a comparison of eeg processing methods to improve the ...using a population of five disable- and...
TRANSCRIPT
-
A Comparison of EEG Processing Methods to
Improve the Performance of BCI
Arjon Turnip, Demi Soetraprawata, and Dwi E. Kusumandari Technical Implementation Unit for Instrumentation Development
Indonesian Institute of Sciences, Bandung, Indonesia
Email: {arjo001, demi001, esti001}@lipi.go.id
Abstract—Electroencephalogram (EEG) recordings provide
an important means of brain-computer communication, but
their classification accuracy is limited by unforeseeable
signal variations due to artifacts or recognizer-subject
feedback. In this paper, we propose a comparison of
processing method (i.e., NPCA, JADE, and SOBI) entailing
time-series EEG signals. Finally, the promising results
reported here (up to 94% average classification accuracy
and 36.4% improvement of maximum average transfer rate)
reflect the considerable potential of EEG for the continuous
classification of mental states.
Index Terms—brain computer interface (BCI), classification
accuracy, transfer rate, NPCA, JADE, SOBI,
electroencephalogram (EEG)
I. INTRODUCTION
Many people with severe motor disabilities require
alternative methods for communication and control.
Numerous studies over the past two decades show that
scalp-recorded electroencephalography (EEG) activity
can be the basis for non-muscular communication and
control systems. With production of advanced bio-
instruments for recording and amplifying the signals as
well as cheap and powerful personal computers, this
dream was realized and Brain-Computer Interface (BCI)
was developed [1], [2]. Brain activity can be measured
using EEG. By extracting specific components from
human brain activity and linking this brain activity to
specifically developed algorithms, an interface between a
computer and the users’ brain is created. Signals from the
brain are processed to extract specific features that reflect
the user’s intentions. Today there exist various techniques
by which to accomplish this [3]-[7]. The user’s brain is
now coupled to a computer or external device, which
allow communication or controlling devices directly,
without implementing any motor action.
There are various properties in EEG that can be used
as a bases for BCI such as rhythmic brain activity (i.e.,
delta, theta, alpha, and beta) [8], event-related potentials
(ERPs), event-related de-synchronization (ERD) and
event-related synchronization (ERS) [9]. In the present
Manuscript received February 20, 2013; revised March 18, 2013;
accepted April 1, 2013. This research was supported by Indonesian
Institute of Sciences, Indonesia.
study, we focused on the use of the ERP-P300 properties.
The P300 represents the unpredictable stimuli presented
in an oddball paradigm, in which low-probability targets
are mixed with high-probability ones. For this paradigm,
the subject is told to respond to a rare stimulus that occurs
randomly and infrequently among other, frequent stimuli
[8]. The presence, magnitude, topography, and time of
the response signal are often used as metrics of cognitive
function in decision making processes. In this paper, a
comparison of extraction method (i.e., NPCA, JADE, and
SOBI) entailing time-series EEG signals is proposed. In
order to examine the performance (i.e., accuracy and
transfer rate) improvements of the proposed method, a
classification using back-propagation neural networks
(BPNN) which has been well developed in the field of
speech recognition is applied.
II. METHODS
The data set used in this study was obtained from the
website of the EPFL BCI group [8]. The data have been
recorded according to the 10-20 international standard
from the 32 electrode configurations [9]. Each recorded
signal has a length of 820 samples with a sampling rate of
2048 Hz (the EEG was down-sampled from 2048 Hz to
32 Hz by selecting each 64th sample from the band pass-
filtered data). A six-choice signal paradigm was tested
using a population of five disable- and four able-bodied
subjects. The subjects were asked to count silently the
number of times a prescribed image flashed on a screen.
Four seconds after a warning tone, six different images (a
television, a telephone, a lamp, a door, a window, and a
radio) were flashed in a random order [8]. Each flash of
an image lasted for 100 ms, and for the following 300 ms
no image was flashed (i.e., the inter-stimulus interval was
400 ms). Each subject completed four recording sessions.
