02/2021 – 08/2021
Psychological stress and its consequences are one of the key social problems in industrialized nations. It not only limits well-being, but also has measurable effects on health . Acute stress induces strong physiological reactions that affect the whole body. As response to stress, the body activates the two main stress pathways, the sympathetic nervous system (SNS), leading to heart rate increase and secretion of alpha-amylase, and the hypothalamic-pituitary-adrenocortical (HPA) axis, leading to the secretion of cortisol .
Such an acute stress reaction is typically assessed by “wet markers”, such as saliva and blood samples, electrophysiological measurements, such as electrocardiogram, as well as via self-reports. These traditional methods are based on complex, often invasive, laboratory procedures, which limit biopsychological research. Hence, novel and non-invasive, potentially even contactless, assessment modalities are required to better understand the consequences of stress on the human body.
Another possibility of capturing an acute stress reaction might be given by the observation of body macro movements. Such movements are often an expression of human emotion and, hence, have the potential to be leveraged for the detection of acute psychosocial stress. Previous work has shown that humans can easily identify negative emotions just from body posture and movements, such as dropping the head, bringing hands to the face or the head, or showing a defensive “freezing behavior” [3, 4]. However, most of previous studies investigated only single movement parameter and not entire body posture and movements. Also, an automatic classification of subjects into stressed/non-stressed conditions based on body posture and movement has not been presented yet
The goal of this master’s thesis is therefore to investigate the influence of acute stress on body posture and movement. For that reason, a pipeline to analyze body posture and motion acquired from IMU-based motion capture data will be implemented. The pipeline should be able to identify and extract relevant features that distinguish stress and non-stress conditions. The extracted features then be used to develop and evaluate machine learning models that are able to automatically classify the two conditions. Furthermore, it will be explored whether body posture and movement can also be used to predict the magnitude of an endocrinological stress response. The IMU-based motion capture data is collected from healthy subjects undergoing the Trier Social Stress Test (TSST)  as stress condition and the friendly version of the Trier Social Stress Test (fTSST)  as non-stress-condition.
 Cohen, D. Janicki-Deverts, und G. E. Miller, „Psychological Stress and Disease“, JAMA, Bd. 298, Nr. 14, S. 1685, Okt. 2007, doi: 10.1001/jama.298.14.1685.
 M. Ulrich-Lai und J. P. Herman, „Neural regulation of endocrine and autonomic stress responses“, Nat Rev Neurosci, Bd. 10, Nr. 6, S. 397–409, Juni 2009, doi: 10.1038/nrn2647.
 P. Atkinson, W. H. Dittrich, A. J. Gemmell, und A. W. Young, „Emotion Perception from Dynamic and Static Body Expressions in Point-Light and Full-Light Displays“, Perception, Bd. 33, Nr. 6, S. 717–746, Juni 2004, doi: 10.1068/p5096.
 Roelofs, M. A. Hagenaars, und J. Stins, „Facing Freeze: Social Threat Induces Bodily Freeze in Humans“, Psychol Sci, Bd. 21, Nr. 11, S. 1575–1581, Nov. 2010, doi: 10.1177/0956797610384746.
 Kirschbaum, K.-M. Pirke, und D. H. Hellhammer, „The ‘Trier Social Stress Test’ – A Tool for Investigating Psychobiological Stress Responses in a Laboratory Setting, Neuropsychobiology, Bd. 28, Nr. 1–2, S. 76–81, 1993, doi: 10.1159/000119004.
 S. Wiemers, D. Schoofs, und O. T. Wolf, „A friendly version of the Trier Social Stress Test does not activate the HPA axis in healthy men and women, null, Bd. 16, Nr. 2, S. 254–260, März 2013, doi: 10.3109/10253890.2012.714427.