Project Group Mathematical Learning in Psychiatry

David Popovic

A schematic of the different themes in the research group, including methods, multimodal data and patterns. — Multimodal analysis workflow

Multimodal analysis workflow

The Mathematical Learning in Psychiatry (MLP) Lab develops mathematical and computational tools to better understand severe mental disorders such as psychosis and depression across their full spectrum—from early vulnerability to chronic stages and forensic settings. We combine methods from statistics, machine learning, and systems neuroscience to disentangle the biological and clinical heterogeneity of mental illness and to derive clinically meaningful, interpretable biomarkers.

As an interdisciplinary team, we integrate clinical assessments, MRI, EEG, genetics, blood biomarkers, and eye tracking data to capture distributed patterns that single modality approaches often miss. Our work aims both to advance mathematical methodology and to improve diagnosis, prognosis, and individualized treatment planning in psychiatry.

The image shows a schematic of the workflow of the research group: The steps include data preparation and alignment, modelling and pre-training, training and evaluation as well as cross-modality insights. — Multimodal data workflow

Multimodal data workflow

All of our tools are open source and designed to be as user friendly as possible. We welcome collaborations with researchers and clinicians interested in data-driven, mathematical approaches to mental health and forensic psychiatry.