Each of the sessions consisted of six runs with one run for
each of the six images. Our goal is to discriminate all
possible combinations of the pairs of mental tasks from
each other using the corresponding EEG signals.
It is difficult to compare the performances of the BCI
systems, because the pertinent studies present the results
in different ways. However, in the present study, the
comparison was made based on the accuracy and the
transfer rate. The speed of a particular BCI is affected by
the trial length, that is, the time needed for one selection.
63
International Journal of Signal Processing Systems Vol. 1, No. 1 June 2013
©2013 Engineering and Technology Publishingdoi: 10.12720/ijsps.1.1.63-67
-
This time should be shortened in order to enhance a
BCI’s effectiveness in communication. The transfer rate
depends on both the speed and the accuracy of selection
and expressed as. If a trial has N possible selections and
each selection has the same probability of being the
desired selection, and if P denotes the probability that the
desired choice is actually selected, then the probability
for the remaining (undesired) selections being selected
will be (1-P)/(N-1). The bit rate (bits/trial) of each
selection can then be expressed as [10]
1
1log)1()(log)(log 222
N
PPPPNb (1)
where N is a number of possible selections of the target
and P denotes the probability that the desired choice is
actually selected. The transfer rate (bits per minute) is
equal to b multiplied by the average speed of selection S
(trial per minute, which is equal to the reciprocal of the
average time required for one selection). Therefore, based
on the data sets information, the desired output signal is
developed.
III. EXTRACTIONS AND CLASSIFICATION METHOD
A. Nonlinear Principle Component Analysis
Let x(k) represent n-dimensional vectors which
correspond to the n continuous time series from the n
EEG channels. Then xi(k) corresponds to the continuous
sensor readings from the ith EEG channel. Because
various underlying sources are summed via volume
conduction to give rise to the scalp EEG, each of the xi(k)
are assumed to be an instantaneous linear mixture of n
unknown components or sources si(k), via the unknown
mixing matrix A [7]
)()()( knkAskx (2)
where TM kxkxkxkx )](,),(),([)( 21 MR is a noisy
sensor vector of EEG signals, NMA R with entries ija
is an unknown NM x mixing matrix,
TN ksksksks )](,),(),([)( 21
NR is an unknown
source vector signals, and MT
M knknkn R)](,),([)( 1 is a vector of additive
noises. The objective of this work is to design a feed-
forward neural network and an associated adaptive
learning algorithm that enable estimation of the source
s(k) and identification of the mixing matrix A and
separating matrix W with a good tracking abilities for
time variable systems. These objective can be achieve
using the flowchart in the Fig. 1.
B. Joint Approximate Diagonalization of Eigen Matrices
Joint approximate diagonalization of eigen matrices
(JADE) is based on diagonalization of cumulant matrices.
In the case of JADE the matrix W diagonalizes F(M) for
any i.e., WF(M)WT is diagonal. Matrix F is a linear
combination of terms of the form Tiiww , w is a column of
W. Ti WMWF )( is made as diagonal as possible for
different combination of iM and ki ,,1 . The
diagonality of matrix Ti WMWFQ )( can be measured
as the sum of the squares of off-diagonal elements:
Observe n-dimentional data
vector x according to: x=As
Determining the pre-separating matrix (V)
Applying the learning rule:
Determine pre-separating vector ( ) by:
)}()({ kxkxET
i
x
)()()( kxkVkx
)]()()()[()()()1( kxkVkxkxkkVkV
Determining whitening matrix (P)
Applying the learning rule:
Determine the whitening vector (u) by:)()()( kxkPku
)()]()()[()()1( kPkukuIkkPkPT
n
2
2
E{P(k)PT(k)}=1
Yes
NoYes
No
Separate the independent components
Determining separating matrix (W)
Applying the learning rule:
)(])()([)()()()1( kWkyfkukyfkkWkW TT
Is decorrelated with)(kx
Is W converged to a
desired value?
