Research

Our current research portfolio spans three interconnected domains. Building on our previous work on PTSD symptom distributions, which revealed that while only few symptom combinations are common, a large part of individuals display a unique symptoms combination, we are now investigating the fundamental building blocks of PTSD diagnosis. This research aims to identify the most diagnostically relevant symptoms and optimize diagnostic criteria, potentially informing future DSM revisions to reduce complexity while maintaining clinical utility.

In parallel, we conduct pragmatic clinical research in collaboration with clinician-scientists at the Psychiatric University Hospital Zurich (PUK). These projects focus on evaluating adherence to clinical guidelines and identifying factors that influence treatment outcomes in real-world settings, bridging the gap between research findings and clinical practice. Our newest research stream explores the role of Large Language Models (LLMs) in psychiatric research and clinical practice.

We are investigating both the potential and limitations of these tools, from their application in research methodology to their potential role in clinical documentation and assessment. This work includes examining how LLMs process and express psychological concepts, their potential biases, and their practical applications in mental health research and care delivery, while maintaining a critical perspective on their limitations and ethical implications.