Hauptinhalt

Under the umbrella of “Network Dynamics”, we investigate how large-scale brain networks support cognition, emotion, and behaviour, and how their dynamic properties are altered in psychiatric disorders. Our group uses structural and functional MRI, combined with network neuroscience and control theory, to characterize the organization and controllability of brain networks in health and disease. Rather than focusing on isolated regions, this work examines how distributed circuits interact over time to generate complex mental states.

A central idea is that psychiatric symptoms may reflect changes in the way brain networks flexibly transition between different configurations, rather than static abnormalities in single areas. To capture this, we apply graph-theoretical metrics, connectivity analyses, and emerging approaches from network control theory. These methods quantify how strongly individual regions contribute to driving the brain from one functional state to another, as well as how global network architecture constrains possible trajectories of brain activity. 

In schizophrenia and related psychosis-spectrum conditions, we investigate whether network-level properties—such as integration, segregation, and hubness—are systematically altered. For example, specific fronto-parietal and limbic regions may show changes in their ability to influence broader network states, which could relate to difficulties in cognitive control, self-monitoring, or emotional regulation. By linking these network measures to dimensional symptom profiles and schizotypy scores, the research aims to identify dynamic correlates of psychosis liability and resilience.

Developmental aspects are also a focus, particularly how white matter maturation and structural connectivity shape the evolution of network controllability across adolescence and early adulthood. This is a critical period for the emergence of many psychiatric disorders, and understanding normative trajectories of network dynamics can help to detect deviations associated with increased risk. Large-scale datasets, often in collaboration with international consortia, enable the group to study these processes with sufficient statistical power.​

Methodologically, our group combines advanced preprocessing pipelines with robust statistical modeling to ensure reproducible network metrics. We integrate multimodal data, such as structural connectivity, functional connectivity, and behavioural measures, to provide converging evidence for network-based hypotheses. In some projects, dynamic functional connectivity and time-resolved analyses are used to characterize moment-to-moment fluctuations in network organization during rest and task states.