Grant: Doctoral Research Award: Canada Graduate Scholarships

People with concurrent disorders (PCD; or co-occurring mental and substance use disorders) have worse health and treatment outcomes than any other mental health group. Research on PCD and their treatment outcomes is severely lacking, as they are often excluded from studies due to the severe and complex nature of their disorders. Recent work has shown that decision-making data, modelled with computational techniques, is able to predict treatment outcomes in PCD. This project will use fMRI to uncover the neural basis of computationally-modelled decision-making data in PCD. Concurrent disorder inpatients from the Red Fish Healing Centre (N=30, 15F) and controls (N=30, 15F) will complete two decision-making tasks: the Cambridge Gambling Task (CGT) and Probabilistic Reinforcement Learning Task (PRLT) while undergoing fMRI scanning. Trial-by-trial decision-making data will be modelled using the Cumulative Model (CM) for CGT data and a Q-Learning model adapted to PRLT data. fMRI data will be used to (1) characterize brain region activation corresponding to the components of decisional processing, and (2) compare differences in patterns of activation between PCD and controls. Trial-by-trial computationally-modelled decision-making data will be regressed against fMRI signals using a general linear model to determine how model parameters correlate with individual brain activity, and group-level random effects analysis of variance will be used to compare regional activation between patients and controls. This data will reveal key neurocognitive underpinnings of risk-taking in PCD and inform the development of much-needed treatment interventions in this population.