ArticlePDF Available

Stress Detection Using Smartphone and Wearable Devices: A Review

Authors:

Abstract and Figures

Stress is the mental condition of the human body that causes it's dis-functioning. It affects adversely on body parts resulting in health disorders. Traditional method of stress detection includes lab tests done by doctor. Besides traditional techniques, sensors are used to measure physiological signals, as these signals make it easy to detect stress. Based on techniques of data collection, this paper is divided into two types, one for In-lab experiment, in which participants wear various sensors on their body which is invasive for real time application while in second, data was collected from sensors which are already available in the handy devices of participant such as smartphone, wearable devices etc. Different types of sensors and their uses are explained in this paper. Automatic real time stress detection systems can be developed. This paper lists various algorithms used to gain more accuracy in detecting stress. This paper is helpful for the fellow researchers who will be working on automatic stress detection. Various studies in this domain have been reviewed and this is a primary effort in summarizing the highlights of the previous research done in stress detection domain.
Content may be subject to copyright.
Asian Journal of Convergence in Technology Volume V Issue I
ISSN NO: 2350-1146 I.F-5.11
Stress Detection Using Smartphone and Wearable
Devices: A Review
Tejaswini Panure
School of Computer Science and
Engineering
Dr. Vishwanath Karad MIT World
Peace University
Pune, India
tejaswinipanure@gmail.com
Shilpa Sonawani
School of Computer Science and
Enigineering
Dr. Vishwanath Karad MIT World
Peace University
Pune, India
shilpa.sonawani@mitpune.edu.in
Abstract—Stress is the mental condition of the human body that
causes it’s dis-functioning. It affects adversely on body parts
resulting in health disorders. Traditional method of stress
detection includes lab tests done by doctor. Besides traditional
techniques, sensors are used to measure physiological signals,
as these signals make it easy to detect stress. Based on
techniques of data collection, this paper is divided into two
types, one for In-lab experiment, in which participants wear
various sensors on their body which is invasive for real time
application while in second, data was collected from sensors
which are already available in the handy devices of participant
such as smartphone, wearable devices etc. Different types of
sensors and their uses are explained in this paper. Automatic
real time stress detection systems can be developed. This paper
lists various algorithms used to gain more accuracy in detecting
stress. This paper is helpful for the fellow researchers who will
be working on automatic stress detection. Various studies in
this domain have been reviewed and this is a primary effort in
summarizing the highlights of the previous research done in
stress detection domain.
Keywords—Stress detection, Physiological signals, Fitbit, E4
I. MOTIVATION
According to World Health Organization (WHO), the count
of people suffering from depression is nearly 350 million.
Stress can lead to depression and further leads to suicide.
Count of suicide is almost 1 million per year [31], so it is
necessary to detect stress at primary level. Positive stress is
like a motivation while negative stress badly affects human
health as it is long lasting [34]. Being a long term it
adversely affects human body, so it is necessary to detect
negative stress. It is mostly seen in the people at working
areas where the workload is more, along with deadlines and
pressure of completion of the work.
II. STRESS DETECTION SYSTEM
A stress detection methodologies studied here can be
divided into three types as follows.
Figure 1. Traditional Stress Detection System
www.asianssr.org 1
Asian Journal of Convergence in Technology Volume V Issue I
ISSN NO: 2350-1146 I.F-5.11
Figure 2. Automatic Stress Detection System
A. Traditional Methods
Following are few traditional methods to detect stress.
Medical Tests – These can be done to know each
physiological signals which reflects stress clearly for e.g.
heart rate, galvanic skin response etc. [16], as shown in
Figure 1.
Facial expression - In our day to day life, we come across
many people. Some are known to us while some are not.
We can easily detect the changes in behavior or the
emotions of a person which are very well known to us.
We can tell whether he/she is tensed by observing his/her
behavior. But same thing doesn't happens with the person
which are not known or very less known to us. Because
knowing someone's behavioral changes happens only when
the person is known to you.
Hormone analysis During stress, level of various
hormones gets changed [33]. So by testing level of these
hormones stress can be detected.
