Our research

Our research group emerged from the department of developmental psychopathology at the Psychiatric University Hospital in Basel. Our goal is to improve clinical practice and theoretical knowledge in this field, by developing innovative computational methods. Faced with the complexity of human interaction in psychotherapy, we started to apply more advanced methods from computational science to our field.

Innovative data sources: We study novel signals in psychiatry, e.g. via machine learning emotion recognition from video analysis, or vocal arousal indices from audio signals in psychotherapy. This also gives us the opportunity to study the absence of these signals, e.g. silence in psychotherapy. We also study "Big data" sources, e.g. lexical corpora or search engine query data for psychiatric epidemiology.

Software systems for psychiatric documentation: We develop MHIRA, an open source psychiatric documentation system, particularly focused on low- and middle-income countries. We aim at making data collection globally more accessible and more sustainable.

Psychiatric treatment and psychotherapy: Our focus is on, but not only, psychiatric and psychotherapeutic treatment.

Modern computational methods in psychology: Considering the complexity of the signals we use, we apply modern computational methods, e.g. machine learning algorithms. However, a method is never a goal in itself. Our philosophy is to aim for the most understandable, yet comprehensive data analysis possible.

Recent publications

A swarm intelligence machine learning algorithm for test development Silence in Psychotherapy Supervised speaker diarization using Random Forests
In this study we used the "intelligence" of a swarm of ants to develop a short questionnaire to detect personality disorders in adolescence. Check out our study on Plos One.
Usually, silence is an important thing in psychotherapy. However, in this study we studied the association of silence in psychotherapies of adolescent patients. Find out more about this study below. Who speaks when? This information is key, when you want to find out about the flow of a conversation, also in psychotherapy. This study by Fürer et al. introduces a methodology to diarize the audio signal for two different speakers.
Steppan / Zimmerann, et al. Zimmermann, et al. Fürer, et al.
Plos One Personality Disorders: Theory, Research and Treatment Frontiers in Psychology