Cognitive and computational Neuroscience in development psychiatry

We investigate how young individuals learn and make decisions and how this is linked to the function and structure of their brains. On the one hand, we aim to understand the motivation underlying approach and avoidance behaviors and how they can lead to rigid and psychopathological behavioral patterns. On the other hand, we are interested in understanding flexible and goal-directed cognitive control that enables behavioral change.

Background

Childhood and Adolescence are characterized by enormous developmental changes, which initialize learning processes on psychological, biological and social levels. These learning processes shape their cognition, motivation and decision-making. This leads to the emergence behavioral patterns, which can flexibly adjusted in a goal-directed manner. However, if development goes awry, aberrant learning may lead to loss-of-control behavioral patterns such as repeated impulsive behaviors. This may show in hyperactivity and cognitive problems but also anger or substance abuse. ADHD is a prominent example of a disorder in developmental psychiatry.

Research objective

Our research aims to identify neurocognitive developmental trajectories to understand the transdiagnostic emergence of impulsive and compulsive symptoms. Exemplary diagnostic entities in developmental psychiatry are ADHD, substance abuse, excessive and restrictive eating disorders and anxious-compulsive disorders. We are particularly interested in dissociating common and distinct neurocognitive processes underlying these symptom dimensions. Empirical and theoretical insights on the role of the neuromodulator dopamine in reinforcement learning play a central role in our research.

Procedures

Our research methods range from questionnaires, behavioral experiments (in the lab, online and on the smartphone) to neural measures (functional und structural MRI, EEG, PET) as well as pharmacological manipulations. We use computational models of learning and decision-making (‘reinforcement learning’) for a detailed understanding of neurocognitive processes and to inform the analysis of neural data (‘computational neuroimaging’). These methods are part of the larger agenda of ‘Computational Psychiatry’.

Relevance

Equipped with these methodological tools, we strive to test the clinical applicability of cognitive and computational neuroscience methods, for example by trying to understand the heterogeneity in treatment responses of our patients to established psychopharmacological agents.