As stress is becoming one of the major risk factor to the
health, it needs to be detected and reduced. To know whether
the person is stressed or tensed, there should be some way
that will detect their daily behavioral patterns and will detect
changes in that patterns.
To know these patterns, system needs to collect required
data of each user, analyze it and make inferences
accordingly.
The main challenge in detecting stress is that it varies from
person to person i.e. although same work is given to all, it
shows different levels of stress for every other person [11].
Stress of a person can be detected by examining the changes
in emotions, daily routine etc. But this cannot be applied for
everyone. For this we need to collect all the data of an
individual and reveal all the changes in the regular behavior.
Collection of data can be done in two ways, one is using
medical tests and other using sensors in those devices
handled by the user. These devices include smartphones,
smartwatches, Fitbit etc., which have many sensors in it,
Few are explained in Table II below. By considering these two
ways, let’s start with stress detection using sensors.
B. Using Sensors for Stress Detection
When an individual is stressed, there are few fluctuations
(depending on how much the person is stressed) in
physiological parameter values [16]. Various physiological
parameters used are explained in Table I. Sensors are
attached to different body parts as shown in Figure 2.
Autonomic nervous system manages various functionalities
of a human body [16]. Sympathetic nervous system and
parasympathetic nervous system are the categories of ANS.
SNS controls the reactions occurred when person is stressed
while PNS controls resting activities of body [17]. To
calculate these parameters, various sensors are available.
These physiological parameters are calculated using various
sensors and used for stress detection as follows.
To record brainwaves, a device which is an interface
between brain and computer is kept on head and later
divided into different bandwidths [1].EEG
(Electroencephalography) is used to record signals of brain
activities. Neurofeedback technique, which is the modern
way to reduce stress is also used. i-care stress is an android
application in which data is taken in two ways, one from
emotive device which records brain activities and other from
questionnaire to know the state of stress. This raw data is
processed, and stress level is calculated. To reduce the
calculated stress three actions are provided, user is free to
choose one of them. Actions include concentration, music
therapy, and relaxation. In concentration user has to control
his/her mind, and by doing so, stress can be reduced. In
relaxation, deep breathing technique is used, which makes
user calm. Third technique is music therapy, which requires
entire mind to work simultaneously, hence relaxes mind.
Stress levels, before and after applying above actions are
recorded and are shown graphically, so as the user can
observe the changes.
TABLE I. PHYSIOLOGICAL PARAMETERS USED
Physiological Signals/ parameters References
Electrodermal Activity of skin (EDA) [17,18,20,22,25,29]
Blood Volume Pulse (BVP) [17]
Electrocardiography (ECG) [17,18]
Electromyography (EMG) [17,25]
Heart Rate Variability (HRV) [7]
Heart Rate (HR) [18,20,22]
Respiration Rate (RR) [20,25]
Systolic Blood Pressure (SBP) [20,29]
Electroencephalography (EEG) [26]
Sioni R discussed how different physiological signals are
affected during stress and various techniques to measure
these signals such as Electrodermal activity of skin, Blood
Volume Pulse, Electrocardiography, Electromyography etc.
To measure these, various sensors and equipment are
required which needs to be kept on particular area of body.
Also it is recommended that these sensors should be
www.asianssr.org 2
Asian Journal of Convergence in Technology Volume V Issue I
ISSN NO: 2350-1146 I.F-5.11
included in the wearable devices. Also, heart rate variability
(HRV), is considered as one of the important factor deciding
stress, as it is a non-stationary signal. Various operations like
preprocessing, analysis, feature extraction are performed on
ECG signals. Feature extraction is done on the basis of time
domain analysis as well as frequency domain analysis [7].
TABLE II. VARIOUS ALGORITHMS USED
Algorithms References
Random Forest [11]
Support Vector Machine (SVM) [18,2,21,24,11]
K-nearest Neighbor (KNN) [18,2]
j48 [8,11]
Naïve Bayes [8,21,24]
Linear discriminant function (LDF [21]
Bagging [11]
DBSCAN [15]
k-means [15,26]
Feature extraction of Heart Rate, Galvanic Skin response,
Heart Rate Variability is done using time and frequency as
parameters, in which 92.75% accuracy is achieved. K-
nearest neighbor and Support Vector Machine algorithms are
used for classification. Using different sensors, physiological
data is taken from working people and dataset is generated
(SWELL-KW dataset) [18].
