Computational Intelligence - August 2016 - 64

The error cancellation property of group work makes
the simultaneous mode much more efficient than the
individual mode.
the individual mode because she "was
always in control of the process".
Another ALS participant said he
"couldn't believe in his teammate and
would rather perform the task alone".
These comments surfaced some very
important research question: How can
C-BCI system provide users with
motor disabilities a greater sense of
control? What training paradigm could
bolster mutual trust? The result proves
that the collaborative BCI is workable
and more efficient than individual
mode. It indicates the feasibility of
developing more collaborative applications for those with motor disabilities in
order to improve their quality of life.
B. Effects of Luminance Contrast

Results showed that luminance contrast
had a significant main effect on task performance and brain activity. Participants
had a significantly higher accuracy, a
shorter completion time, a bigger ITR
and a bigger spectral power in high
luminance contrast compared to low
luminance contrast. Nam et al. [37] and
Li, Bahn, Nam & Lee [57] have obtained
similar results in a P300-based BCI.
One explanation is that participants
could perceive more intense stimuli in
high luminance contrast when all other
environmental f actor s remained
unchanged. High luminance contrast
could result in brighter, more noticeable
visual stimuli compared to low luminance
contrast, even when the visual stimuli
were the same luminance. From visual
sensory theory, rods will function more in
high luminance contrast and make green
and blue LED lights appear brighter [58].
Since more green and blue were used
than red, high luminance contrast produced brighter visual stimuli. As revealed
by Campbell & Maffei [35], the amplitude of evoked potentials was linearly
increasing as the logarithm of the contrast
of the stimulus. The spectral power in
high luminance contrast was bigger than

that in low luminance contrast. Meanwhile, participants could stay more alert
and have better focus in high luminance
contrast. Li et al. [57] showed that participants can lose focus when using BCIs.
Higher luminance contrast may help participants to remain more focused so as to
produce better task performance.
User preference also supports the popularity of the high luminance contrast.
Results showed that significantly more
participants preferred high luminance
contrast. About 85.7% ALS participants
and 78.6% able-bodied participants preferred high luminance contrast. Most participants commented that they could
"catch" the stimuli more easily in high
luminance contrast condition. