Limits...
Towards Measuring Stress with Smartphones and Wearable Devices During Workday and Sleep.

Muaremi A, Arnrich B, Tröster G - Bionanoscience (2013)

Bottom Line: But, when the person constantly experiences high workload over a longer period of time and is not able to recover, then work may lead to prolonged negative effects and might cause serious illnesses like chronic stress disease.Using only smartphone features, we get an accuracy of 55 %, and using only heart rate variability features, we get an accuracy of 59 %.The combination of all features leads to a rate of 61 % for a three-stress level (low, moderate, and high perceived stress) classification problem.

View Article: PubMed Central - PubMed

Affiliation: Wearable Computing Lab, ETH Zurich, Gloriastrasse 35, 8092 Zurich, Switzerland.

ABSTRACT

Work should be a source of health, pride, and happiness, in the sense of enhancing motivation and strengthening personal development. Healthy and motivated employees perform better and remain loyal to the company for a longer time. But, when the person constantly experiences high workload over a longer period of time and is not able to recover, then work may lead to prolonged negative effects and might cause serious illnesses like chronic stress disease. In this work, we present a solution for assessing the stress experience of people, using features derived from smartphones and wearable chest belts. In particular, we use information from audio, physical activity, and communication data collected during workday and heart rate variability data collected at night during sleep to build multinomial logistic regression models. We evaluate our system in a real work environment and in daily-routine scenarios of 35 employees over a period of 4 months and apply the leave-one-day-out cross-validation method for each user individually to estimate the prediction accuracy. Using only smartphone features, we get an accuracy of 55 %, and using only heart rate variability features, we get an accuracy of 59 %. The combination of all features leads to a rate of 61 % for a three-stress level (low, moderate, and high perceived stress) classification problem.

No MeSH data available.


Related in: MedlinePlus

An example of the visualization of DSP for the class probabilities  =
© Copyright Policy - OpenAccess
Related In: Results  -  Collection


getmorefigures.php?uid=PMC4269214&req=5

Fig5: An example of the visualization of DSP for the class probabilities =

Mentions: The daily stress score DS is a continuous value between 0 and 1 and reflects the stress level of the previous day. This score can also be seen as the acute stress level of a person. DSP and DSH are the individual scores using smartphone data and HRV data. They are computed as 5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \text{DS}_{\text{K}} = [p_{0\text{K}}, p_{1\text{K}}, p_{2\text{K}}] \cdot [0, 0.5, 1]^{T}, $$\end{document}. Figure 5 shows an example of the visualization of DSP for the class probabilities . If training data from both modalities are available at the same time, then a common logit model is trained with features from both smartphone and HRV, and DS is computed as6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \text{DS}_{\text{K}} = [p_{0}, p_{1}, p_{2}] \cdot [0, 0.5, 1]^{T}, $$\end{document}where are the outcome probabilities of the common model with the input . However, in a practical case, a common trained model is not available, if for daily training, data from one modality are missing. In that case, DS is computed using DSP and DSH as7\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \text{DS} = w_{\text{P}} \text{DS}_{\text{P}} + w_{\text{H}} \text{DS}_{\text{H}}. $$\end{document} and are the a priori weights which correspond to the normalized classification accuracies of DSP and DSH.Fig. 5


Towards Measuring Stress with Smartphones and Wearable Devices During Workday and Sleep.

Muaremi A, Arnrich B, Tröster G - Bionanoscience (2013)

An example of the visualization of DSP for the class probabilities  =
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC4269214&req=5

Fig5: An example of the visualization of DSP for the class probabilities =
Mentions: The daily stress score DS is a continuous value between 0 and 1 and reflects the stress level of the previous day. This score can also be seen as the acute stress level of a person. DSP and DSH are the individual scores using smartphone data and HRV data. They are computed as 5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \text{DS}_{\text{K}} = [p_{0\text{K}}, p_{1\text{K}}, p_{2\text{K}}] \cdot [0, 0.5, 1]^{T}, $$\end{document}. Figure 5 shows an example of the visualization of DSP for the class probabilities . If training data from both modalities are available at the same time, then a common logit model is trained with features from both smartphone and HRV, and DS is computed as6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \text{DS}_{\text{K}} = [p_{0}, p_{1}, p_{2}] \cdot [0, 0.5, 1]^{T}, $$\end{document}where are the outcome probabilities of the common model with the input . However, in a practical case, a common trained model is not available, if for daily training, data from one modality are missing. In that case, DS is computed using DSP and DSH as7\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \text{DS} = w_{\text{P}} \text{DS}_{\text{P}} + w_{\text{H}} \text{DS}_{\text{H}}. $$\end{document} and are the a priori weights which correspond to the normalized classification accuracies of DSP and DSH.Fig. 5

Bottom Line: But, when the person constantly experiences high workload over a longer period of time and is not able to recover, then work may lead to prolonged negative effects and might cause serious illnesses like chronic stress disease.Using only smartphone features, we get an accuracy of 55 %, and using only heart rate variability features, we get an accuracy of 59 %.The combination of all features leads to a rate of 61 % for a three-stress level (low, moderate, and high perceived stress) classification problem.

View Article: PubMed Central - PubMed

Affiliation: Wearable Computing Lab, ETH Zurich, Gloriastrasse 35, 8092 Zurich, Switzerland.

ABSTRACT

Work should be a source of health, pride, and happiness, in the sense of enhancing motivation and strengthening personal development. Healthy and motivated employees perform better and remain loyal to the company for a longer time. But, when the person constantly experiences high workload over a longer period of time and is not able to recover, then work may lead to prolonged negative effects and might cause serious illnesses like chronic stress disease. In this work, we present a solution for assessing the stress experience of people, using features derived from smartphones and wearable chest belts. In particular, we use information from audio, physical activity, and communication data collected during workday and heart rate variability data collected at night during sleep to build multinomial logistic regression models. We evaluate our system in a real work environment and in daily-routine scenarios of 35 employees over a period of 4 months and apply the leave-one-day-out cross-validation method for each user individually to estimate the prediction accuracy. Using only smartphone features, we get an accuracy of 55 %, and using only heart rate variability features, we get an accuracy of 59 %. The combination of all features leads to a rate of 61 % for a three-stress level (low, moderate, and high perceived stress) classification problem.

No MeSH data available.


Related in: MedlinePlus