To develop a real time system, neural networks are trained
using signals like Respiration Rate, Heart Rate, and Systolic
Blood Pressure. These proved most effective for accuracy
[20]. The sensors capture data continuously resulting in
more energy consumption, to avoid this, first data of
galvanic skin response is taken and analyzed If it shows
symptoms of stress then level of above mentioned
physiological signals is taken. Stress detection process takes
place only when these signals show the symptoms of stress.
The study in which devices or machines are trained such
that it recognizes human emotions and stimuli like artificial
intelligence [5].
Instead of using various sensors to know stress, a single
sensor i.e. Electro-dermal Activity sensor is used in [21]. For
this, MIT Media lab 'stress database is used. Model is
trained, tested and then used. 81.82% accuracy is achieved
[21].
Using just two signals i.e. Heart Rate [22] and Galvanic
Skin Response [29,22] an accuracy of 99.5% is achieved. In-
Lab experimentation is done. To collect data, sensors were
wore by the participants [22].
Electroencephalography signals of a patient are continuously
taken for medical diagnosis. EEG detects the change in brain
activities which can be noticed by their frequency changes.
These signals are used in [6,26]. Using k-means clustering
[26], and deep learning methodology [6], patients are
divided into different categories of stress. Just EEG signal is
used, therefore consumption of human efforts and time is
reduced. Dataset is obtained from Matrix Radiotherapy from
Kolhapur [26].
C. Automatic Sensing Methodologies
Almost everyone uses smartphone frequently. It has various
inbuilt sensors for different purposes as explained in Table
III. Less or more of these sensors are available in wearable devices
like Fitbit, E4 etc. Uses of these sensors is explained in the
www.asianssr.org 3
Asian Journal of Convergence in Technology Volume V Issue I
ISSN NO: 2350-1146 I.F-5.11
table. Also no special handling of these sensors is required.
Pattern of using smartphone and sensing data impacts a lot on
person’s stress [14]. Along with this, sleeping pattern and
social interaction is also taken for better results [4]. Data
obtained from these sensors is used for stress detection
by various researchers, as explained below.
Accelerometer sensor in smartphone collects physical
activity of the user.
But when smartphone is motionless (kept somewhere
during sleep, meeting etc.) creates obstructions in
collecting physical activity data. Same issue doesn’t
occur when wearable devices like smartwatches are
used because same sensors are available in smartwatch
also. Here smartphone and smartwatch are used for
data collection [13].
Orientation and placement of these sensors in smartphone or
other wearable devices greatly influence the success of this
system. But in real time application the variability in the
placement and orientation of these devices should be handled
properly. Fusion of sensors is a technique in which data from
sensors is analyzed together to get more accurate results [13].
According to Thapliyal H., using IoT, various devices can be
developed for automatic stress detection. Various sensors that
collect different types of data required to detect stress are
explained. A mobile application is necessary to function these
sensors and analyze data.
Here five wearable devices are taken for this study, these are
HeartMath Inner Balance, Spire, WellBe, Zensorium Being,
Zensorium Tinke. All these devices are attached with at least
one, either android or iOS application having access to the
location and calendar with the users permissions. According to
this study, HRV method is important to detect stress.
TABLE III. SENSORS AND THEIR USES
permission of user to his/her highly private and sensitive
contents. Data from sensors of smartphone such as
microphone, accelerometer, electronic compass, light sensor
and app usage history is taken from user for this study. Mood-
Explorer is an android app which collects sensing information.
This app sends notifications to user and reports their emotional
state. Around 76% accuracy is taken from ML model [14].
Jin Lu’s combination of smartphone and Fitbit along with
questionnaire data are used. In the beginning of this study,
clinical diagnosis of each participant is done. From
smartphone, location and physical activity of the user is
taken. For clustering DBSCAN is used. Activity data, heart
rate, sleeping patterns are extracted from Fitbit. Here multi-
task learning is a proposed method [15].
