Content uploaded by Carlos A Almenara
Author content
All content in this area was uploaded by Carlos A Almenara on Jul 23, 2020
Content may be subject to copyright.
Abstract— This work proposes the analysis in time and
frequency of EEG and EMG waves with the purpose of obtaining
stress states in 5 levels. Due to the advance and evolution of
technology, it is possible to obtain low-cost brain-computer
interfaces with greater ease in neurofeedback sessions, which, in
turn, helps to reduce stress levels in individuals and this requires a
deeper analysis due to a scarce investigation. Different studies are
limited to only obtaining binary results, that is to say, if an
individual is in a state of stress or not, but they do not present
results in a scale of levels. We analyzed 6 EEG channels with a
sampling frequency of 250 Hz following the 10-20 standard and 1
EMG channel decomposed in the time and frequency domain
obtaining parameters with the discrete wavelet transform and
energy per band. The parameters obtained from each signal were
entered into a k-NN classifier. In the same way, for the validation,
the stress level was established by the graphic analysis of the heart
rate variability following Baevsky's method, following with the
measurement of the relaxation and stress of differents students
while subjecting them to psychotechnical tests. The proposed
algorithm was able to differentiate in most cases and quantify the
states, reaching an accuracy level of 92%.
Keywords— EEG, EMG, HRV, stress, wavelet transform, k-NN,
Baevsky stress index.
I. INTRODCTION
Stress in people is caused by work, studies, economic
problems and family problems among others, and it affects
their welbeing causing deterioration in mental and physical
health of the individual.
For this reason, it is becoming increasingly necessary to
develop methods, procedures and technological tools to
quantify the level of stress of people, so that specialists,
through biofeedback sessions, can obtain more reliable
information to achieve a more efficient control of the
problem and thus achieve greater welbeing of the individual.
At present, the procedure used by specialist and
psychologists to determine a person's state of mind is
carried out by means of perceptual tests that are subjective,
since they depend on the experience of the specialist and the
willingness of the patient to collaborate. This method can
cause an erroneous diagnosis of the patient's condition and,
therefore, a treatment that does not generate improvement
and well-being.
In the scientific literature several proposed methods have
also been found with limitations in the precision and the lack
quantifitaction levels of the stress level.
Subhani et al. proposes for example an algorithm in [1]
oriented to differentiate between a state of stress and a state
of relaxation by analyzing brain waves in two channels. In
this case, an accuracy of 93.4% was achieved. However, the
method is limited to discerning two states (stressed and
relaxed) and does not allow results to be obtained at
multiple levels.
Similarly, the paper presented by Richer et al. in [2]
proposes an algorithm that analyzes the power of the
frequency bands of brain signals. However, they achieve an
accuracy no greater than 83% by differentiating between a
state of stress and a state of relaxation.
Amores et al., on the other hand, proposes in [3] a method
to detect the level of relaxation of an individual on several
scalesby the analysis of electroencephalogram waves (EEG).
However, they do not specify the precision of the proposed
method and use subjective relaxation perception methods
for the validation of the stress leve lof an individual.
Wang et al. in [4] proposes an algorithm to differentiate
between a state of relaxation or meditation and a state of
stress using a brain-computer interface. In this case, an
accuracy of 90.3% was achieved. However, like some of the
above methods, it is limited to differentiating only two
states.
Liu et al. on the other hand proposes in [5] a method to
detect between a state of attention or relaxation obtained by
brain waves of students. However, this study was less than
80% accurate and is limited to only two states of mind.
The studies conducted by Al-Shargie et al. in [6] and [7]
propose a method for determining the stress level of an
individual quantified at multiple levels from EEG waves. In
the first study, an accuracy of 86.3% was achieved by
differentiating three stress levels, and in the second study,
accuracy was improved to 94.79%. However, this method is
limited to an analysis with few levels to differentiate.
