Systems, Man & Cybernetics - October 2016 - 12

EEG

Table 3. Classification of movement types of the same limb.
Study

Movement Task

Feature/Technique

Result (Classification Accuracy)

[77]

Imagined fast and slow wrist
extension and rotation

Rebound rate of MRCP and the mu
and beta band powers

Average misclassification rate of
21% (binary)

[78]

Elbow versus shoulder
torque intention

Classifier-enhanced TF synthesized
spatial pattern

(92%, 75%) and (100%, >80%) accuracy
for (healthy, stroke) subjects

[80]

Motor imagery patterns for
finger and wrist

Mahalanobis distance clustering
and artificial neural networks

65% and 71%

[81]

Imagined wrist flexion and
extension of wrists

Elman's neural networks

63%

[82]

Self-initiated seven
movement tasks

Bayesian classifier

62.9% (30.2% baseline)

[83]

Rest, imaginary grasp, and
elbow movements

CSP, filter bank CSP, etc.

66.9% (grasp versus elbow)
and 60.7% (three-class)

[84], [85]

Finger movement

Spectral principal component
analysis, band powers, and direct
temporal data

77.11% over each finger pairs
91.28% (epileptic patients)

[86]

Nine index finger position
during key-pressing

Random forest classifier

12.29% (11.1% chance level)

[87]

Real and imagined movements using fingers

Symbolic regression-based features

45% (SVM) and 38% (artificial neural
networks)

and continuous motor control to an interfaced device
directly from brain signals. In this section, a few studies
that demonstrate noninvasive BCI systems used to control movement of an interfaced device are listed. It is
interesting to note that, in all these studies, continuous
control is achieved using discrete motor tasks or combinations of motor tasks, using specific control strategies.
The applicability of movement kinematics decoding to
achieve higher-dimensional motor control, as an alternative for this strategy, can thus be investigated.
In [78], robotic arm control to execute a multistep
grasping task was achieved using EEG and was demonstrated both in healthy and poststroke subjects. Discrete
motor tasks (imagine left/right hand open-close, imagine
open-close of both hands, imagine tap foot) were performed, and control commands (move right-left, vertical,
open/close robot hand) were provided for the robot. In [79],
a virtual helicopter control was achieved using EEG. The
(left, right, up, down) controls were provided by imagining
movement of left hand, right hand, both hands up, and rest,
respectively. In [80], an EEG-based asynchronous BCI system was developed that allowed driving a car in a 3-D virtual reality environment. The imagined hand and feet
movements were used to determine the direction of the
steering wheel and car speed. Similarly, in [81] and [82],
continuous 3-D control was achieved using EEG in a virtual helicopter and quadcopter. The SMR-based features
were used in [81] to achieve helicopter control and features
from selected electrodes and frequency bins were used in
[82] to achieve quad copter control.
12

IEEE SyStEmS, man, & CybErnEtICS magazInE October 2016

Limitations and Future Scope
The noninvasive BCI research discussed in this review
demonstrated the potentials of scalp-recorded signals to
provide higher-dimensional motor control. In this regard,
we have provided a comprehensive review of recent studies using noninvasive BCI for classification and decoding
of hand-movement parameters and classification of movement types performed by localized areas of a single limb.
SMR-BCI that aims to obtain precise motor control using
relevant features from spectral, temporal, and spatially
localized neural signals recorded by EEG, fNIRS, and
MEG were reported in these studies. Most studies report
the increase in BCI performance to decode (classify or
reconstruct) movement kinematics when low-frequency
features (<6-7 Hz) are considered [21]-[23], [29], [31], [33],
[35], [37], [44], [48], [50]-[60], [88]. Further, the contribution
of parietal and supplementary motor areas along with primary motor cortex is also reported [23], [27]-[31], [46]-
[48], [56]. These results introduce novel research areas for
BCI research that will look into low-frequency signals and
the regions outside motor area and how these correlate
with the human motor activity.
It will be interesting to look into the temporal variations in connectivity between regions of the brain for different spectral regions and how it correlates to movement
intention, planning, and execution. Research in these
directions can help to better understand the neurophysiology of human locomotion and thus contribute to SMR-BCI
systems. The results indicate the feasibility of such BCI
systems to impart higher degrees of freedom of movement