III. DISCUSSION
According to World Health Organization (WHO), in future one
individual among four will be suffering from mental disorder
[32]. As we saw in above studies, to know the person’s stress, it
is important to know all the activities of a user along with their
physiological parameters. Data is captured in two different
forms, one in constrained environment, taking few participants
for particular period of time (having some restrictions) for the
purpose of study and second in unconstrained environment i.e.
participants were free to do anything without any restriction.
Some studies captured this data using sensors. But to develop a
system as a real time application, user will be in unconstrained
environment.
Now to capture data, few studies used special sensors which
were attached to the user’s body, and some took data from the
devices which user uses daily such as smartphone and various
wearable devices.
But having these sensors attached, user’s body created
disturbance in their daily routine. So capturing data from
devices which user uses daily will be more comfortable.
Various algorithms and techniques are applied as shown in
Table II.
Gjoreski M., proved that using only smartphone creates many
challenges. Though the work profile is same for the people,
stress observed is different for each person. Here person wise
different model worked better using random forest. Along with
the use of sensors that detects physiological signals, voice
analysis can also be done [11]. Person specific model works
better [12], but for this, more labelled data is required and in
case of general model, no labelling is required. Garcia-Ceja
only accelerometer is used for sensing data.
Even for mood prediction, blood pressure, heart beats, blood
pressure etc. can be considered. SMS, Email contents also
gives basic idea of the mood of a person, but it requires
www.asianssr.org 4
Sensors Uses
Accelerometer Gives Activity Data.
To detect activity of a person such as whether
he/she is stationary, walking or running [11]
Microphone To detect audio surrounding of a person (silence,
voice or noise) [11]
Conversation data To detect if person is near conversation [11]
GPS data
(GPS Chip) To know the location of person [11]
Call log data Information about time and duration of calls [11]
Light sensor data
If the phone is in dark, charging or locked for a
longer period of time(battery charging andlocking
data) [11]
Optical Sensor
Variations in blood volume and heart rate are
detected using LED light that reflects off tothe skin.
Available in Zensorium Being device [9].
Infrared Blood
Flow Sensor
Using IR light, blood flow speed in ear is calculated
[9].
Heart Rate
Monitoring Sensors Using Heart Rate Variability, heart rate is taken [9]
Asian Journal of Convergence in Technology Volume V Issue I
ISSN NO: 2350-1146 I.F-5.11
IV. CONCLUSION
World Health Record of World Health Organization,
mentioned that among four individuals, one will suffer from
mental disorder. 450 million people around world are
suffering from stress [32]. Although much of study is done
in the area of stress detection, more accuracy in
unconstrained environment is needed. As we have seen
above, no extra handling of sensors will be preferred by a
user. It is fruitful to take data using sensors which are
already present in the devices handled by user frequently.
Combination of various sensors will result in better
accuracy.
REFERENCES
[1]Mitrpanont J., Phandhu-fung, J., Klubdee, N., Ratanalaor, S., Pratiphakorn, P.,
Damrongvanakul, K., Piyawat Chuanvaree, P. and Mitrpanont, T. (2019).
iCare-Stress: Caring system for stress - IEEE Conference Publication .
[online] Ieeexplore.ieee.org. Available at:
https://ieeexplore.ieee.org/document/8075319/ [Accessed 7 Jan. 2019].
[2]Akmandor, A. and Jha, N. (2017). Keep the Stress Away with SoDA: Stress
Detection and Alleviation System. IEEE Transactions on Multi- Scale
Computing Systems, 3(4), pp.269-282.
[3]de Santos Sierra, A., Sanchez Avila, C., Guerra Casanova, J. and Bailador del
Pozo, G. (2011). A Stress-Detection System Based on
Physiological Signals and Fuzzy Logic. IEEE Transactions on Industrial
Electronics, 58(10), pp.4857-4865.
[4] Kostopoulos, P., Kyritsis, A., Deriaz, M. and Konstantas, D. (2016). Stress
Detection Using Smart Phone Data. Lecture Notes of the Institute for
Computer Sciences, Social Informatics and Telecommunications
Engineering, pp.340-351.