Karthikeyan et al. in [8] presents a novel algorithm to
determine a person's stress level from electromyographic
(EMG) signals measured in the trapezium muscle,
differentiating between four levels with an accuracy of
90.7%. However, the proposed method does not specify the
procedure or the index used to measure the accuracy
obtained and only works with four levels of stress.
Finally, the process carried out by Ahani et al. in [9] used
2 physiological signals to measure a person's stress level,
obtaining an accuracy of 85% for 2 states and an accuracy of
78% when differentiating in multiple levels. This novel
Diego E. Ugarte1, David Linares2, Guillermo Kemper3 and Carlos A. Almenara4
1,2,3Faculty of Engineering - School of Electronic Engineering
4Faculty of Health Sciences
Universidad Peruana de Ciencias Aplicadas, Lima, Peru
1u201210914@upc.edu.pe, 2u201214233@upc.edu.pe, 3pcelgkem@upc.edu.pe, 4carlos.almenara@upc.pe
An Algorithm to Measure the Stress Level from
EEG, EMG and HRV Signals
Please cite:
Ugarte, D. E., Linares, D., Kemper, G., & Almenara, C. A. (2019). An algorithm to measure the stress level from EEG, EMG and HRV signals. In 2019 International Conference on Information Systems and Computer
Science (INCISCOS) (pp. 346–353). Quito, Ecuador: IEEE. https://doi.org/10.1109/INCISCOS49368.2019.00061
method uses the analysis of EEG and breathing signals to
determine the degree of stress presented by the person.
However, the low accuracy of this method prevents a correct
analysis for specialists and quantifies stress on a scale of
three levels at most.
The method proposed in this work presents the novelty
of using parameters extracted simultaneously from EEG and
EMG signals in order to find the stress level of an individual
while meditating. The EMG signals are obtained from the
involuntary movements of the left trapezium in order to
detect the stress level on a five-level scale. In addition, in the
present work the five levels established by the stress index
were used, extracted from the graphical analysis of the
variation of the cardiac pulse (HRV), which is an objective
measure to quantify the stress of an individual.
All this solves the subjectivity in obtaining the stress level,
which is obtained by traditional methods, such as surveys or
long term examinations.
The 2 types of signals are processed and analyzed offline
through the algorithm proposed in this paper. The algorithm
extracts characteristic parameters from both signals and
then submits them to a classifier to obtain the stress level.
The computational tool developed can be used by specialists
in order to improve the effectiveness of their diagnoses and
know if the method used is the right one.
The results obtained were validated using the reference
stress level obtained from the HRV. The results show an
accuracy in the detection of the stress level of 92%.
II. DESCRIPTION OF THE PROPOSED ALGORITHM
The present algorithm employs the use of biomedical
signals to determine a scale of stress in people while
meditating with the mindfulness method, which consists of
a meditation that involves the perception internal and
external factors of an individual.
Fig. 1 illustrates the block diagram of the proposed
algorithm. Its parts will be described in the following
sections.
A. EEG data acquisition
The OpenBCI Cyton module [10] was used for data
acquisition. It has 8 measurement channels with a sampling
frequency of 250 samples per second. It also has a wireless
connection to a computer through an RFduino module. This
module allows the simultaneous measurement of different types
of biomedical signals such as EEG, EMG and ECG; it also
works with a resolution of 24 bits per sample or 0.02235 µV per
level.
In the acquisition process, data was obtained from 20
volunteers in the age range of 20 to 30 years who practice
meditation at a basic, intermediate or advanced level. The
process is based on a protocol that consists of closing the eyes
for 15 minutes by adapting an upright posture while seated (see
Fig. 2). The meditation period is delimited by the use of a tingsha
at the beginning and at the end of the session. Because the
objective is to measure involuntary movements caused by stress
on the trapezium muscle, the patient is asked to avoid moving
the arms. In addition, the measurement of the heart rate is also
affected by the movement, so this requirement is essential.
