Research

The core aim of my research is to advance precision psychiatry by improving the understanding of the development and maintenance of mental disorders, and ultimately to enhance individualized diagnosis and treatment. To achieve this, predictive models and biomarkers are essential, as they can provide person-specific insights. A deep understanding of the interplay between biological, psychological, and environmental networks is crucial for generating mechanistic insights that enable tailored interventions. This goal is structured along six main pillars.

Conceptualizing Mental Disorders as Complex Systems

The search for clinically useful biomarkers requires a paradigm shift in how we conceptualize and model psychiatric disorders. My work has shown that traditional approaches are insufficient to capture the multicausal nature of disorders such as depression. I approach psychiatric conditions and brain networks as dynamic systems. This includes symptom-based models, such as network theory of psychopathology, and neurobiological models grounded in dynamical systems theory and control theory. These frameworks provide novel quantitative models that can evolve through falsification and ultimately improve clinical practice.

Longitudinal Imaging and Behavioral Data

Studying these systems requires comprehensive, multimodal, longitudinal datasets. My work has used resting-state and structural MRI, genetic and environmental data, as well as ecological momentary assessment (EMA) to capture real-time fluctuations in symptom networks. Collecting and integrating such datasets is critical for linking dynamic symptom changes with underlying neurobiology.

Predictive Modeling and Machine Learning

Individual-level prediction depends on advanced predictive modeling and machine learning. I combine data-driven methods (classical machine learning, deep learning, normative modeling) with theory-driven approaches (dynamical systems, control theory, computational psychiatry). This integrative perspective is key to identifying mechanisms and improving diagnostic and therapeutic tools.

Infrastructure and Software

Robust software and computational infrastructure are vital for sustainable progress. I contribute to the development of open-source machine learning frameworks (e.g. PHOTONAI) and tools for brain–behavior analyses. My focus is on creating modular, documented, community-driven resources that enable large-scale neuroimaging research and the development of clinically relevant predictive models.

Interdisciplinary Collaboration

The complexity of psychiatric disorders requires collaboration across psychology, neuroscience, computer science, and clinical research. I leverage my background at the intersection of clinical psychology and informatics to bridge these domains. Interdisciplinary work is crucial for integrating machine learning methods into psychiatric research pipelines.

Open Science

Transparent and reproducible research practices are essential. I am committed to preregistration, sharing of preprints, data, code, and software, to ensure replicability and accelerate scientific progress.