Table of Contents for the Digital Edition of Systems, Man & Cybernetics - October 2016

Systems, Man & Cybernetics - October 2016 - Cover1
Systems, Man & Cybernetics - October 2016 - Cover2
Systems, Man & Cybernetics - October 2016 - 1
Systems, Man & Cybernetics - October 2016 - 2
Systems, Man & Cybernetics - October 2016 - 3
Systems, Man & Cybernetics - October 2016 - 4
Systems, Man & Cybernetics - October 2016 - 5
Systems, Man & Cybernetics - October 2016 - 6
Systems, Man & Cybernetics - October 2016 - 7
Systems, Man & Cybernetics - October 2016 - 8
Systems, Man & Cybernetics - October 2016 - 9
Systems, Man & Cybernetics - October 2016 - 10
Systems, Man & Cybernetics - October 2016 - 11
Systems, Man & Cybernetics - October 2016 - 12
Systems, Man & Cybernetics - October 2016 - 13
Systems, Man & Cybernetics - October 2016 - 14
Systems, Man & Cybernetics - October 2016 - 15
Systems, Man & Cybernetics - October 2016 - 16
Systems, Man & Cybernetics - October 2016 - 17
Systems, Man & Cybernetics - October 2016 - 18
Systems, Man & Cybernetics - October 2016 - 19
Systems, Man & Cybernetics - October 2016 - 20
Systems, Man & Cybernetics - October 2016 - 21
Systems, Man & Cybernetics - October 2016 - 22
Systems, Man & Cybernetics - October 2016 - 23
Systems, Man & Cybernetics - October 2016 - 24
Systems, Man & Cybernetics - October 2016 - 25
Systems, Man & Cybernetics - October 2016 - 26
Systems, Man & Cybernetics - October 2016 - 27
Systems, Man & Cybernetics - October 2016 - 28
Systems, Man & Cybernetics - October 2016 - 29
Systems, Man & Cybernetics - October 2016 - 30
Systems, Man & Cybernetics - October 2016 - 31
Systems, Man & Cybernetics - October 2016 - 32
Systems, Man & Cybernetics - October 2016 - Cover3
Systems, Man & Cybernetics - October 2016 - Cover4
https://www.nxtbook.com/nxtbooks/ieee/smc_202110
https://www.nxtbook.com/nxtbooks/ieee/smc_202107
https://www.nxtbook.com/nxtbooks/ieee/smc_202104
https://www.nxtbook.com/nxtbooks/ieee/smc_202101
https://www.nxtbook.com/nxtbooks/ieee/smc_202010
https://www.nxtbook.com/nxtbooks/ieee/smc_202007
https://www.nxtbook.com/nxtbooks/ieee/smc_202004
https://www.nxtbook.com/nxtbooks/ieee/smc_202001
https://www.nxtbook.com/nxtbooks/ieee/smc_201910
https://www.nxtbook.com/nxtbooks/ieee/smc_201907
https://www.nxtbook.com/nxtbooks/ieee/smc_201904
https://www.nxtbook.com/nxtbooks/ieee/smc_201901
https://www.nxtbook.com/nxtbooks/ieee/smc_201810
https://www.nxtbook.com/nxtbooks/ieee/smc_201807
https://www.nxtbook.com/nxtbooks/ieee/smc_201804
https://www.nxtbook.com/nxtbooks/ieee/smc_201801
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_1017
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0717
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0417
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0117
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_1016
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0716
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0416
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0116
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_1015
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0715
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0415
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0115
https://www.nxtbookmedia.com