[5] Greene, S., Thapliyal, H. and Caban-Holt, A. (2016). A Survey of Affective
Computing for Stress Detection: Evaluating technologies in stress detection
for better health. IEEE Consumer Electronics Magazine, 5(4), pp.44-56.
[6] Liao, C., Chen, R. and Tai, S. (2018). Emotion stress detection using EEG
signal and deep learning technologies. 2018 IEEE International Conference
on Applied System Invention (ICASI).
[7] Ramteke, R. and Thool, V. (2017). Stress detection of students at academic
level from heart rate variability. 2017 International Conference on Energy,
Communication, Data Analytics and Soft Computing (ICECDS).
[8] Gjoreski, M., Luštrek, M., Gams, M. and Gjoreski, H. (2017). Monitoring
stress with a wrist device using context. Journal of Biomedical Informatics,
73, pp.159-170.
[9] Thapliyal, H., Khalus, V. and Labrado, C. (2017). Stress Detection and
Management: A Survey of Wearable Smart Health Devices. IEEE
Consumer Electronics Magazine, 6(4), pp.64-69.
[10] Ollander, S., Godin, C., Campagne, A. and Charbonnier, S. (2016). A
comparison of wearable and stationary sensors for stress detection. 2016
IEEE International Conference on Systems, Man, and Cybernetics (SMC).
[11] Gjoreski, M., Gjoreski, H., Lutrek, M. and Gams, M. (2015). Automatic
Detection of Perceived Stress in Campus Students Using Smartphones.
2015 International Conference on Intelligent Environments.
[12] Garcia-Ceja, E., Osmani, V. and Mayora, O. (2016). Automatic Stress
Detection in Working Environments From Smartphones’ Accelerometer
Data: A First Step. IEEE Journal of Biomedical and Health Informatics,
20(4), pp.1053-1060.
[13] Vaizman, Y., Ellis, K. and Lanckriet, G. (2017). Recognizing Detailed
Human Context in the Wild from Smartphones and Smartwatches. IEEE
Pervasive Computing, 16(4), pp.62-74.
[14] Zhang, X., Li, W., Chen, X. and Lu, S. (2018). MoodExplorer. Proceedings
of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies,
1(4), pp.1-30.
[15] Lu, J., Bi, J., Shang, C., Yue, C., Morillo, R., Ware, S., Kamath, J., Bamis,
A., Russell, A. and Wang, B. (2018). Joint Modeling of Heterogeneous
Sensing Data for Depression Assessment via Multi-task Learning.
Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous
Technologies, 2(1), pp.1-21.
[16] Xia, L., Malik, A. and Subhani, A. (2018). A physiological signal- based
method for early mental-stress detection. Biomedical Signal Processing and
Control, 46, pp.18-32.
[17] Sioni, R. and Chittaro, L. (2015). Stress Detection Using Physiological
Sensors. Computer, 48(10), pp.26-33.
[18] Sriramprakash, S., Prasanna, V. and Murthy, O. (2017). Stress Detection in
Working People. Procedia Computer Science, 115, pp.359- 366.
[19] Smith, A. and Thomas, A. (2018). Reducing the Consequences of Acute
Stress on Memory Retrieval. Journal of Applied Research in Memory and
Cognition, 7(2), pp.219-229.
[20] Koussaifi, M., Habib, C. and Makhoul, A. (2018). Real-time stress
evaluation using wireless body sensor networks. 2018 Wireless Days(WD).
[21] Liu, Y. and Du, S. (2018). Psychological stress level detection based on
electrodermal activity. Behavioural Brain Research, 341, pp.50-53.
[22] de Santos Sierra, A., Sanchez Avila, C., Guerra Casanova, J. and Bailador
del Pozo, G. (2011). A Stress-Detection System Based on Physiological
Signals and Fuzzy Logic. IEEE Transactions on Industrial Electronics,
58(10), pp.4857-4865.
[23] Han, L., Zhang, Q., Chen, X., Zhan, Q., Yang, T. and Zhao, Z. (2017).
Detecting work-related stress with a wearable device. Computers in
Industry, 90, pp.42-49.
[24] Zhang, X., Lyu, Y., Luo, X., Zhang, J., Yu, C., Yin, H. and Shi, Y. (2018).