Yes
No
Separate independent componets
according to: y(k)=W(k)u(k)
Estimate the mixing matrix (A)
Calculate the estimating matrix (Q)
Applying the learning rule:
)()]()()(ˆ)[()()1( kykykQkxkkQkQT
Observed the observed signals by:
)()()(ˆ kykQkx
Determine the correlation
coefficient of signals
Is the value
acdepted?
End
2
1
1
3
3
Figure 1. Nonlinear principal component cnalysis flowchart.
lk klq2
. Minimization of sum of squares of diagonal
elements is same as maximization of sum of squares of
diagonal elements [10], [11].
2)()( iT
iJADE WMWFdiagWJ (3)
Joint approximate diagonalization of )( iMF can be
obtained by maximizing JADEJ . Choice of the matrix
iM would be to take the eigen matrices of the cumulant.
After algebraic manipulations, the above equation
becomes
2),,,()( iiklijkl lkjiJADE yyyycumWJ (4)
when the above equation is minimized the sum of squares
of cross-cumulants of iy is also minimized.
C. Second-Order Blind Identification
The Second-Order Blind Identification (SOBI) is
applied as a blind source separation to EEG data collected
based visual stimulation with the goal of performing
classification of event-related potentials (ERPs) elicited
under different stimulation condition. SOBI uses the EEG
64
International Journal of Signal Processing Systems Vol. 1, No. 1 June 2013
©2013 Engineering and Technology Publishing
-
measurements x(k) and nothing else to generate an
unmixing matrix W that approximates A-1
in the Eq. 2,
and the vector of the estimated component values y(k) in
the Fig. 1. Sensor space projections, which indicate the
effect of a given component on all sensors are given by
the estimated mixing matrix 1ˆ WA . The SOBI
algorithm proceeds in two stages: First, the sensor signals
are zero-meaned and presphered as follows [12]
)()()( kxkxBky (5)
The angle brackets denote an average over time, so
the subtraction guarantees that y will have a mean of zero.
The matrix B is chosen so that the correlation matrix of
Tkykyy )()(, , becomes the identity matrix. This is
accomplished by moving to the PCA basis using
Ti UB 2/1diag (6)
where i are the eigenvalues of the correlation matrix,
Tkxkxkxkx )()()()( (7)
and U is the matrix whose columns are the corresponding
eigenvalues, that is, the PCA components of x. The
second stage, one constructs a set of matrices that, in the
correct separated basis, should be diagonal. A set of time
delay values s is chosen to compute symmetrized
correlation matrices between the signal y(k) and a
temporally shifted version:
TkykyR )()(sym (8)
where 2/)()(symTMMM is a function that takes
an asymetric matrix and returns a closely related
symmetric one. This symetrization discards some
information, but the problem is already highly valid,
albeit slightly weaker, constraints on the solution. The
rotation of V that jointly diagonalizes all of them is
calculated by minimazing 2
ijjiTV VR , the sum
of the squares of the off-diagonal entries of the matrix
products VRTV through an iterative process. The final
estimate of the separation matrix is
BTVW (9)
which is used to derive the separated component of y(k).
To measure the performance of algorithms, we use the
performance index (PI) as in [11], [12] defined by
n
i
n
k jij
kin
k ijj
ik
g
g
g
g
nnPI
1 11
1max
1max1
1 (10)
where G is the global transformation matrix from s to y,
gij is the (i,j) -element of the global system matrix G=HW
and maxj gij represents the maximum value among the
elements in the ith row vector of G, maxj gji does the
maximum value among the elements in the ith column
vector of G. When the perfect separation is achieved, the
performance index is zero. In practice, the values of
performance index around 10-2
gives quite a good
performance.
