Computational Modelling of Decision Biases in Anxiety Disorders

Anxiety is a very common type of mental illness which is estimated to affect up to a third of the world's population and which manifests itself in different forms and with different intensities. The nascent field of Computational Psychiatry aims to investigate psychiatric disorders using computational modelling and techniques in order to provide mechanistic understanding of the causes and mechanisms of mental illnesses and to provide more refined and better treatments to patients.



In anxiety, decision making is impaired in different ways, for example by an increase in threat-avoidance behaviours. My interest lies in understanding how anxiety distorts the expectation of negative outcomes (threats) and how it affects the balance between exploration and exploitation as well as risk and loss aversion. These differences in decision making will be modelled using Reinforcement Learning models, which have already been successfully used in the context of anxiety disorders. Reinforcement Learning models can also be complemented with neural computational modelling based on biophysical findings of brain regions connected to anxiety.



The literature reports contradictory results regarding the differences in decision making in anxious individuals. Therefore, being able to formalise the way in which patients use and predict negative information is particularly useful for targeting and refining treatments such as cognitive behavioural therapy.