Content uploaded by Gaurav --
Author content
All content in this area was uploaded by Gaurav -- on Mar 12, 2021
Content may be subject to copyright.
Abstract—Meditation is a mental practice to achieve focus of
mind and emotional clarity. Meditation has been used for
cognitive enhancement, rehabilitation and reducing stress and
anxiety. In the present study, we are doing a comparative
analysis between various levels of meditators based on EEG as
psychophysiological indicator; and possibility of EEG as a
neurofeedback for meditators. An analytical experiment on
three categories of subjects (A: an expert meditator, B: five
moderate meditators and C: five non-meditators) was done.
Each subject was guided to perform two visual tasks; first to sit
relaxed with eyes closed (REC) and second to gaze on a dot on
screen (RDOT); supplied, EEG being recorded in parallel. The
first subject was recorded with absorbed state of meditation
(Samādhi). For psychophysiological analysis, wavelet transform
based features from each recording of EEG was evaluated.
Topographical mapping of brain functioning based on features
were plotted and analyzed. It was observed that theta, alpha
and beta were comparatively higher for expert meditator in
frontal and central region during REC and RDOT. Also,
during absorbed meditative state, the alpha and beta are higher
at midline central region (Cz) and theta is higher at C3 and C4.
Keywords- EEG, Meditation, Neurofeedback
I. INTRODUCTION
Neurofeedback is a tool for conditioning the
psychophysiological functioning such as brain functioning
into measurable quantities which can act as a feedback for
clinical assessment. It is a fast growing field of study in the
areas of neuroscience, brain machine interface (BMI) and
cognitive psychology. In last few decades, neurofeedback has
grown into a wide field with variety of applications due to
advancement in neuronal monitoring systems. Tools like
fMRI, PET, EEG, etc. have created advancement in neuronal
monitoring, due to growth in generic use and with widely
developing methods of analysis. Neurofeedback has been
very helpful in various application such as cognitive disorder
diagnoses [1]; attention, anxiety and sleep study [2];
cognitive enhancement and therapies; and in non-clinical
applications such as emotion and mood indicators while
performing activities such as listening music, videos, dance,
Yoga, meditations and mindfulness, etc. [3] [4].
Electroencephalography (EEG) is the monitoring of
electrophysiological signal generated due to brain cortical
activity. Based on EEG rhythmic activity, viz. delta, theta,
alpha, beta and gamma; physiological and mental state of a
subject can be determined [5]. EEG has been widely used in
clinical, cognitive neuroscience and in neurofeedback
applications [6]. There have been a wide range of study done
to characterize neuronal effects of meditation using EEG and
event related potential. Meditation is a mental and
consciousness based practice, to focus on particular
representation and achieve high level of attention, focus and
emotional clarity. There have been a wide range of study
done to study neuronal effects of meditation using EEG and
event related potential [7]. EEG studies on meditation
practitioners has observed prominent increase in theta and
alpha activity during meditation [8]. Other studies have
reported an increase in the specific frequencies expressed in
the alpha range and reduction in frequency of EEG activity in
experienced meditators versus less experienced meditators
while meditating [9].
In the Buddhist and the Yogic literatures, various states of
minds are discussed while going deep into the meditative or
mindful states, termed in Sanskrit as dhyana or in Pali as
jhāna. The mind reaches such a trance absorbed state where a
high level of attended representation is activated to a highly
sustained degree and all other representations of perceptions
are deactivated. This state of high sustained attention and
pure mindfulness is termed as ‘samādhi’ in Sanskrit or
‘Samāpatti’ in Pali [10]. A study done on transcendental
meditation (TM) indicates slowing of alpha frequency and
increase in alpha amplitude during deep state of TM
(presumably the same as Samādhi), signaled by the subject
through pressing a push button [11]. Progressing to deepness
in meditation, first the alpha frequency decreases and
amplitude increases, then theta rhythm occurs intermixing
with alpha and then transcendence burst to higher beta
frequencies (20-30 Hz) [12].
In the present study, we are experimentally analyzing the
effect of meditation on brain activation. Quantitative and
comparative analysis of EEG recorded for different groups of
subjects was done. First group is of expert meditator (with
self-assessment of Samādhi), second group is of moderate
meditation practitioners and third group (control group),
having no exposure, was done. Based on task oriented EEG
evaluation and topographical mapping, the difference in
cortical activity of advanced meditator, moderate meditation
practitioners and non-practitioners was established. This
study may be used as a neurofeedback indicator for
meditation practitioner to self-assess the growth and deepness
of meditative focus reached.
