skip to main content
10.1145/3674029.3674051acmotherconferencesArticle/Chapter ViewAbstractPublication PagesicmltConference Proceedingsconference-collections
research-article

Explaining Predicted Stress Levels in employed Individuals

Published: 11 September 2024 Publication History

Abstract

When emotional, physical, or mental challenges go beyond manageability, stress can occur and long-term exposure to stress leads to negative health consequences. Stress index, a threshold value to detect stress levels, can be linked with sensor data to get valuable insights for healthcare. Research shows that deep learning models can be applied to select heart rate variability features that are deemed important for the prediction of stress, thereby create stress index. However, there is a limitation in existing literature regarding its explainability. In this study, we used LIME and CNN on an ECG dataset called SWELL-KW to explain the predicted stress at job. We applied multiclass classification using a dependent variable ’condition’ having states ’time pressure’ and ’interruption’ as stressor activities and ’no stress’ as a non-stressor activity. Accordingly, we found out that, out of the 34 features of the SWELL-KW dataset, only 5 (namely, HR, PNN25, PNN50, Mean_RR, SDRR_REL_RR) have contributed to the resulting model, positively and negatively in all the three classes.

References

[1]
Saskia Koldijk, Maya Sappelli, Suzan Verberne, Mark A. Neerincx, and Wessel Kraaij. 2014. The SWELL Knowledge Work Dataset for Stress and User Modeling Research. In Proceedings of the 16th International Conference on Multimodal Interaction (ICMI '14). Association for Computing Machinery, New York, NY, USA, 291–298. https://doi.org/10.1145/2663204.2663257.
[2]
Jorn Bakker, Leszek Holenderski, Rafal Kocielnik, Mykola Pechenizkiy, and Natalia Sidorova. 2012. Stess@Work: from measuring stress to its understanding, prediction and handling with personalized coaching. In Proceedings of the 2nd ACM SIGHIT International Health Informatics Symposium (IHI '12). Association for Computing Machinery, New York, NY, USA, 673–678. https://doi.org/10.1145/2110363.2110439.
[3]
Abdullah Alharbi, Wael Alosaimi, Radhya Sahal, and Hager Saleh. 2021. Real-Time System Prediction for Heart Rate Using Deep Learning and Stream Processing Platforms. 2021, (February 2021).
[4]
Hoang ChuDuc, Kien NguyenPhan, and Dung NguyenViet. 2013. A Review of Heart Rate Variability and its Applications. 7, (July 2013), 80–85.
[5]
Marek Malik. 1996. Heart rate variability: Standards of measurement, physiological interpretation, and clinical use. 93, (March 1996), 1043–1065.
[6]
Pooja Prajod and Elisabeth André. 2022. On the Generalizability of ECG-based Stress Detection Models. (Dec. 2022), 549–554.
[7]
Leonidas Liakopoulos, Nikolaos Stagakis, Evangelia I. Zacharaki, and Konstantinos Moustakas. 2021. CNN-based stress and emotion recognition in ambulatory settings. In 2021 12th International Conference on Information, Intelligence, Systems & Applications (IISA).
[8]
Huiwen Loh, Chui Ooi, Silvia Seoni, Prabal Datta Barua, Filippo Molinari, and U Rajendra Acharya. 2022. Application of Explainable Artificial Intelligence for Healthcare: A Systematic Review of the Last Decade (2011–2022). 226, (September 2022).
[9]
Tianhua Chen, Elpida Keravnou-Papailiou, and Grigoris Antoniou. 2021. Medical analytics for healthcare intelligence–Recent advances and future directions. 112, (July 2021), 1-5.
[10]
Inês Neves, Duarte Folgado, Sara Santos, Marilia Barandas, Andrea Campagner, Luca Ronzio, Federico Cabitza, and Hugo Gamboa. 2021. Interpretable Heartbeat Classification using Local Model-Agnostic Explanations on ECGs. 133, (April 2021), 104393.
[11]
Marco Ribeiro, Sameer Singh, and Carlos Guestrin. 2016. “Why Should I Trust You?”: Explaining the Predictions of Any Classifier.
[12]
Andreas Holzinger, Anna Saranti, Christoph Molnar, Przemyslaw Biecek, and Wojciech Samek. 2022. Explainable AI Methods - A Brief Overview.
[13]
Talal Abdullah, M. Zahid, Waleed Ali, and Shahab Ul Hassan. 2023. B-LIME: An Improvement of LIME for Interpretable Deep Learning Classification of Cardiac Arrhythmia from ECG Signals. 11, (February 2023), 595.
[14]
Giorgio Visani, Enrico Bagli, Federico Chesani, Alessandro Poluzzi, and Davide Capuzzo. 2021. Statistical stability indices for LIME: Obtaining reliable explanations for machine learning models. 73, (February 2021), 1–11.
[15]
J. Hughes, Jeffrey Olgin, Robert Avram, Sean Abreau, Taylor Sittler, Kaahan Radia, Henry Hsia, Tomos Walters, Byron Lee, Joseph Gonzalez, and Geoffrey Tison. 2021. Performance of a Convolutional Neural Network and Explainability Technique for 12-Lead Electrocardiogram Interpretation. 6, (August 2021).
[16]
