Research

WHAT WE STUDY

Understanding Mental Illness: Computational Psychiatry

Translating advances in neuroscience into benefits for patients with mental illness presents enormous challenges because it involves both the most complex organ, the brain, and its interaction with a similarly complex environment. One promising approach to this challenge is computational psychiatry. This nascent field integrates diverse computational methods with multifaceted data to enhance our understanding, prediction, and treatment of mental illness (Petzschner et al, 2017).

In the lab, we have developed experiments and complex models to capture various aspects of cognition, spanning perception, learning, and action selection in clinical populations. Specifically, we study the origins of perceptual biases in Autism Spectrum Disorder (Schneebeli et al., 2022), and the role of action selection and beliefs about agency for the development of compulsions in Disordered Gambling and Obsessive-Compulsive Disorder (Rigoux et al., 2024; Conceição*, Petzschner* et al., 2023). Our research also examines how these cognitive processes interact with pharmacological and psychological therapies, offering insights that could lead to more effective treatment strategies (Conceição*, Petzschner* et al., 2023).

Key PubliCATIONS

Petzschner F.H., Weber L.A.E., Gard T., Stephan K.E. Computational Psychosomatics and Computational Psychiatry: Towards a joint framework for differential diagnosis. Biological Psychiatry 82(6):421-430, 2017

Rigoux, L, Stephan K.E., and Petzschner F.H. Beliefs, Compulsive Behavior and Reduced Confidence in Control. PLoS Comput Biol 20(6): e1012207. https://doi.org/10.1371/journal.pcbi.1012207, 2024

Conceição V.A.*, Petzschner F.H.*, Cole D.M., Wellstein K.V. , Müller D., Raman S., Maia T.V.: Serotonin Reduces Belief Stickiness. bioRxiv, 2023

Schneebli M., Haker H., Rüesch A., Zahnd N., Marino S., Paolini G., Petzschner F.H.*, Stephan K.E.*. Disentangling «Bayesian brain» theories of Autsim Spectrum Disorder. medRxiv XX(X):XX-XX, 2022

RESOURCES

Computational Psychiatry Code Toolbox

Computational Psychosomatics: Understanding Somatic Symptoms and Affect

Our research in Computational Psychiatry has led to a crucial insight that is now central to the research agenda of the lab: many symptoms observed in psychiatric disorders—particularly those related to somatic reactions, emotions, mood, and self-perception—can only be effectively understood and predicted if our models incorporate how we perceive and regulate our body. To address this, we have expanded Computational Psychiatry frameworks to include body perception, culminating in our pioneering paper on Computational Psychosomatics published in 2017 (Petzschner et al., 2017). This paper introduces a comprehensive systems theory framework that integrates brain-body interactions, encompassing interoception, exteroception, and both homeostatic and allostatic regulation. Notably, it proposes a novel computational approach for using model-based indices to facilitate formal differential diagnosis of psychosomatic symptoms.

Building on this framework, we have developed detailed mechanisms that may underlie chronic somatic symptoms, such as Chronic Fatigue (Stephan et al., 2016), and Chronic Pain, and their impact on affective states, particularly Anhedonia in Depression (Stephan et al., 2016). Recently, we have begun testing predictions derived from these models in large-scale clinical studies involving large sample sizes in clinical patients over extended periods, employing a combination of mechanistic modeling and machine learning techniques (e.g. Müller-Schrader M., 2022).

Key PubliCATIONS

Khalsa S.S., Adolphs R., Cameron O.G., Critchley H.D., Devenport P.W., Feinstein J.S., Feusner J.D., Garfinkel S.N., Lane R.D., Mehling W.E., Meuret A.E. Nemeroff C.B., Oppenheimer S., Petzschner F.H., Pollatos O., Rhudy J.L., Schramm L.P., Simmons W.K., Stein M.B., Stephan K.E., Van Den Bergh O., Van Diest I., von Leupoldt A., Paulus M.P. and the Interoception Summit 2016 participants: Interoception and Mental Health: a Roadmap. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging 3(6): 501-513, 2018

Stephan K.E., Manjaly Z.M., Mathys C.D., Weber L.A.E., Paliwal S., Gard T., Tittgemeyer M., Fleming S.M., Haker H., Seth A.K., Petzschner F.H. Allostatic self-efficacy: a metacognitive theory of dyshomeostasis-induces fatigue and depression. Frontiers in Human Neuroscience 10:550, 2016

Müller-Schrader M., Heinzle J., Müller A., Lanz C., Häussler O., Suter M., Eggspühler A., Mare S., Toussaint B., Pereira I., Petzschner F.H., Wiech K., Barth J., Witt C.M., Stephan K.E., Manjaly Z.-M. Individual treatment expectations predict clinical outcome after lumbar injections against low back pain. Pain 164 (1), 132-141, 2023

Towards Precision Psychiatry: Real-World Data Sets and Large-Scale Modeling

While our theoretical concepts offer strong predictions about the mechanisms underlying mental disorders, there are few rigorous attempts to empirically test these predictions and validate them in clinical settings. To address this gap, we need a dual approach: integrating deep data—detailed assessments of individual, computationally-informed markers—with wide data—comprehensive coverage of a diverse population beyond specific disease labels, and utilizing rigorously validated machine learning models (Petzschner, Science, 2024).