II. EXPERIMENT AND DATA
A. Hardware set up and preprocessing
In the present work, EEG signal is recorded using B-
Alert X10 device. B-Alert X10 (Advanced Brain
Monitoring, Inc., Carlsbad, CA) is a wireless acquisition
system equipped with headset to get connected with one lead
ECG and nine channel EEG visually; Poz, Fz, Cz, C3, C4,
An EEG based Quantitative Analysis of Absorbed Meditative State
Gaurav G1, Ashish Kumar Sahani2, and Abhijit Sahoo3
1Department of Electrical Engineering, IIT Roorkee
2Center for Biomedical Engineering, IIT Ropar
3Department of Computer Science, IIT Madras
978-1-5386-7921-0/19/$31.00 ©2019 IEEE
9th International IEEE EMBS Conference on Neural Engineering
San Francisco, CA, USA, March 20 - 23 , 2019
57
Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY ROORKEE. Downloaded on March 12,2021 at 19:47:45 UTC from IEEE Xplore. Restrictions apply.
F3, F4, P3 and P4, following 10/20 electrode system. The
headset is responsible for analog to digital conversion, which
generally operates at 256 samples per second, and digital
encoding at 16 bit resolution. Before recording electrical
impedance between the scalp and the electrodes is checked
to remain below 30KΩ. The signal is formatted and
transmitted using a USB host device between 2.4 to 2.48
GHz radio transmitters. For recording and analysis,
Acknowledge 4.2 software of Biopac™ is used. The
recording was conducted in a silent and disturbance free
environment.
The signal while recording is prone to motion artifact
and powerline interference. A pair of electrode fixed at
mastoid position records EMG along with EEG to remove
motion artifact due to skull and scalp movement using
adaptive filtering technique. Fig. 1 depicts the schematic of
B-Alert X10 placement following 10/20 standard electrode
positions. Those segments of signal which are majorly
corrupted with motion artifact were removed. Signal was
filtered using IIR Butterworth filter of order 4 between cut-
off frequencies 0.5 Hz and 40 Hz.
B. Experimental task and subject
For this study, we took three category of subjects based
on their meditation practice and level of experience. First
group (A) has only one person, Om Swami (male, age 38
years), a Himalayan mystic with broad training and
experience of Yoga (25 years of practice in meditation and
mantras) [13]. During the experiment subject went into
Samādhi sate, which was recorded and studied in this paper.
Second group (B) have five meditative practitioners with
more than 3 years of meditation practice (all male, age: 21-
30 years). Third group (C) is control group of five subjects
(all male, age: 21-30 years) without any practice or exposure
to meditation. Written consent of each participant was taken
and prior guidance about the experiment was provided.
Each participant was asked to perform two tasks, each
of five minutes approximately. In first session, the subject
has to keep their eyes closed and sit state on a chair (rest
eyes closed, REC). In the second session, each subject is
guided to perform a task to continuously focus their visual
attention on a red colored dot shown on a dark computer
screen (restfully gazing at a dot on screen, RDOT). Each
task was performed in dark and noiseless environment.
While performing the tasks for each participants nine
channel EEG was recorded.
III. METHOD AND DATA ANALYSIS
A. Wavelet decomposition and features
For EEG analysis, frequency band features of the
recorded signal has to be evaluated. The corresponding
bands in EEG are delta (δ) between 0.1 to 4 Hz, theta (θ)
between 4 to 8 Hz, alpha (α) between 8 to 13Hz, beta (β)
between 13 to 30 Hz and gamma (γ) beyond 30 Hz,
consecutively. EEG being non-stationary and rhythmic,
discrete wavelet decomposition has been widely used for
feature extraction [14].
For the feature extraction, we use an observation
window of 2s (512 samples) to select segments pf the full
sequence. Every next segment is selected by shifting the
observation window by 256 samples (one second). Thus,
each segment has an overlap of 50% (256 samples from both
sides) shown in fig. 2. On each segment, a discrete wavelet
decomposition is performed using Daubechies 10 and 20
levels decomposition. The first five levels are selected for
time-localized EEG frequency band features (first level:
delta, second level: theta, third level: alpha, fourth level:
beta and fifth level: gamma). All these decomposed bands
are windowed with a 512 sample Kaiser window and then
point to point summed together to get absolute value of band
feature. The absolute value is normalized by subtracting
minimum value from each feature and dividing by maximum
value. This step is repeated for each segment of the EEG
sample for all nine channels.