Touch Sense. Proceedings of the ACM on Interactive, Mobile, Wearable
and Ubiquitous Technologies, 2(2), pp.1-18.
[25] “Stress Detection Using Physiological Sensors”, Riccardo Sioni and Luca
Chittaro, Volume 48, Issue 10, October 2015.
[26] Kalas, M. and Momin, D. (2019). Stress detection and reduction using EEG
signals - IEEE Conference Publication. [online] Ieeexplore.ieee.org.
Available at: https://ieeexplore.ieee.org/document/7755604 [Accessed 8
Jan. 2019]. [27] Aigrain, J., Spodenkiewicz, M., Dubuiss, S., Detyniecki,
M., Cohen, D. and Chetouani, M. (2018). Multimodal Stress Detection
from Multiple Assessments. IEEE Transactions on Affective Computing,
9(4), pp.491-506.
[28] VAIZMAN,, Y., WEIBEL,, N. and LANCKRIET,, G. (2019). Context
Recognition In-the-Wild: Unified Model for Multi-Modal Sensors and
Multi-Label Classification. [online] Extrasensory.ucsd.edu. Available at:
http://extrasensory.ucsd.edu/papers/vaizman2017b_imwutAcceptedVersion
.pdf [Accessed 8 Jan. 2019].
[29] Fernandes, A., Helawar, R., Lokesh, R., Tari, T. and V. Shahapurkar,
A. (2019). Determination of stress using Blood Pressure and Galvanic Skin
Response - IEEE Conference Publication. [online] Ieeexplore.ieee.org.
Available at: https://ieeexplore.ieee.org/document/7062747 [Accessed 8
Jan. 2019].
[30] Marina Marcus, M. Taghi Yasamy, Mark van Ommeren, and Dan Chisholm,
Shekhar Saxena. (2012). Depression: A Global Public Health Concern |
HESP News Briefing – RSS feeds. [Online]. Available:
http://hespnews.org /2012 /10/ 05/depression-a-global-public-health-
concern/.
[31] http://www.who.int/mental_health/management/depression/who_paper
_depression_wfmh_2012.pdf
[32] https://www.who.int/whr/2001/media_centre/press_release/en/
[33] Ranabir, S. and Reetu, K. (2011). Stress and hormones. Indian Journal of
Endocrinology and Metabolism, 15(1), p.18.
[34]https://www.uml.edu/docs/IntroductionToJobStress_tcm18-42460.pdf
www.asianssr.org 5
ResearchGate has not been able to resolve any citations for this publication.
Article
The early detection of mental stress is critical for efficient clinical treatment. As compared with traditional approaches, the automatic methods presented in literature have shown significance and effectiveness in terms of diagnosis speed. Unfortunately, the majority of them mainly focus on accuracy rather than predictions for treatment efficacy. This may result in the development of methods that are less robust and accurate, which is unsuitable for clinical purposes. In this study, we propose a comprehensive framework for the early detection of mental stress by analysing variations in both electroencephalogram (EEG) and electrocardiogram (ECG) signals from 22 male subjects (mean age: 22.54 ± 1.53 years). The significant contribution of this paper is that the presented framework is capable of performing predictions for treatment efficacy, which is achieved by defining four stress levels and creating models for the individual level. The experimental results indicate that the framework has realised an accuracy, a sensitivity, and a specificity of 79.54%, 81%, and 78%, respectively. Moreover, the results indicate significant neurophysiological differences between the stress and control (stress-free) conditions at the individual level.
Article
Automatic recognition of behavioral context (location, activities, body-posture etc.) can serve health monitoring, aging care, and many other domains. Recognizing context in-the-wild is challenging because of great variability in behavioral patterns, and it requires a complex mapping from sensor features to predicted labels. Data collected in-the-wild may be unbalanced and incomplete, with cases of missing labels or missing sensors. We propose using the multiple layer perceptron (MLP) as a multi-task model for context recognition. Based on features from multi-modal sensors, the model simultaneously predicts many diverse context labels. We analyze the advantages of the model's hidden layers, which are shared among all sensors and all labels, and provide insight to the behavioral patterns that these hidden layers may capture. We demonstrate how recognition of new labels can be improved when utilizing a model that was trained for an initial set of labels, and show how to train the model to withstand missing sensors. We evaluate context recognition on the previously published ExtraSensory Dataset, which was collected in-the-wild. Compared to previously suggested models, the MLP improves recognition, even with fewer parameters than a linear model. The ability to train a good model using data that has incomplete, unbalanced labeling and missing sensors encourages further research with uncontrolled, in-the-wild behavior.
