Luca Abel

Luca Abel

Master's Thesis

Machine Learning-Based Detection of Acute Psychosocial Stress from Dynamic Movements

Robert Richer (M.Sc.), Arne Küderle (M.Sc), Prof. Dr. Nicolas Rohleder (Chair of Health Psychology), Prof. Dr. Björn Eskofier

12 / 2021 – 06 / 2022

Stress is universally present in all our daily lives. Although stress is a healthy reaction of our body, extensive stress can severely affect health and mental wellbeing [1]. Therefore, stress is nowadays recognized as a leading cause for long term sickness in many countries around the globe [2], [3]. Various neuroendocrine reactions happen as a response to stress. The two main stress pathways are the sympathetic nervous system (SNS) and the hypothalamic-pituitary-adrenocortical (HPA) axis, which, among other functions, are responsible for the secretion of the well-established stress markers: alpha-amylase and cortisol [4].

Although stress and its consequences are relatively well studied, the traditional approach of measuring neuroendocrine and electrophysiological markers requires a great deal of effort on the researchers’ side. Furthermore, these traditional assessment techniques are mostly invasive, especially those regarding the inflammatory response, which can only be measured reliably via blood samples [5]. Even when noninvasive methods are used, they require at least some interference with natural human behavior [6].

A step toward noninvasive, possibly even contactless, measurement of stress can be made by observing body movements and posture. A defensive “freezing” behavior due to acute stress has already been shown in multiple studies [7], [8]. Dynamic movements, such as walking, might have the potential to yield even more information about motion pattern alterations which would not be observable in static poses. Previous work has shown that several mental and physical illnesses can be linked to distinct variations of gait and body posture. For instance, patients suffering from depression had a slower gait and slumped posture [8]. Lasselin et. al. showed that acute systemic inflammation, which is also induced as a response to acute stress, can be connected to gait and posture alterations [10]. Hence, it is expected that similar gait and postural changes can be observed following a stressful event.

The goal of this master’s thesis is therefore to extend the knowledge about how acute stress can be unobtrusively measured from body posture and movements. To achieve that, a randomized cross-over study will be conducted as a part of the EmpkinS collaborative research center [11]. As part of this study, the participants will perform the Trier Social Stress Test (TSST) [12] and the friendly version (fTSST) [13] in randomized order on two consecutive days. In addition to the traditional stress markers, IMU-based motion capture data will be recorded during the stress exposure as well as during standardized gait and mobility tests which will be performed before and after the stress task. The existing analysis pipeline from a pre-study will be extended to additionally extract information from dynamic movement. Further, it will be analyzed how the movement changes during the standardized gait tests and how these changes correlate with biological and psychological stress responses.


Full Thesis


[1] D. B. O’Connor, J. F. Thayer, and K. Vedhara, “Stress and Health: A Review of Psychobiological Processes,” Annu. Rev. Psychol., vol. 72, no. 1, pp. 663–688, 2021, doi: 10.1146/annurev-psych-062520-122331.
[2] APA, “Stress in America 2019,” Am. Psychol. Assoc., 2019.
[3] I. Polanowski, “Work-related stress, anxiety or depression statistics in Great Britain, 2021,” UK Health Saf., 2021.
[4] Y. M. Ulrich-Lai and J. P. Herman, “Neural regulation of endocrine and autonomic stress responses,” Nat. Rev. Neurosci., vol. 10, no. 6, pp. 397–409, Jun. 2009, doi: 10.1038/nrn2647.
[5] D. C. Slavish, J. E. Graham-Engeland, J. M. Smyth, and C. G. Engeland, “Salivary markers of inflammation in response to acute stress,” Brain. Behav. Immun., vol. 44, pp. 253–269, Feb. 2015, doi: 10.1016/j.bbi.2014.08.008.
[6] A. E. Kazdin, “Unobtrusive measures in behavioral assessment,” J. Appl. Behav. Anal., vol. 12, no. 4, pp. 713–724, 1979, doi: 10.1901/jaba.1979.12-713.
[7] K. Roelofs, M. A. Hagenaars, and J. Stins, “Facing Freeze: Social Threat Induces Bodily Freeze in Humans,” Psychol. Sci., vol. 21, no. 11, pp. 1575–1581, Nov. 2010, doi: 10.1177/0956797610384746.
[8] M. A. Hagenaars, K. Roelofs, and J. F. Stins, “Human freezing in response to affective films,” Anxiety Stress Coping, vol. 27, no. 1, pp. 27–37, Jan. 2014, doi: 10.1080/10615806.2013.809420.
[9] R. Feldman, S. Schreiber, and E. Been, “Gait, Balance and Posture in Major Mental Illnesses: Depression, Anxiety and Schizophrenia,” Austin Med Sci, vol. 5, no. 1, p. 1039, 2020.
[10] J. Lasselin et al., “Biological motion during inflammation in humans,” Brain. Behav. Immun., vol. 84, pp. 147–153, Feb. 2020, doi: 10.1016/j.bbi.2019.11.019.
[11] “EmpkinS – Website für den SFB-Antrag Empathokinästhetische Sensorik.” (accessed Jan. 04, 2022).
[12] C. Kirschbaum, K.-M. Pirke, and D. H. Hellhammer, “The ‘Trier Social Stress Test’ – A Tool for Investigating Psychobiological Stress Responses in a Laboratory Setting,” in Neuropsychobiology, 1993, vol. 28, no. 1–2, pp. 76–81. doi: 10.1159/000119004.
[13] U. S. Wiemers, D. Schoofs, and O. T. Wolf, “A friendly version of the Trier Social Stress Test does not activate the HPA axis in healthy men and women,” Stress, vol. 16, no. 2, pp. 254–260, Mar. 2013, doi: 10.3109/10253890.2012.714427.