IV. RESULT
A. Topographic mapping of wavelet features
The wavelet based EEG band features are evaluated for
all the subjects for both experimental sessions (REC and
RDOT). Taking these band features, topographical mapping
of EEG power spectrum is developed using interpolated
pixel values on a head image. These topographical
projections are analyzed to compare the cortical activity in
different bands for all subjects and both sessions. Similarity
in topographical pattern of each subject of same group was
observed and listed. Fig. 3 shows the topographical mapping
of features for task REC for one member of each group A, B
and C. Fig. 4 shows the topographical mapping of features
for task RDOT for one member of each group A, B and C.
(a) (b)
Fig. 1. (a) B-Alert X10 system scalp placement schematic and (b)
10/20 standard of electrode placement for nine channel EEG.
Fig. 2. Steps involved in generating wavelet decomposition based
EEG frequency band features.
58
Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY ROORKEE. Downloaded on March 12,2021 at 19:47:45 UTC from IEEE Xplore. Restrictions apply.
It has been evident from the fig. 3 that theta, alpha and
beta are consistently higher in frontal region (F3 and F4) for
the case of group A in comparison to group B and group C,
while performing task REC. Also in the central region (C3,
C4 and Cz) theta and alpha are relatively higher while
performing REC for group A. Similarly, it has been
observed that the frontal region (F3, F4 and Fz) and central
region (C3, C4 and Cz) has consistently high intensity theta,
alpha and beta for group A in comparison to group B and
group C, while performing RDOT, as shown in fig. 4.
In case of absorbed meditative state for group A, it has been
observed that the whole cortical activity is intense at central
region. C3 and C4 entertain high theta, whereas alpha, beta
and delta are concentrated high Cz position, as shown in fig.
5. This represents a very high adaptive attention level [15].
Along with the topographical mapping, a boxplot
distribution of the features generated from both the tasks
show the contrast between their cortical activities. Fig. 6
represents the boxplot of features for task REC. Fig. 7
represents the boxplot of features for task RDOT.
In both fig. 6 and fig. 7, the distribution of all the
features of delta, theta, alpha, beta and gamma of all nine
channels(POz, Fz, Cz, C3, C4, F3, F4, P3 and P4) are
plotted for group A, B and C and compared with each other.
It is observed that group A has most of the cortical region
very stable through-out the session REC and RDOT and
deviation from central tendency is very less. Alpha and beta
is overall higher for group A, whereas, for symmetric region,
central region and F4 region central tendency of theta, alpha
and beta are higher for group A than group B and C.
Similarly, in case of RDOT, alpha and beta in midline
region (POz, Cz and Fz) are significantly higher for group
A. Also in central (Cz, C3 and C4) and frontal (F4) region;
alpha and beta are higher. A high alpha to theta ratio is an
indicator of higher auditory and visual working memory
[16].
Fig. 3. Topographical map of the features during REC session for on
member of (a) group A, (b) group B and (c) group C
Fig. 4. Topographical map of the during RDOT session for on member
of (a) group A, (b) group B and (c) group C
Fig. 6. Boxplot of wavelet EEG band features for group A, B and C
while performing REC for all nine channels.
59
Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY ROORKEE. Downloaded on March 12,2021 at 19:47:45 UTC from IEEE Xplore. Restrictions apply.
V. CONCLUSION
A quantitative comparative study of meditation
practitioners and non-practitioners was done using EEG as a
psychophysiological indicator. Wavelet decomposition
based features from nine channel EEG was taken for the
analysis. Topographical mappings of the features were done
to observe the neuronal activation of different EEG bands
during eyes closed at rest state and then focused gazing a dot
on screen. The activation of brain cortex during REC and
RDOT task performed by three groups was evaluated and a
comparative analysis was made about the positions and EEG
frequency bands. The absorbed meditative state was
separately analyzed and various position of brain cortex
activation during absorbed state was reported. Overall, the
wavelet feature based method adopted to evaluate the brain
rhythmic activities was successfully used in creating
discrimination as per the experience of and level of
meditation of subject groups. The efficacy of the quantitative
results may be utilized as neurofeedback for self-assessment
of meditation.