D. Backpropagation Neural Networks Classifier
Neural networks have been proposed in the fields of
neural sciences following research into the mechanisms
and structures of the brain. The back-propagation
algorithm allows exponential acquisition of input-output
mapping knowledge within multilayer networks. If a
pattern is submitted and its classification is determined to
be erroneous, the current least mean-square classification
error is reduced. The error is expressed as [13]
n
i
ijijjT
ijijjj wxydwxydE1
)),(()),((2
1 (11)
where jd denotes the desired output of node j
corresponding to input ix , n is the number of training
patterns and ),( ijij wxy denotes the vector output of the
networks corresponding to input ix and weight matrix
ijwW . During the association or classification phase, the trained neural network itself operates in a feed-
forward manner. The error is therefore a function of the
weights of the input and output layers. The back-
propagation algorithm is a gradient descent method
minimizing the mean square error between the actual and
target outputs of a multilayer perceptron. Using the
sigmoid nonlinearity
neti enetf
1
1)( (12)
the back-propagation algorithm consists of the following
steps: First, initialize all weights and node offsets to small
random values. Second, present continuous input vector
ix and specify desired output jd . The output vector
elements are set to zero values except for that
corresponding to the class of the current input. Third,
calculate the actual output vector y using the sigmoid
nonlinearity. Fourth, adjust the weights by
ijijij xtwtw )()1( (13)
where j is the sensitivity of node j. Fifth, repeat the
steps from the second step. A better approach is a cross-
validation technique, which stops training when the error
on a separate validation set reaches a minimum. We
observe records a vectors of EEG signals )(tx
Tm txtxtx )](,),(),([ 21 from a multiple-input/ multiple-
output nonlinear dynamical system. The objective is to
find an inverse system, termed a reconstruction system
with back-propagation neural networks (BPNN), in order
to estimate the primary input source of brain signals
65
International Journal of Signal Processing Systems Vol. 1, No. 1 June 2013
©2013 Engineering and Technology Publishing
-
Tn tstststs )](,),(),([)( 21 corresponding to particular
stimulus, which are represented by )(ty
Tn tytyty )](,),(),([ 21 .
IV. RESULTS AND DISCUSSION
In brain research, the ability to measure single-trial
ERPs is one important step toward the understanding of
how the relative timing of neuronal activity can affect
learning and how memory of a particular experience can
be encoded rapidly with a single or very few exposures.
In the present study, a BPNN classifier was used. In order
to cope with nonlinearly separable problems, additional
layers of neurons placed between the input layer and the
output neuron are needed, leading to the multilayer
perceptron architecture. Performance is measured
according to the specified performance function such as
iteration speed and signal noise to ratio (SNR) criteria
[14]. The robustness of the each algorithm (NPCA, JADE,
and SOBI) was evaluated by comparing its separation
performance as shown in Fig. 2. After 500 iteration, the
NPCA algorithm perform slightly better performance and
reach around 10-2
PI after 2500 iteration. This values
indicate that NPCA give shortest extraction time compare
to others algorithm. Fig. 3 shows typical performances of
the three algorithms discussed in this paper. At high SNR,
all tested algorithms perform very well. At low SNR, one
can observe that the NPCA gives better performance than
the other algorithms in most SNR ranges. In 0 - 4dB
range JADE is worse than the others.
0 500 1000 1500 2000 2500 3000 3500 40000
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
Number of iterations k
PI(
k)
NPCA
JADE
SOBI
Figure 2. Evolutions of PI(k) of the NPCA, JADE, and SOBI algorithms.
The data sets for subject 5 were not included in the
simulation since the subject misunderstood the
instructions given before the experiment. Comparative
plots of the classification accuracies and transfer rates
(obtained with the BPNN classifier method and averaged
over eight electrode configurations) for the disable- (S1 -
S4) and able-bodied subjects (S6 - S9) are depicted in Fig.
4 and Fig. 5, respectively. All of the subjects, except for
subjects 6 and 9, achieved an average classification
accuracy of 100% after ten blocks of stimulus
presentations were averaged (i.e., 24.5 s). The reason for
the poorer performance of subject 9 might be fatigue.
