Activity-Aware Mental Stress Detection Using Physiological Sensors

Abstract

Continuous stress monitoring may help users better understand their stress patterns and provide physicians with more reliable data for interventions. Previously, studies on mental stress detection were limited to a laboratory environment where participants generally rested in a sedentary position. However, it is impractical to exclude the effects of physical activity while developing a pervasive stress monitoring application for everyday use. The physiological responses caused by mental stress can be masked by variations due to physical activity. We present an activity-aware mental stress detection scheme. Electrocardiogram (ECG), galvanic skin response (GSR), and accelerometer data were gathered from 20 participants across three activities: sitting, standing, and walking. For each activity, we gathered baseline physiological measurements and measurements while users were subjected to mental stressors. The activity information derived from the accelerometer enabled us to achieve 92.4% accuracy of mental stress classification for 10-fold cross validation and 80.9% accuracy for between-subjects classification.

Wearables and Location Tracking Technologies for Mental-State Sensing in Outdoor Environments

“…These types of deficiencies, typical to most field research, weaken the internal validity of results (Wilhelm and Grossman 2010). Third, and related to the previous drawback, real-world measurements of physiological signals-especially those conducted out of doors-are often fraught with noise and measurement errors, making data interpretation even more demanding (Sun et al 2010;Osborne and Jones 2017). For example, skin conductivity which rises during emotional arousal will also increase as a result of extraneous variables such as high ambient temperature that increases sweating.…”

“… Finally, and of key importance to ecological approaches, ambulatory measurements facilitate the investigation of people's physiological signals during their daily routines in real-life situations, offering greater ecological validity than do lab studies (Wilhelm and Grossman 2010). However, with a few exceptions (see for example: Sun et al 2010;Osborne and Jones 2017;Shoval, Schvimer, andTamir 2018a, 2018b), real-world measurements of physiological signals have by and large been restricted to static postures such as a sedentary driving position (Healey, Seger, and Picard 1999;Healey and Picard 2005) and/or to studies that focus on long-term behavioral trends (i.e., hourly/daily changes) rather than second-by-second physiological reactions (Wilhelm and Grossman 2010, 553). There are two primary reasons for this.…”

“…The second is that the social and physical context, which is essential for interpreting the results (Bakker, Pechenizkiy, and Sidorova 2011;Osborne and Jones 2017), changes frequently when people move. In order to deal with this problem, contextual information needs to be collected continuously (Sun et al 2010), complicating data collection and research design.…”

What to Put on the User

“…Hence and by using factorial design, the conditions relaxed walking (RW), relaxed stationary (RS), MAT walking (MW), and MAT stationary (MS) resulted (see Figure 1). A similar setup of conditions has been used by Sun et al [67].…”

Real-time Tracking of Stress Propagation using Distributed Granger Causality

“…, M − 1. Figure 2(a) depicts the workflow for stress detection and stress-level rating [13,4], which uses vital-sign data acquired from biomedical as well as kinematic sensors. Stage 1 of this workflow is purely composed of data-analysis tasks, whereas Stages 2 and 3 are composed of data-manipulation or decision-making tasks.…”