Data collection is performed (measuring brain signals)
following the international 10-20 standard at points Fp1, Fp2,
T6, T7, O1 and O2 with Ag-AgCl dry electrodes. The
reference and ground are placed at points A1 and A2 as shown
in Fig. 3 [11].
Frontal points are chosen because of their large variation in
beta waves which increase with a state of stress and relaxation
[9]. The temporal and occipital points represent great
variation of theta waves as shown in [9] and finally, alpha
waves are reflected with greater intensity in the temporal and
occipital points as concluded in [12].
Fig. 2. Data acquisition diagram
To facilitate data collection, an elastic band was designed
with electrodes in the established points to obtain good contact
with different head sizes. Previous measurements with the
OpenBCI UltraMark IV helmet had errors due to an unreiable
contact.
Fig. 1. Block diagram of the presented algorithm
B. EMG data acquisition
To acquire EMG signals, two gold electrodes [13] with
conductive gel are placed on the left trapezium to measure the
energy present in the muscle at rest (the same position used in
[8] is applied). This muscle was chosen because it has favorable
results in the classification of stress states.
The 6 EEG signal electrodes and the 2 EMG signal
electrodes are connected to the Cyton module and are configured
so that the acquisition of EMG waves are in differential mode
between the two electrodes. This allows the signals to be
decoupled from the reference, so that it does not interfere with
the EEG waves that are of lower voltage. Like brain signals,
EMG waves are acquired with a sampling frequency of 250 Hz.
C. EEG signal improvement
The EEG signals obtained present noise caused by the
induction of the 60Hz component of the electrical network. It
also has a DC component on which the acquired signals are
mounted (see Fig. 4). To reduce this interference, the wavelet
denoising method was applied.
In this case the wavelet transform breaks down a signal into
its components in time and frequency. Unlike the Fourier
transform, this type of transformation has a lower resolution in
frequency and in time, but more accurately describes non-
stationary signals.
.
Fig. 3. 10-20 system for EEG. Source: Adapted from [11]
According to the experiments and results obtained in [14],
the best wavelet to eliminate induction noise in EEG signals
is the Daubechies 8 ("db8").
The procedure of the denoising method is described
below:
• The wavelet wave type and the decomposition level
are chosen.
• The type and the threshold rule are chosen and
implemented.
• The thresholded signal is reconstructed with the
inverse wavelet transform.
The decomposition tree for obtaining the wavelets sub-
bands of interest for the wavelet denoising procedure is shown
in Fig. 5 [15]. In this case represents the signal that is
subject to decomposition.
Each resulting wavelet coefficient is defined as ,
where i is the level of decomposition and j is the number of
the sub-band obtained at level .
Table I shows the mapping of the EEG signals and the sub-
bands obtained with the decomposition tree.
As specified in [16] the best method to determine the
threshold is by absolute deviation of the median, which is
normalized to the value to scale the numerator and
thus is a suitable estimator of the white Gaussian additive
noise. This is expressed as [16]:
(1)
Where represents the wavelet coefficient vector of the
sub-band.
(a)
(b)
Fig. 4. EEG signal from Fp1 position without. (a) Extract of the EEG
signal. (b) Zoomed EEG signal from the black box in graph (a)
TABLE I. OBTAINED SUB-BANDS AND EEG BANDS MAPPING
Coeficiente Wavelet
Banda de frecuencia
Onda EEG
0 – 3.91 Hz
Delta
3.91 – 7.81 Hz
Theta
7.81 – 15.63
Alpha
15.63 – 31.25 Hz
Beta
31.25 – 62.5 Hz
Gamma
62.5 – 125 Hz
Gamma
Fig. 5. Decomposition tree [15]
From (1) the threshold is determined for each
sub-band resulting from the decomposition tree:
(2)
Where is the total number of samples from sub-band
. Finally, the threshold coefficients are obtained from the
following expression:
(3)
The coefficients other than zero are estimated to be
free of noise and distortion. Therefore, they more reliably
reflect the characteristics of EEG signals.