What is an electrocardiogram (ECG. Retrieved from https://www.ncbi.nlm.nih.gov/books/NBK536878/.
[17]
S. Sriramprakash, Prasanna Vadana, and O.V. Murthy. 2017. Stress Detection in Working People. 115, (December 2017), 359–366.
[18]
Markus Plass, Michaela Kargl, Tim-Rasmus Kiehl, Peter Regitnig, Christian Geißler, Theodore Evans, Norman Zerbe, Rita Carvalho, Andreas Holzinger, and Heimo Müller. 2023. Explainability and causability in digital pathology. 9, (April 2023).
[19]
Theodore Evans, Carl Orge Retzlaff, Christian Geißler, Michaela Kargl, Markus Plass, Heimo Müller, Tim-Rasmus Kiehl, Norman Zerbe, and Andreas Holzinger. 2022. The explainability paradox: Challenges for xAI in digital pathology. 133, (July 2022), 281–296.
[20]
Jon Andreas Mortensen, Martin Efremov Mollov, Ayan Chatterjee, Debasish Ghose, and Frank Y. Li. 2023. Multi-Class Stress Detection Through Heart Rate Variability: A Deep Neural Network Based Study. 11, (July 2023), 57470–57480.
[21]
Bosun Hwang, Jiwoo You, Thomas Vaessen, Inez Myin-Germeys, Cheolsoo Park, and Byoung-Tak Zhang. 2018. Deep ECGNet: An Optimal Deep Learning Framework for Monitoring Mental Stress Using Ultra Short-Term ECG Signals. 24, (February 2018).
[22]
Jürg Schläpfer and Hein Wellens. 2017. Computer-Interpreted Electrocardiograms. 70, (August 2017), 1183–1192.
[23]
Matthew Oyeleye, Tianhua Chen, Sofya Titarenko, and Grigoris Antoniou. 2022. A Predictive Analysis of Heart Rates Using Machine Learning Techniques. 19, (February 2022), 2417.
[24]
[1]Bin Wang, Wenbin Pei, Bing Xue, and Mengjie Zhang. 2022. Explaining Deep Convolutional Neural Networks for Image Classification by Evolving Local Interpretable Model-agnostic Explanations. (November 2022).
[25]
Pramod Bobade and Vani M. 2020. Stress Detection with Machine Learning and Deep Learning using Multimodal Physiological Data.
[26]
Maciej Dzieżyc, Martin Gjoreski, Przemysław Kazienko, Stanisław Saganowski, and Matjaž Gams. 2020. Can We Ditch Feature Engineering? End-to-End Deep Learning for Affect Recognition from Physiological Sensor Data. 20, 22 (July 2020), 6535.
[27]
Pengfei Zhang, Fenghua Li, Rongjian Zhao, Ruishi Zhou, Lidong Du, Zhan Zhao, Xianxiang Chen, and Zhen Fang. 2021. Real-Time Psychological Stress Detection According to ECG Using Deep Learning. 11, 9 (July 2021), 3838.
[28]
Fredric Shaffer, Rollin McCraty, and Christopher Zerr. 2014. A healthy heart is not a metronome: An integrative review of the heart's anatomy and heart rate variability. 5, (September 2014).
[29]
Chiranjeevi Endukuru, Endukuru Tutor, null Demonstrator, Aiims Bhopal, and Sonali Tripathi. 2016. Medical Science Evaluation of Cardiac Responses to Stress in Healthy Individuals -A Non Invasive Evaluation by Heart Rate Variability and Stroop Test. (August 2016).
[30]
Fredric Shaffer and Jp Ginsberg. 2017. An Overview of Heart Rate Variability Metrics and Norms. 5, (September 2017), 258.
[31]
Radwa El Shawi, Mouaz Al-Mallah, and Sherif Sakr. 2019. On the interpretability of machine learning-based model for predicting hypertension. 19, (July 2019).
[32]
Marco Tulio Correia Ribeiro. 2021. LIME. GitHub repository. Retrieved from https://github.com/marcotcr/lime.
[33]
SAS® Help Center. Retrieved September 1, 2023 from https://documentation.sas.com/doc/en/emref/14.3/n061bzurmej4j3n1jnj8bbjjm1a2.htm.
[34]
Hyun-Myung Cho, Heesu Park, Suh-Yeon Dong, and Inchan Youn. 2019. Ambulatory and Laboratory Stress Detection Based on Raw Electrocardiogram Signals Using a Convolutional Neural Network. 19, (October 2019), 4408.

Recommendations

Comments

Information & Contributors

Information

Published In

cover image ACM Other conferences
ICMLT '24: Proceedings of the 2024 9th International Conference on Machine Learning Technologies
May 2024
336 pages
ISBN:9798400716379
DOI:10.1145/3674029
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected].

Publisher

Association for Computing Machinery

New York, NY, United States

Publication History

Published: 11 September 2024

Permissions

Request permissions for this article.

Check for updates

Author Tags

  1. Electrocardiogram
  2. LIME
  3. XAI
  4. heart rate

Qualifiers

  • Research-article
  • Research
  • Refereed limited

Conference

ICMLT 2024

Contributors

Other Metrics

Bibliometrics & Citations

Bibliometrics

Article Metrics

  • 0
    Total Citations
  • 23
    Total Downloads
  • Downloads (Last 12 months)23
  • Downloads (Last 6 weeks)5
Reflects downloads up to 13 Feb 2025

Other Metrics

Citations

View Options

Login options

View options

PDF

View or Download as a PDF file.

PDF

eReader

View online with eReader.

eReader

HTML Format

View this article in HTML Format.

HTML Format

Figures

Tables

Media

Share

Share

Share this Publication link

Share on social media