ReBreathe: A Calibration Protocol that Improves Stress/Relax Classification by Relabeling Deep Breathing Relaxation Exercises

Beena Ahmed, Hira Mujeeb Khan, Jongyong Choi, Ricardo Gutierrez-Osuna

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Training stress-prediction models is challenging due to the difficulty in reliably eliciting stress and relaxation responses in participants. For example, a task intended to elicit a relaxation response (e.g., deep breathing) can have the opposite effect depending on the participant's appraisal of and familiarity with the exercise. Including such instances in a training set undermines the accuracy of the resulting prediction model. This paper presents a technique, ReBreathe, to identify such instances based on respiratory patterns and determine their accurate stress/relax labels. We compared this relabeling approach against two labeling techniques: 1) nominal labels obtained from the experimental protocol and 2) labels obtained from subjective assessments. We then trained generalized estimating equation regression models to predict the resulting stress/relax labels from measures of heart rate variability and electrodermal activity. Training the model using protocol labels achieved a classification rate of 0.53 on participants not included in the training set. Relabeling the exercises based on each participant's subjective ratings increased classification rates but only marginally (0.61). In contrast, relabeling the exercises based on respiratory patterns increased classification rates to 0.88, or a four-fold reduction in error rates. These results illustrate the unreliability of protocol and subjective labels during stress/relax exercises and the potential benefits of ReBreathe.

Original languageEnglish
Article number7164293
Pages (from-to)150-161
Number of pages12
JournalIEEE Transactions on Affective Computing
Volume7
Issue number2
DOIs
Publication statusPublished - 1 Apr 2016

Fingerprint

Labels
Calibration
Labeling
Pattern recognition

Keywords

  • deep breathing
  • electrodermal activity
  • heart rate variability
  • stress prediction
  • training protocols

ASJC Scopus subject areas

  • Software
  • Human-Computer Interaction

Cite this

ReBreathe : A Calibration Protocol that Improves Stress/Relax Classification by Relabeling Deep Breathing Relaxation Exercises. / Ahmed, Beena; Khan, Hira Mujeeb; Choi, Jongyong; Gutierrez-Osuna, Ricardo.

In: IEEE Transactions on Affective Computing, Vol. 7, No. 2, 7164293, 01.04.2016, p. 150-161.

Research output: Contribution to journalArticle

@article{653159045bfc4f7fba11a81cdc19ff4d,
title = "ReBreathe: A Calibration Protocol that Improves Stress/Relax Classification by Relabeling Deep Breathing Relaxation Exercises",
abstract = "Training stress-prediction models is challenging due to the difficulty in reliably eliciting stress and relaxation responses in participants. For example, a task intended to elicit a relaxation response (e.g., deep breathing) can have the opposite effect depending on the participant's appraisal of and familiarity with the exercise. Including such instances in a training set undermines the accuracy of the resulting prediction model. This paper presents a technique, ReBreathe, to identify such instances based on respiratory patterns and determine their accurate stress/relax labels. We compared this relabeling approach against two labeling techniques: 1) nominal labels obtained from the experimental protocol and 2) labels obtained from subjective assessments. We then trained generalized estimating equation regression models to predict the resulting stress/relax labels from measures of heart rate variability and electrodermal activity. Training the model using protocol labels achieved a classification rate of 0.53 on participants not included in the training set. Relabeling the exercises based on each participant's subjective ratings increased classification rates but only marginally (0.61). In contrast, relabeling the exercises based on respiratory patterns increased classification rates to 0.88, or a four-fold reduction in error rates. These results illustrate the unreliability of protocol and subjective labels during stress/relax exercises and the potential benefits of ReBreathe.",
keywords = "deep breathing, electrodermal activity, heart rate variability, stress prediction, training protocols",
author = "Beena Ahmed and Khan, {Hira Mujeeb} and Jongyong Choi and Ricardo Gutierrez-Osuna",
year = "2016",
month = "4",
day = "1",
doi = "10.1109/TAFFC.2015.2459682",
language = "English",
volume = "7",
pages = "150--161",
journal = "IEEE Transactions on Affective Computing",
issn = "1949-3045",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "2",

}

TY - JOUR

T1 - ReBreathe

T2 - A Calibration Protocol that Improves Stress/Relax Classification by Relabeling Deep Breathing Relaxation Exercises

AU - Ahmed, Beena

AU - Khan, Hira Mujeeb

AU - Choi, Jongyong

AU - Gutierrez-Osuna, Ricardo

PY - 2016/4/1

Y1 - 2016/4/1

N2 - Training stress-prediction models is challenging due to the difficulty in reliably eliciting stress and relaxation responses in participants. For example, a task intended to elicit a relaxation response (e.g., deep breathing) can have the opposite effect depending on the participant's appraisal of and familiarity with the exercise. Including such instances in a training set undermines the accuracy of the resulting prediction model. This paper presents a technique, ReBreathe, to identify such instances based on respiratory patterns and determine their accurate stress/relax labels. We compared this relabeling approach against two labeling techniques: 1) nominal labels obtained from the experimental protocol and 2) labels obtained from subjective assessments. We then trained generalized estimating equation regression models to predict the resulting stress/relax labels from measures of heart rate variability and electrodermal activity. Training the model using protocol labels achieved a classification rate of 0.53 on participants not included in the training set. Relabeling the exercises based on each participant's subjective ratings increased classification rates but only marginally (0.61). In contrast, relabeling the exercises based on respiratory patterns increased classification rates to 0.88, or a four-fold reduction in error rates. These results illustrate the unreliability of protocol and subjective labels during stress/relax exercises and the potential benefits of ReBreathe.

AB - Training stress-prediction models is challenging due to the difficulty in reliably eliciting stress and relaxation responses in participants. For example, a task intended to elicit a relaxation response (e.g., deep breathing) can have the opposite effect depending on the participant's appraisal of and familiarity with the exercise. Including such instances in a training set undermines the accuracy of the resulting prediction model. This paper presents a technique, ReBreathe, to identify such instances based on respiratory patterns and determine their accurate stress/relax labels. We compared this relabeling approach against two labeling techniques: 1) nominal labels obtained from the experimental protocol and 2) labels obtained from subjective assessments. We then trained generalized estimating equation regression models to predict the resulting stress/relax labels from measures of heart rate variability and electrodermal activity. Training the model using protocol labels achieved a classification rate of 0.53 on participants not included in the training set. Relabeling the exercises based on each participant's subjective ratings increased classification rates but only marginally (0.61). In contrast, relabeling the exercises based on respiratory patterns increased classification rates to 0.88, or a four-fold reduction in error rates. These results illustrate the unreliability of protocol and subjective labels during stress/relax exercises and the potential benefits of ReBreathe.

KW - deep breathing

KW - electrodermal activity

KW - heart rate variability

KW - stress prediction

KW - training protocols

UR - http://www.scopus.com/inward/record.url?scp=84973469374&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84973469374&partnerID=8YFLogxK

U2 - 10.1109/TAFFC.2015.2459682

DO - 10.1109/TAFFC.2015.2459682

M3 - Article

AN - SCOPUS:84973469374

VL - 7

SP - 150

EP - 161

JO - IEEE Transactions on Affective Computing

JF - IEEE Transactions on Affective Computing

SN - 1949-3045

IS - 2

M1 - 7164293

ER -