Computational psychiatry attempts to translate advances in computational neuroscience and machine learning into improved outcomes for patients. Here, I describe recent work on mechanistic approaches to support the correct assignment of psychotherapeutic interventions to individuals.

Psychotherapies are one of the core treatment options available for depression. However, despite an extensive theoretical basis for interventions, our understanding of the underlying mechanisms mediating treatment response remains poor. Here, I will describe work suggesting that a combination of computational models and cognitive tasks may enable the measurement of the cognitive processes engaged in therapies. Critically, for the case of cognitive behavioural therapy for depression, we find a double dissociation, with effort-reward tradeoffs engaged preferentially altered by behavioural activation, and learning about attributions preferentially altered by cognitive restructuring. Furthermore, improvement in symptoms in a realistic treatment setting is related, and possibly mediated, by changes in pavlovian biases measuring using a task and a computational model. Finally, the cognitive measurement process enables us to design novel training interventions.

This seminar is hosted in person in the Department of Psychiatry, to join online, please use the joining link below:

zoom.us/j/94567124781?pwd=sVxXabbSWibdU8A9W2clQlG9neRGbQ.1
Meeting ID: 945 6712 4781
Passcode: 470970