REFERENCES
[1] M. Arns, S. D. Ridder, U. Strehl, M. Breteler, A. Coenen, “Efficacy of
neurofeedback treatment in ADHD: the effects on inattention,
impulsivity and hyperactivity: a meta-analysis,” Clinical EEG and
neuroscience, vol. 40, no. 3, pp.180-189, July 2009.
[2] D. C. Hammond, “Neurofeedback with anxiety and affective
disorders,” Child and Adolescent Psychiatric Clinics, vol. 14, no. 1,
pp. 105-123, Jan. 2005.
[3] J. Gruzelier, “A theory of alpha/theta neurofeedback, creative
performance enhancement, long distance functional connectivity and
psychological integration,” Cognitive processing, vo.10, no. 1, pp.
101-109, Feb, 2009.
[4] A. Chiesa, A. Serretti, “Mindfulness-based stress reduction for stress
management in healthy people: a review and meta-analysis,” The
journal of alternative and complementary medicine, vol. 15, no. 5, pp.
593-600, Nov. 2009.
[5] V. P. Perlo, “Fundamentals of Electroencephalography,” JAMA, vol.
218, no. 12, pp. 1833-1833, Dec. 1971
[6] J. R. Evans, A. Abarbanel, “Introduction to quantitative EEG and
neurofeedback,” Elsevier, May 1999 May.
[7] B. R. Cahn, J. Polich, "Meditation states and traits: EEG, ERP, and
neuroimaging studies," Psychology of Consciousness: Theory,
Research, and Practice, vol. 1, no. S, pp. 48-96, 2013.
[8] Chiesa, Alberto, and Alessandro Serretti. "A systematic review of
neurobiological and clinical features of mindfulness
meditations." Psychological medicine, vol. 40, no. 8, pp. 1239-1252,
Aug. 2010.
[9] B. Stigsby, J. C. Rodenberg, H. B. Moth, “Electroencephalographic
findings during mantra mediation (transcendental meditation). A
controlled, quantitative study of experienced
meditators,” Electroencephalography and Clinical
Neurophysiology, vol. 51, no. 4, pp. 434-442, Apr. 1981.
[10] R. Shankman, “The experience of Samadhi: An in-depth exploration
of Buddhist meditation,” Shambhala Publications; Dec. 2008.
[11] J. P. Banquet JP, “Spectral analysis of the EEG in meditation,”
Electroencephalography and clinical neurophysiology, vol. 35, no. 2,
pp. 143-51, Aug. 1973.
[12] P. B. C. Fenwick, S. Donaldson, L. Gillis, J. Bushman, G. W. Fenton,
I. Perry, C. Tilsley, H. Serafinowicz, “Metabolic and EEG changes
during transcendental meditation: an explanation,” Biological
Psychology, vo. 5, no. 2, pp. 101-118, Jun. 1977.
[13] O. Swami, “Om Swami- A mystic’s viewpoint,” Internet:
www.omswami.com [Sep. 21, 2018].
[14] W. Ting, Wu, Y. Guo-zheng, Y. Bang-hua, and S. Hong, “EEG
feature extraction based on wavelet packet decomposition for brain
computer interface,” Measurement vol. 41, no. 6, pp. 618-625, Jul.
2008.
[15] S. Sadaghiani, R. Scheeringa, K. Lehongre, B. Morillon, A. L. Giraud,
M. D'Esposito, A. Kleinschmidt, “Alpha-band phase synchrony is
related to activity in the fronto-parietal adaptive control
network,” Journal of Neuroscience, vol. 32, no. 41, pp. 14305-14310,
Oct. 2012.
[16] M. Kawasaki, K. Kitajo, and Y. Yamaguchi, “Dynamic links between
theta executive functions and alpha storage buffers in auditory and
visual working memory,” European Journal of Neuroscience, vol. 31,
no. 9, pp. 1683-1689, May 2010.
Fig. 7. Boxplot of wavelet EEG band features for group A, B and C
while performing RDOT for all nine channels.
60
Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY ROORKEE. Downloaded on March 12,2021 at 19:47:45 UTC from IEEE Xplore. Restrictions apply.
