Computational psychiatry
While psychiatric disorders are very common, no objective diagnostic measures exist and predicting for whom which treatment will work is difficult. We are part of a movement to address this (computational psychiatry) by using computational modelling to understand and obtain objective measures of the cognitive and neural bases of psychiatric disorders.
Cognitive mechanisms of bipolar disorder (Scholl et al., preprint )
Bipolar disorder is a psychiatric disorder characterized by strong mood fluctuations. Here, combining longitudinal data, patients and drug manipulation, we capture the computational and neural correlates of (risk for) bipolar disorder. In healthy participants, when making decisions between different gambles (‘wheel of fortunes’), brain areas in prefrontal cortex (ventromedial prefrontal cortex, medial frontal pole) maintain a memory of rewards won or lost in the past. We found this memory to be linked to the ability to adapt ones behaviour over time to wins and losses. In contrast, severity of bipolar disorder symptoms reduced both computational and corresponding brain markers.
Naturalistic tasks for computational psychiatry (Scholl and Klein-Flügge, 2018)
In this review, we focus on more naturalistic tasks capturing decision-making and learning experimentally and how they can be used to understand psychiatric disorders.