In response, our lab has developed an award-winning digital health tool, SOMA. SOMA is a smartphone app designed for the longitudinal tracking of symptoms (mood, pain), daily life activities (emotions, activities, expectations), and treatments (medications and interventions) over extended periods (Gunsilius et al., JMIR, 2024). The app generates rich, multidimensional data that significantly enhances our capacity to monitor and analyze patient experiences. Complementing this, we have established the SOMAScience platform, which facilitates large-scale, multi-site data acquisition and clinical trial tracking (Gunsilius et al., JMIR, 2024). To further extend the application of translational computational models in psychiatry, we have also published open-source software packages (Frässle et al., 2021), enabling broader access to these advanced tools and promoting collaborative research in the field.

Key Publications

Petzschner F. H.: Practical Challenges for Precision Medicine. Science 383 (6679): 149-150. PMID: 38207033, 2024

Zimmerman C.S., Heffner J., Bruinsma S., Corinha M., Cortinez M., Dalton H., Duong E., Lu J., Omar A., Owen L.L.W., Roarr B.N., Tang K., Petzschner F.H.SOMAScience: A novel platform for multidimensional, longitudinal pain assessment. JMIR mHealth and uHealth 12 (1), e47177

Frässle S., Aponte E.A., Bollman S., Bordersen K.H., Do C.T., Harrison O.K., Harrison S.J., Heinzle J., Iglesias S., Kasper L., Lomakina E.I., Mathys C., Müller-Schrader M., Pereira I., Petzschner F.H., Raman S., Schöbi D., Toussaint B., Weber L.A., Yao Y., Stephan K.E. TAPAS: an open-source software package for Translational Neuromodeling and Computational Psychiatry. Frontiers in Psychiatry 12:857, 2021

RESOURCES

Computational Psychiatry Code Toolbox

Perception and Psychophysics: Reconciling the laws of psychophysics

For centuries, scientists have been fascinated by the brain’s extraordinary capacity to construct coherent and reliable representations of both the self and the external world from limited and often noisy sensory information. A widely accepted concept is that perception is fundamentally an inference process, where the brain interprets sensory input through its internal models of the world. These models generate predictions about incoming sensory data, which are continuously refined by comparing predictions to actual sensory input via sensory prediction errors. This process has been formalized through mathematical frameworks, including Predictive Coding, Hierarchical Bayesian Modeling, and Active Inference. Our lab works on advancing our understanding of these perceptual processes by developing and applying cross-dimensional computational frameworks.

Weber, L.A., Yee, D. M., Small, D. M., & Petzschner, F. H. Rethinking reinforcement learning: The interoceptive origin of reward. https://doi.org/10.31234/osf.io/be6nv, 2024

Key PubliCATIONS

Petzschner F.H., Glasauer S., Stephan K.E. A Bayesian Perspective on Magnitude Estimation. Trends in Cognitive Sciences 19(5):285-293, 2015

Petzschner F.H., Glasauer S. Iterative Bayesian estimation as an explanation for regression and range effects – a study on human path integration. Journal of Neuroscience 31(47):17220-9, 2011

RESOURCES

Distance Estimation Task

Interoception

The brain evolved to ensure our survival, a process that depends critically on our ability to accurately sense and regulate our physiological state. Interoception—the brain’s capacity to process and interpret internal bodily signals from visceral organs like the heart and lungs—plays a pivotal role in this regulatory function (Petzschner et al., 2022). Unsurprisingly, dysfunctions in interoception have been linked to a wide range of somatic and psychiatric conditions, including depression, anxiety, eating disorders, and chronic fatigue (Khalsa et al., 2018). Traditionally, interoception and exteroception have been studied in isolation due to their distinct sensory receptors and neural pathways. However, we and others have recently extended perceptual frameworks to interoception, demonstrating that interoception and exteroception are deeply intertwined computational processes governed by common algorithmic principles (Petzschner et al., 2021). This integrated framework offers powerful explanations for how bodily signals shape emotions, cognition, and the sense of self.

Key Publications

Weber, L.A., Yee, D. M., Small, D. M., & Petzschner, F. H. Rethinking reinforcement learning: The interoceptive origin of reward. https://doi.org/10.31234/osf.io/be6nv, 2024

Petzschner F.H., Critchley H., Tallon-Baudry C., Interoception, Scholarpedia, 17(5):55569, 2022

Petzschner F.H.*, Garfinkel S.N., Paulus M.P., Koch C., Khalsa S.S. Computational Models of Interoception and Body regulation. Trends in Neurosciences 44(1):63-76, 2021

Petzschner F.H.*, Weber L.*, Wellstein K.V., Paolini G., Tri Do C, Stephan K.E. Focus of attention modulates the heartbeat evoked potential. NeuroImage 186:595-606, 2020

RESOURCES

Task Code: Heartbeat Attention Task

Data: Heartbeat Attention Task

Decision-Making

Our decision are a window into the computational processes performed by the brain & they can also be a window into consciousness.

Key Publications

Kang Y.*, Petzschner F.H.*, Wolpert D.M., Shadlen M.N. Piercing of consciousness as a threshold-crossing operation. Current Biology 27:2285-2295, 2017

RESOURCES

Data: Random Dots Task for “Piercing of consciousness paper”

Code: for “Piercing of consciousness paper”