to entirely avoid neurocognitive side effects during deep brain stimulation treatment of Parkinson’s disease. This study is currently under review for publication in a peer-reviewed scientific journal.
We wrote a comprehensive review on the anatomy and functional organization of the subthalamic nucleus, the primary target for implantation of deep brain stimulation electrodes for treatment of Parkinson’s disease. This knowledge is important for unravelling the role of the subthalamic nucleus in controlling limbic, cognitive, and motor functions, and for understanding whether these functions can be seen as clearly independent of one another. The review was published in a peer-reviewed scientific journal.
Throughout the timeline of the Fellowship, project plans and results have been communicated to other researchers at international conferences (Cosyne 2019, HBM 2019), institutional seminars (Amsterdam, Moscow), and lab meetings (Amsterdam, London, Shanghai). In addition, two book chapters were written for a lay audience on the principles behind EEG and neural oscillations in motor control and decision-making.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

The new software implementation of the biologically-detailed computational model of cortico-basal ganglia networks has been made freely accessible to the research community via one the largest open-source software programs for the analysis of neuroimaging data. It has already found its application in other research projects that have led to scientific publications and in projects that are currently ongoing.
The EEG studies in this project are the first to combine mathematical modelling of behavioural responses with computational modelling of EEG responses. This is a crucial step towards a full integration of cognition and neurobiology in a single model, which would be the ultimate goal for explaining how the brain controls the decisions we make. Although cognitive models of behavioural responses and neural models of EEG responses are both established methods in their respective disciplines, it is their combination that brings novel insights. Importantly, the novel modelling approach hols the potential for bringing abstract theories in psychology a step closer towards their neurobiological implementation. The approach can be readily extended to more complex decision-making processes that involve, e.g. strategic choices, values and emotions, and social factors, and to other research domains in (cognitive) neuroscience such as memory, perception, and language. It is therefore expected to find wider research applications in the next years.
Atypical decision-making behaviour in the form of systematic impulsive actions or choices may become problematic in ADHD, obsessive-compulsive disorder, gambling and drug addictions. Unravelling how our decisions are rooted in neurobiological mechanisms helps us to understand the factors that influence our behaviour, and inter-individual differences in behaviour. In the long-term, the novel modelling approach might significantly contribute to predicting effects of pharmacological interventions, stimulation protocols, and lesions on decision-making behaviour.

figure-brainbehaviourmodel.png