Subject 6 reported that he accidentally concentrated on
the wrong stimulus during one run in session 1 [8].
Shown alongside the classification accuracies using
BPNN for all of the subjects, in Table I, are the
corresponding 92% confidence intervals which the
classified based NPCA extraction gives the best
accuracies and transfer rates with smallest standard
deviations
-4 0 4 8 12 16 20 24 28 32 36-10
2
-101
-100
SNR [dB]P
erf
orm
ance I
ndex
NPCA
JADE
SOBI
Figure 3. Comparison of performance index of the NPCA, JADE, and SOBI algorithms as a function of signal to noise ratio (SNR).
0 5 10 15 20 25 30 35 40 45 500
10
20
30
40
50
60
70
80
90
100
Accura
cy (
%)
Time (s)
0 5 10 15 20 25 30 35 40 45 500
5
10
15
20
25
30
35
40
45
50
0 5 10 15 20 25 30 35 40 45 500
5
10
15
20
25
30
35
40
45
50
0 5 10 15 20 25 30 35 40 45 500
5
10
15
20
25
30
35
40
45
50
Subject 3
0 5 10 15 20 25 30 35 40 45 500
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25 30 35 40 45 500
5
10
15
20
25
30
35
40
45
50
0 5 10 15 20 25 30 35 40 45 500
5
10
15
20
25
30
35
40
45
50
0 5 10 15 20 25 30 35 40 45 500
5
10
15
20
25
30
35
40
45
50
Tra
nsfe
r ra
te (
bits/m
in)
Subject 2
0 5 10 15 20 25 30 35 40 45 500
10
20
30
40
50
60
70
80
90
100
Accura
cy (
%)
Time (s)
0 5 10 15 20 25 30 35 40 45 500 5 10 15 20 25 30 35 40 45 500 5 10 15 20 25 30 35 40 45 50
Subject 1
Accuracy
Transfer rate
0 5 10 15 20 25 30 35 40 45 500
10
20
30
40
50
60
70
80
90
100
Time (s)
0 5 10 15 20 25 30 35 40 45 500
5
10
15
20
25
30
35
40
45
50
0 5 10 15 20 25 30 35 40 45 500
5
10
15
20
25
30
35
40
45
50
0 5 10 15 20 25 30 35 40 45 500
5
10
15
20
25
30
35
40
45
50
Tra
nsfe
r ra
te (
bits/m
in)
NPCA
SOBI
JADE
Subject 4
Accuracy
Transfer rate
Figure 4. Comparison of classification accuracy and transfer rate plots (averaged over eight electrode configurations) obtained with
BPNN classifier for disabled subjects.
0 5 10 15 20 25 30 35 40 45 500
10
20
30
40
50
60
70
80
90
100
Accura
cy (
%)
Time (s)
0 5 10 15 20 25 30 35 40 45 500 5 10 15 20 25 30 35 40 45 500 5 10 15 20 25 30 35 40 45 50
Subject 8
0 5 10 15 20 25 30 35 40 45 500
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25 30 35 40 45 500
5
10
15
20
25
30
35
40
45
50
0 5 10 15 20 25 30 35 40 45 500
5
10
15
20
25
30
35
40
45
50
0 5 10 15 20 25 30 35 40 45 500
5
10
15
20
25
30
35
40
45
50
Tra
nsfe
r ra
te (
bits/m
in)
Subject 7
0 5 10 15 20 25 30 35 40 45 500
10
20
30
40
50
60
70
80
90
100
Accura
cy (
%)
0 5 10 15 20 25 30 35 40 45 500 5 10 15 20 25 30 35 40 45 500 5 10 15 20 25 30 35 40 45 50
Subject 6
0 5 10 15 20 25 30 35 40 45 500
10
20
30
40
50
60
70
80
90
100
Time (s)
0 5 10 15 20 25 30 35 40 45 500
5
10
15
20
25
30
35
40
45
50
0 5 10 15 20 25 30 35 40 45 500
5
10
15
20
25
30
35
40
45
50
0 5 10 15 20 25 30 35 40 45 500
5
10
15
20
25
30
35
40
45
50
Tra
nsfe
r ra
te (
bits/m
in)
NPCA
SOBI
JADE
Subject 9
Accuracy
Transfer rate
Figure 5. Comparison of classification accuracy and transfer rate plots (averaged over eight electrode configurations) obtained with
BPNN classifier for able-bodied subjects.
