Identifizierung verschiedener Verlaufsformen der Posttraumatischen Belastungsstörung durch maschinelles Lernen mit hochdimensionalen kognitiven, emotionalen und biologischen longitudinalen Daten
Allgemeine, Kognitive und Mathematische Psychologie
Biologische Psychiatrie
Zusammenfassung der Projektergebnisse
In this project we first reviewed the relevant literature on how machine learning was previously used for examining stress pathologies and resilience after traumatic events. We also discussed the challenges of examining PTSD and other stress pathologies and how machine learning can be useful to address these challenges. Furthermore, we summarized the current knowledge about the biological underpinnings and the etiology of PTSD with the aim to integrate the findings of previous studies in a unified theoretical account. There is a vast number of studies that report alterations in neuroendocrine and neurochemical systems in patients with PTSD. Based on these findings, a novel pathophysiologydriven strategy was developed in this project to predict the risk of PTSD symptoms and to forecast the course of symptom development leveraging on a vast and diverse set of biological, immune, behavioral and cognitive markers. In an experimental trauma film paradigm, we showed that a heighted biological stress response during the encoding and consolidation of the experience of a trauma analogue was associated with more intrusive memories - a hallmark of PTSD. In addition, we examined whether inflammatory markers collected in the Emergency Department early after the experience of a traumatic event can be used to develop a prospective biomarker for PTSD. We showed that the proinflammatory immune response to trauma exposure is a relevant predictor for long-lasting PTSD symptoms following trauma. In addition, we examined whether we can develop a noninvasive mobile device to use a physiological marker of autonomic nervous system activation by measuring skin conductance response. Finally, the utilization of Latent Growth Mixture Modelling and machine learning on data from large prospective longitudinal study cohorts has led to the development of predictive models for PTSD risk. We identified early risk factors in trauma survivors using routinely collectable data in the ED and built an algorithm that is implementable in medical systems. In a further naturalistic cohort study, we identified early markers of non-response to routine clinical care in a day-clinic in Zürich using routinely collectable data and we were able to accurately forecast therapy response for depression. In addition, we identified risk factors in high risk populations such as veterans of the US military deployed to Afghanistan and United Nations workforce from around the world. This research is ongoing.
Projektbezogene Publikationen (Auswahl)
- (2018). Neurobiological pathways involved in fear, stress, and PTSD. In C. B. Nemeroff & C. R. Marmar (Eds.), Post-traumatic stress disorder (pp. 331–352). Oxford, United Kingdom: Oxford University Press. ISBN: 9780190259440
Heim, C., Schultebraucks, K., Marmar, C. R., & Nemeroff, C. B.
- (2019). Association of Prospective Risk for Chronic PTSD Symptoms With Low TNFα and IFNγ Concentrations in the Immediate Aftermath of Trauma Exposure. American Journal of Psychiatry
Michopoulos, V., Beurel, E., Gould, F., Dhabhar, F. S., Schultebraucks, K., Galatzer-Levy, I. & Nemeroff, C. B.
(Siehe online unter https://doi.org/10.1176/appi.ajp.2019.19010039) - (2019). Heightened biological stress response during exposure to a trauma film predicts an increase in intrusive memories. Journal of abnormal psychology, 128(7), 645
Schultebraucks, K., Rombold-Bruehl, F., Wingenfeld, K., Hellmann-Regen, J., Otte, C., & Roepke, S.
(Siehe online unter https://doi.org/10.1037/abn0000440) - (2019). Increased Skin Conductance Response in the Immediate Aftermath of Trauma Predicts PTSD Risk. Chronic Stress, 3
Hinrichs, R., van Rooij, S. J., Michopoulos, V., Schultebraucks, K., Winters, S., Maples-Keller, J. & Ressler, K. J.
(Siehe online unter https://doi.org/10.1177/2470547019844441) - (2019). Machine Learning for Prediction of Posttraumatic Stress and Resilience Following Trauma: An Overview of Basic Concepts and Recent Advances. Journal of traumatic stress, 32(2), 215-225
Schultebraucks, K., & Galatzer‐Levy, I. R.
(Siehe online unter https://doi.org/10.1002/jts.22384) - (2019). Suicidal imagery in borderline personality disorder and major depressive disorder. Journal of personality disorders, 1-19
Schultebraucks, K., Duesenberg, M., Simplicio, M. D., Holmes, E. A., & Roepke, S.
(Siehe online unter https://doi.org/10.1521/pedi_2019_33_406)