It is important to note finally that the sub-band was
not taken into account in all the processing that was applied
to the EEG signals. This is because this sub-band contains the
unwanted DC component that is present in the acquired EEG
signals.
D. EMG signal improvement
Like the EEG signals, these signals have induction noise from
the mains and a DC component. The wavelet denoising method
was also used to solve this problem.
In this case the decomposition tree to obtain the wavelets sub-
bands of interest is similar to the one shown in Fig. 5 with 7
levels of decomposition.
The wavelet used in this case was the wavelet "coif5".
The wavelets coefficients of the resulting sub-bands are
defined in this case as .
For the elimination of the DC component and the baseline
wandering, the coefficients of the sub-band 7.0 were set to zero
i.e. . This sub-band corresponds to the frequency
range from 0 to 0.98 Hz.
Then the signal is finally reconstructed with the
corresponding reconstruction tree [16].
Subsequently, in order to reduce the noise, a threshold was
applied to the sub-band coefficients of the second level
since this corresponds to the frequency band in the range
[31.25Hz, 62.5Hz] and therefore contains the 60 Hz component
induced by the electrical network.
The type of threshold chosen is soft obtained from the
following expression:
(4)
In this case is the threshed wavelet coefficient,
is the wavelet coefficient of the sub-band to be threshed
and the value of the threshold.
The threshold value is obtained from the rigorouse SURE
(rigrsure) algorithm which generates an adaptive threshold
based on the Stein risk estimation principle [17]. This method
was chosen because it gives the best results compared to other
types of thresholds [18].
E. EEG feature extraction
The EEG classification parameters are obtained from the
normalized energies of the sub-bands using the following
expressions:
(5)
(6)
(7)
In this case, , and correspond to the energies of theta
waves (), alpha (), and beta ()
respectively. The subscript indicates the position of the
measured EEG points: Fp1 ( ), Fp2 (), T6 ( ), T7
(), O1 ( ), and O2 ( ). This implies that the
parameter extraction procedure is applied on the EEG wave
obtained from each of the 6 points. This resulted in a total of 18
EEG parameters for classification.
F. EMG feature extraction
In this process, the 4-level decomposition proposed in [8] was
used as a reference. However, since this method uses a sampling
frequency of 500Hz which is twice the frequency used in the
present work (250Hz), we opted for a 2-level balanced tree
wavelet packets decomposition.
This decomposition verifies that the most important
information is in the range of 0 - 32.5 Hz.
To characterize the information located in that band, the
minimum, maximum, mean, standard deviation, power, energy
and entropy are extracted from the corresponding coefficients.
G. Classification
A k-NN classifier (k-nearest neighbors) is used in which 25
EEG and EMG parameters are introduced as inputs to obtain the
stress level. The order of the classifier with which the best result
was obtained in terms of precision was 3. The classifier was
trained following the mindfulness protocol. This protocol
establishes that the participant must meditate for 15 minutes in
a seated position with closed eyes. During this time, the left arm
should not be moved because the aim is to measure the
involuntary movements of the trapeium.
Together with the EEG and EMG electrodes, a pulse sensor
is placed on the index finger of the left arm to measure the
variability of the heart rate. As stated in [19], the minimum time
to obtain the stress level from the heart pulse is 100 seconds, so
we opted for a window that covers that time and consists of a
number of samples that is a power of two for optimization of the
algorithm. That is why a window of 32768 samples is chosen,
equivalent to 131 seconds with a sampling frequency of 250 Hz.
After obtaining the data with the OpenBCI software, the
generated file is analyzed and the features corresponding to each
signal is extracted. First, the cardiac pulse is analyzed and the
time between pulses is obtained. These values are stored to
obtain the variability of the HRV heart rate and then analyzed
graphically to obtain the Baevsky’s index (which will be
explained in detail in the validation process). The use of this
index is due to the objectivity of the physiological signals as
opposed to the values obtained through perceptual tests. That is
why it has been used in studies to detect the presence of stress in
Russian astronauts before and after performing emotionally
demanding missions [20].