66
International Journal of Signal Processing Systems Vol. 1, No. 1 June 2013
©2013 Engineering and Technology Publishing
-
TABLE I. AVERAGE CLASSIFICATION ACCURACY (%).
Subject SOBI JADE NPCA
S1 94.00 93.00 95.75
S2 92.25 94.5 94.00
S3 94.00 95.5 95.00
S4 93.00 94.25 94.50
S6 90.50 91.85 92.55
S7 94.75 94.00 96.20
S8 94.75 96.00 97.75
S9 93.70 92.95 94.00
Average (S1–S4) 93.30.8 94.31.0 94.50.7
Average (S6-S9) 93.42.0 93.71.7 94.00.5
Average (all) 93.41.4 94.01.3 94.30.6
V. CONCLUSION
The results presented in this study show that compared
with the JADE and SOBI algorithms, a better extraction
result can be obtained when using the NPCA algorithm
for single-trial ERPs based on the P300 component from
specific brain regions. With NPCA extraction, the data
indicate that a P300-based BCI system can communicate
at the rate around 25.4 bits/min and 36.5 bits/min for the
disable- and able-bodied subjects, respectively. The
average of 100% classification accuracy is achieved after
nine blocks (average) for disabled subjects and after
fourteen blocks (average) for able-bodied subjects. These
results indicate that the system allowed several disabled
users to achieve transfer rates significantly beyond those
reported previously in the literature.
ACKNOWLEDGMENT
This research was supported by the tematic program
(No. 3425.001.013) through the Bandung Technical
Management Unit for Instrumentation Development
(Deputy for Scientific Services) and the competitive
program (No. 079.01.06.044) through the Research
Center for Metalurgy (Deputy for Earth Sciences) funded
by Indonesian Institute of Sciences, Indonesia.
REFERENCES
[1] F. Aloise, F. Schettini, P. Arico, F. Leotta, S. Salinari, D. Mattia, F. Babiloni, and F. Cincotti, “P300-based brain-computer
interface for environmental control: An asynchronous approach,” Journal of Neural Engineering, vol. 8, no. 2, pp. 025025, 2011.
[2] A. Belitski, J. Farquhar, and P. Desain, “P300 audio-visual speller,” Journal of Neural Engineering, vol. 8, no. 2, pp. 025022,
2011.
[3] H. Cecotti, and A. Graser, “Convolutional neural networks for P300 detection with application to brain-computer interfaces,”
IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 33, no. 3, pp. 433-445, 2011.
[4] Z.-G. Che, T.-A. Chiang, and Z.-H. Che, “Feed-forward neural networks training: A comparison between genetic algorithm and back-propagation learning algorithm,” Int. J. Innov. Comp. Inf.
Control, vol. 7, no. 10, pp. 5839-5851, 2011. [5] J. Escudero, R. Hornero, D. Abasolo, and A. Fernandez,
“Quantitative evaluation of artifact removal in real
magnetoencephalogram signals with blind source separation,” Annals of Biomedical Engineering, vol. 39, no. 8, pp. 2274-2286,
2011.
[6] A. Turnip and K.-S. Hong, “Classifying mental activities from EEG-P300 signals using adaptive neural network,” Int. J. Innov.
Comp. Inf. Control, vol. 8, no. 9, pp, 6429-6443, 2012.
[7] A. Turnip, K.-S. Hong, and M.-Y. Jeong, “Real-time feature extraction of P300 component using adaptive nonlinear principal
component analysis,” BioMedical Engineering OnLine, vol.10:83, 2011.