Article
Psychological stress has been shown to impair episodic memory retrieval. Implicated in this memory impairment is the physiological stress response, which interferes with retrieval-related neural processing. An important next step in research is to determine how to improve post-stress memory accessibility. In this review, we first consider methodological differences in studies that have examined stress and memory, as they lend insight into the conditions under which stress does and does not impede retrieval. Motivated by these variations in methodology, we advocate for two potential approaches to intervention. One approach is to employ evidence-based techniques that reduce the physiological stress response. A second approach is to target the processes that occur during initial learning to promote the formation of highly accessible memories. Thus, this review serves to both critically evaluate the methods used to examine the effects of stress on memory retrieval and encourage research on interventions for stress-related memory impairment.
Article
This paper presents a method for detecting psychological stress levels. It aims to explore the feasibility of using a single physiological signal to create a more practical alternative for detecting stress in people than current multiple physiological signals approaches involve. In particular, the approach explored uses linear discriminant analysis (LDA) based on the electrodermal activity (EDA) signal which aims at discriminating between three stress levels: low, medium and high. We used the MIT Media lab 'stress database' from which we selected eleven 'foot' based EDA data sets for our experiments. Using this eighteen EDA features were extracted from (sixty-six) five-minutes data segments equating to three driving conditions: at rest, on the open road (highway) and city driving. After that, Fisher projection and Linear discriminant analysis (LDA) were used to classify the stress levels with feature vectors, that included both leaving one out and test cross-validation strategy. The results showed that these methods achieved recognition rate of 81.82% which we argue, while less that multiple signal systems, may be a better balance between recognition performance and computational load, that could be a promising line of research for the development of practical personal stress monitors.
Article
Stress detector classifies a stressed individual from a normal one by acquiring his/her physiological signals through appropriate sensors such as Electrocardiogram (ECG), Galvanic Skin Response (GSR) etc,. These signals are pre-processed to extract the desired features which depicts the stress level in working individuals. Support Vector Machine (SVM) and K-Nearest Neighbour (KNN) are investigated to classify these extracted feature set. The result indicates feature vector with best features having a strong influence in stress identification. An attempt is made to determine the best feature set that results in maximum classification accuracy. Proposed techniques are applied on benchmark SWELL-KW dataset and state-of-art results are obtained.
Article
In today's world, many people feel stressed out from school, work, or other life events. Therefore, it is important to detect stress and manage it to reduce the risk of damage to an individual's well being. With the emergence of the Internet of Things (IoT), devices can be made to detect stress and manage it effectively by using cloudbased services and smartphone apps to aggregate and compute large data sets that track stress behavior over long periods of time. Additionally, there is added convenience via the connectivity and portability of these IoT devices. They allow individuals to seek intervention prior to elevated health risks and achieve a less stressful life.
Article
Being able to detect stress as it occurs can greatly contribute to dealing with its negative health and economic consequences. However, detecting stress in real life with an unobtrusive wrist device is a challenging task. The objective of this study is to develop a method for stress detection that can accurately, continuously and unobtrusively monitor psychological stress in real life. First, we explore the problem of stress detection using machine learning and signal processing techniques in laboratory conditions, and then we apply the extracted laboratory knowledge to real-life data. We propose a novel context-based stress-detection method. The method consists of three machine-learning components: a laboratory stress detector that is trained on laboratory data and detects short-term stress every 2 minutes; an activity recognizer that continuously recognizes the user's activity and thus provides context information; and a context-based stress detector that uses the outputs of the laboratory stress detector, activity recognizer and other contexts, in order to provide the final decision on 20-minute intervals. Experiments on 55 days of real-life data showed that the method detects (recalls) 70% of the stress events with a precision of 95%.