After the analysis of the HRV, the parameters explained in the
previous points are obtained and stored in a matrix of the same
size as the vector of the level obtained by the stress index that
are obtained in the same instants of time.
Once the parameters and the corresponding level are obtained,
a k-NN classifier is used to determine the stress value from the
EEG and EMG signals.
For each person, a classifier is trained so that the level obtained
corresponds to the unique changes an individual present.
The 20 trained classifiers present an accuracy of 92% in the
determination of the 5 levels. The levels that represented the
greatest error are 1 and 5 because they are at the extremes of the
normal distribution. Figs. 6 and 7 show the stress index obtained
by the HRV analysis and the index obtained with the classifier
for one of the participants. It is observed that the index found by
the proposed method has a high correlation with Baevsky's
reference index.
Fig. 6. Stress level obtained with HRV analysis.
Fig. 7. Stress level obtained with the classifier.
III. RESULTS
The level of stress in people is reflected in signals from the
sympathetic and parasympathetic nervous system that regulate
the changes in the different bioelectric waves.
In this context, heart rate variability (HRV) is an important
reference measure as it directly reflects the level of stress from
the graphical analysis of the RR intervals of the ECG wave (time
period between R waves). Histograms with bin widths of 50 ms
and RR time periods of 0.4 and 1.3 seconds [21] are used for the
analysis. The relationship between height in percent and width
of the intervals establishes the stress index proposed by R. M.
Baevsky in [22]. This index is expressed as:
(8)
Where is the amplitude of fashion in percent, is the
fashion expressed in seconds and is the range of the
interval of RR values expressed in seconds. All these values are
obtained from the histogram of the RR periods [22].
For a normal state of rest the value of is between 50-150
c.u. (arbitrary units) [22].
Table II shows the range of stress index values for each level.
Each of them is associated with a state of the person's emotional
state: state of relaxation (low level), normal state (normal level),
state of medium emotional stress caused by a low mental or
physical load (moderate level), state of high stress caused by
significant mental or physical loads (high level) and a state of
very high stress caused by very high emotional or physical loads
that could trigger illnesses if kept in that state for a prolonged
time (very high level) as specified by R. M. Baevsky in [23].
TABLE II. STRESS INDEX () INTERVALS
Level
Very High
≥500
High
300 – 500
Moderated
150 – 300
Normal
50 – 150
Low
≤50
To determine the stress level, HRV values are used as
reference values. These values are then mapped to the 5 levels
shown in Table II. This involves moving from quantitative to
qualitative valuation.
The process seeks to obtain objective values rather than using
perception tests based on subjective interpretation.
The PulseSensor module, which emits an analog signal and a
green light to measure the variation of blood volume changes,
was used to measure the heart rate. This equipment was modified
with a thin layer of silicone on the conductive tracks of the green
LED, as the contact of the 5V signal with the skin interfered with
the measurement of the other signals.
The sensor is placed on the index finger of the left hand and
immobilized with a velcro strap attached to the finger. The
sensor output is placed on an auxiliary analog input of the Cyton
with a resolution of 16 bits per sample.
To find the accuracy of the proposed algorithm, SVM and k-
NN classifiers were trained. In this case, a higher percentage of
accuracy was obtained with the k-NN classifiers, so this type of
classifier was chosen.
Fig. 8. Accuracy percentage of each patient.
The accuracy of the classifier is obtained from the following
expression:
(9)
Where is the number of true positives (representing the
number of hits), is the true negatives (representing the
number of correct rejections of the classifier), is the total
number of tests performed and is the accuracy of the stress
level in the test. Since the confusion matrix is multivariable, then
the true positives and negatives correspond to the diagonal
where the predicted and expected values coincide. Finally, in
order to find the total accuracy of the classifier per person, the
average of the 5 levels is obtained.