[8] U. Hoffmann, J.-M. Vesin, and T. Ebrahimi, “An efficient P300-based brain–computer interface for disabled subjects,” Journal of Neuroscience Methods, vol. 167, no. 1, pp. 115-125, 2008.
[9] E. W. Sellers, D. J. Krusienski, D. J. McFarland, T. M. Vaughan, and J. R. Wolpaw, “A P300 event-related potential brain-
computer interface (BCI): The effects of matrix size and inter
stimulus interval on performance,” Biological Psychology, vol. 73, no. 3, pp. 242-252, 2006.
[10] A. Hyvarinen, J. Karhunen, E. Oja, Independent Component Analysis, John Wiley & Sons, Inc, 2001.
[11] A. Cichocki and S. Amari. Adaptive Blind Signal and Image Processing. New York, USA: Wiley, 2002, pp. 161-162.
[12] A. Belouchrani, K. Abed-Meraim, J. F. Cardoso, and E. Moulines, “A blind source separation technique using second-order statistics,” IEEE Trans. Signal Processing, vol. 45, no. 2, pp.
434-444, 1997.
[13] Y. Yang, S. S. Ge, T. H. Lee, and C. Wang, “Facial expression recognition and tracking for intelligent human-robot interaction,”
Intelligent Service Robotics, vol. 1, no. 2, pp. 143-157, 2008. [14] S. Choi and A. Cichocki, “Blind separation of nonstationary
sources in noisy mixtures,” Electronics Letters, vol. 37, no. 1, pp.
61-62, 2001.
Arjon Turnip received the B.Eng. and M.Eng. degrees
in Engineering Physics from the Institute of Technology Bandung (ITB), Indonesia, in 1998 and 2003,
respectively, and the Ph.D. degree in Mechanical Engineering from Pusan National University, Busan,
Korea, under the World Class University program in
2012. He is currently work in the Technical Implementation Unit for Instrumentation Development, Indonesian Institute of Sciences,
Indonesia as a research coordinator. He received Student Travel Grand Award for the best paper from ICROS-SICE International Joint
Conference 2009, Certificate of commendation: Superior performance
in research and active participation for BK21 program from Korean government 2010, and JMST Contribution Award for most citations of
JMST papers 2011. His research areas are integrated vehicle control, adaptive control, nonlinear systems theory, estimation theory, signal
processing, brain engineering, and brain-computer interface.
Demi Soetraprawata received the Bachelor degree in
Engineering Physics from National University of Jakarta (UNAS) and the Master degree in Instrumentation and
Control from Institute of Technology Bandung (ITB),
Indonesia. He is currently work in the Technical Implementation Unit for Instrumentation Development
(UPT BPI), Indonesian Institute of Sciences (LIPI), Indonesia as a Chairman. Several activities that he had attended related to the
instrumentation are about analytical instrumentation at Queensland
University of Technology (QUT) and medical instrumentation at the Royal Brisbane Hospital, in 1990 and 1992, respectively, Brisbane,
Australia. His research areas are control engineering, instrument technology, engineering measurements, and instrumentation for
calibration and metrology.
Dwi Esti Kusumandari received the B.Eng. degrees in Engineering Physics from the Institute of Technology
Bandung (ITB), Indonesia, in 1999, and the M.Eng.
degree in Biomedical Engineering from the same institute in 2006. She is currently work in the Technical
Implementation Unit for Instrumentation Development, Indonesian Institute of Sciences, Indonesia as a junior
researcher. Her research interest are in biomedical engineering, image
processing, signal processing, and instrumentation.
67
International Journal of Signal Processing Systems Vol. 1, No. 1 June 2013
©2013 Engineering and Technology Publishing
http://www.springer.com/engineering/robotics/journal/11370http://www.springerlink.com/content/lh58213r4733/?p=ab56e1c5f4fb423e9b404bc674ed6a19&pi=0