Fig. 8 shows a graph of the accuracy obtained by each patient.
Note that an average percentage of 92% is obtained in the
accuracy of the classifier.
In the cases in which the measurement presents a low
accuracy, it is due to the fact that the individual had some
movement during the data acquisition.
Tables III, IV, V, VI and VII take as an example the confusion
matrices of 5 of the 20 participants. The columns represent the
expected values obtained from the HRV analysis and the rows
represent the results obtained from the classifier with the EEG
and EMG parameters. As it can be seen, the diagonal of
coincidences is where the classification stress levels is applied
correctly and the other cells represent the classification errors.
The analysis of the diagonal of coincidence in relation to the
other cells shows the efficiency of the classifier used, for
example, if the great majority of values are in this diagonal, it
can be said that the method used can correctly predict the
expected values.
As mentioned above, the diagonal of hits corresponds to the
true positives. The true negatives, in this case, are not present in
the matrix, since we want to find the accuracy of all levels.
Therefore, the accuracy in the classification of the stress level in
patients 4, 8, 12, 18 and 19 is 94%, 96%, 85%, 96% and 94%
respectively.
Fig. 9. Kappa coefficient of each patient.
Cohen's kappa coefficient is another statistical measure that
represents the concordance between predicted and expected
values. The equation for finding the kappa coefficient is defined
in (10). Where is the sum of the expected values of each
level, which is defined in (11), and is Cohen's Kappa
coefficient. The value is a measure of the expected number
of hits and is calculated in each of the cells of the hit diagonal.
In (11) it is defined as finding the expected value of hits for each
level, where is the sum of all the values in the row of the
cell of the level to be measured and corresponds to the sum
of the values of the column. In the same way as in the case of
accuracy, the kappa coefficient is calculated for each patient and
the values 0.89, 0.92, 0.67, 0.91 and 0.89 respectively are
obtained. Fig. 9 shows a graph of the kappa coefficient for each
participant and obtains an average value of 0.83, which indicates
that the values predicted by the classifier have high concordance
with the values obtained with the , the value of 1 being a
perfect concordance.
(10)
(11)
In synthesis, the steps to follow to find the values chosen for
the validation of the classifier and the algorithm used are
summarized in the following points:
• The band with the 6 dry electrodes is placed around the
head, 2 gel electrodes are placed on the left trapezium,
and finally the pulse sensor is attached to the index of the
left hand.
• The data is stored while the individual meditates by
taking an upright, sitting posture without movement for
15 minutes.
• The data is processed, and 18 EEG wave parameters and
7 EMG wave parameters are extracted. In addition, the
stress index is found by analyzing the heart rate.
• 70% of the collected data is used to train the classifier.
The EEG and EMG wave parameters are taken as inputs
and the stress index is set as the desired output of the
classifier.
• The validity of the classifier is tested with 30% of the
remaining data with Cohen's kappa index and accuracy.
TABLE III. CONFUSION MATRIX OF PARTICIPANT 4
Values obtained from HRV
(Stress Index)
1
2
3
4
5
Predicted
Values
(k-NN)
1
14
5
0
0
0
2
8
326
9
0
0
3
0
12
1415
19
0
4
0
0
15
310
14
5
0
0
0
3
30
TABLE IV. CONFUSION MATRIX OF PARTICIPANT 8
Values obtained from HRV
(Stress Index)
1
2
3
4
5
Predicted
Values
(k-NN)
1
81
15
0
8
0
2
4
310
38
2
0
3
0
26
1667
32
0
4
0
0
20
340
2
5
0
0
0
1
53
TABLE V. CONFUSION MATRIX OF PARTICIPANT 12
Values obtained from HRV
(Stress Index)
1
2
3
4
5
Predicted
Values
(k-NN)
1
18
8
0
0
0
2
10
294
16
0
0
3
0
9
1422
21
0
4
0
0
18
316
9
5
0
0
0
4
35
TABLE VI. CONFUSION MATRIX OF PARTICIPANT 18
Values obtained from HRV
(Stress Index)
1
2
3
4
5
Predicted
Values
(k-NN)
1
96
12
0
0
0
2
8
302
41
2
0
3
0
15
1649
41
0
4
0
0
28
356
2
5
0
0
0
1
38
TABLE VII. CONFUSION MATRIX OF PARTICIPANT 19
Values obtained from HRV
(Stress Index)
1
2
3
4
5
Predicted
Values
(k-NN)
1
120
18
2
0
0
2
39
107
17
0
0
3
2
17
1309
135
10
4
0
0
44
142
16
5
0
0
2
1
3
CONCLUSIONS
An algorithm was developed to quantify the level of stress in
people on a 5-level scale based on EEG and EMG signals and
find its consistency with the stress index obtained by HRV
analysis. In addition, an accuracy of 92% was obtained in the
results, which is superior to many of the methods found in the
literature and with a scale of levels higher than previous
research.
It is concluded that parallel extraction of biomedical signal
parameters (EEG and EMG) increases the accuracy in
recognizing a person's stress level.
A larger database could establish a global classifier to work
with any type of person within a homogeneous group as opposed
to establishing a classifier per person.
In the future, a portable equipment developed in an embedded
system could be designed to analyze the data in real time with
the proposed method for integral neurofeedback and
biofeedback sessions in which a virtual environment is
established from the variations of the state of stress in which a
meditation practitioner is found.
REFERENCES
[1] A. R. Subhani, W. Mumtaz, M. N. B. M. Saad, N. Kamel, and A. S. Malik,
“Machine Learning Framework for the Detection of Mental Stress at
Multiple Levels,” IEEE Access, vol. 5, pp. 13545–13556, 2017.
[2] R. Richer, N. Zhao, J. Amores, B. M. Eskofier, and J. A. Paradiso, “Real-
time Mental State Recognition using a Wearable EEG,” 2018 40th Annual
International Conference of the IEEE Engineering in Medicine and Biology
Society (EMBC), 2018.
[3] J. Amores, R. Richer, N. Zhao, P. Maes, and B. M. Eskofier, “Promoting
relaxation using virtual reality, olfactory interfaces and wearable
EEG,” 2018 IEEE 15th International Conference on Wearable and
Implantable Body Sensor Networks (BSN), 2018.
[4] H. Wang, Q. Song, T. Ma, H. Cao, and Y. Sun, “Study on brain-computer
interface based on mental t asks,” 2015 IEEE International Conference on
Cyber Technology in Automation, Control, and Intelligent Systems
(CYBER), 2015.
[5] N.-H. Liu, C.-Y. Chiang, and H.-C. Chu, “Recognizing the Degree of
Human Attention Using EEG Signals from Mobile Sensors,” Sensors, vol.
13, no. 8, pp. 10273–10286, 2013.
[6] F. M. Al-Shargie, T. B. Tang, N. Badruddin, and M. Kiguchi, “Mental Stress
Quantification Using EEG Signals,” IFMBE Proceedings International
Conference for Innovation in Biomedical Engineering and Life Sciences,
pp. 15–19, 2015.
[7] F. Al-Shargie, T. B. Tang, N. Badruddin, and M. Kiguchi, “Towards
multilevel mental stress assessment using SVM with ECOC: an EEG
approach,” Medical & Biological Engineering & Computing, vol. 56, no. 1,
pp. 125–136, 2017.
[8] P. Karthikeyan, M. Murugappan, and S. Yaacob, “EMG Signal Based
Human Stress Level Classification Using Wavelet Packet
Transform,” Communications in Computer and Information Science Trends
in Intelligent Robotics, Automation, and Manufacturing, pp. 236–243, 2012.
[9] A. Ahani, H. Wahbeh, H. Nezamfar, M. Miller, D. Erdogmus, and B. Oken,
“Quantitative change of EEG and respiration signals during mindfulness
meditation,” Journal of NeuroEngineering and Rehabilitation, vol. 11, no.
1, p. 87, 2014.
[10] OpenBCI, Cyton Biosensing Board (8-channels). 2019.
[11] G. Rojas, C. Alvarez, C. Montoya, M. de la Iglesia-Vayá, J. Cisternas and
M. Gálvez, "Study of Resting-State Functional Connectivity Networks
Using EEG Electrodes Position as Seed", Frontiers in Neuroscience, vol.
12, 2018. Available: 10.3389/fnins.2018.00235.
[12] Sharma, A. and Singh, M. (2015). Assessing alpha activity in attention and
relaxed state: An EEG analysis. 2015 1st International Conference on Next
Generation Computing Technologies (NGCT).
[13] OpenBCI, Gold Cup Electrodes. 2019.
[14] I. Johnstone and B. Silverman, "Wavelet Threshold Estimators for Data with
Correlated Noise", Journal of the Royal Statistical Society: Series B
(Statistical Methodology), vol. 59, no. 2, pp. 319-351, 1997. Available:
10.1111/1467-9868.00071.
[15] G. Kemper and Y. Iano, "An Audio Compression Method Based on
Wavelets Subband Coding", IEEE Latin America Transactions, vol. 9, no.
5, pp. 610-621, 2011. Available: 10.1109/tla.2011.6030967.
[16] M. Mamun, M. Al-Kadi and M. Marufuzzaman, "Effectiveness of Wavelet
Denoising on Electroencephalogram Signals", Journal of Applied Research
and Technology, vol. 11, no. 1, pp. 156-160, 2013. Available:
10.1016/s1665-6423(13)71524-4.
[17] D. Valencia, D. Orihuela J. Salazar, and J. Valencia, " Comparison analysis
between rigrsure, sqtwolog, heursure and minimaxi techniques using hard
and soft thresholding methods", Symposium on Signal Processing, Image
and Artificial Vision, 2016. Available: 10.1109/STSIVA.2016.7743309.
[18] P. Karthikeyan, M. Murugappan and S. Yaacob, "ECG Signal Denoising
Using Wavelet Thresholding Techniques in Human Stress Assessment",
International Journal on Electrical Engineering and Informatics, vol. 4, no.
2, pp. 306-319, 2012. Available: 10.15676/ijeei.2012.4.2.9.
[19] P. Fulpatil and Y. Meshram, “Analysis of EEG Signals with the Effect of
Meditation”, International Journal of Engineering Research & Technology
(IJERT), vol. 3, no. 6, 2014.
[20] F. Shaffer and J. Ginsberg, "An Overview of Heart Rate Variability Metrics
and Norms", Frontiers in Public Health, vol. 5, 2017. Available:
10.3389/fpubh.2017.00258.
[21] R. Baevsky and A. Chernikova, “Heart rate variability analysis:
physiological foundations and main methods”, Cardiometry, pp. 66-76,
2017. Available: 10.12710/cardiometry.2017.10.6676.
[22] R. Baevsky, B. Bennett, M. Bungo, J. Charles, A. Goldberger and G.
Nikulina, “Adaptive responses of the cardiovascular system to prolonged
spaceflight conditions: assessment with Holter monitoring”, Journal of
Cardiovascular Diagnosis and Procedures, vol. 14, no. 2, pp. 53-57, 1997.
[23] R. Baevsky and A. Berseneva “Use of Kardivar System for Determination
of the Stress Level and Estimation of the Body Adaptability”, 2008.
Please cite:
Ugarte, D. E., Linares, D., Kemper, G., & Almenara, C. A. (2019). An algorithm to measure the stress level from EEG, EMG
and HRV signals. In 2019 International Conference on Information Systems and Computer Science (INCISCOS) (pp. 346–
353). Quito, Ecuador: IEEE. https://doi.org/10.1109/INCISCOS49368.2019.00061
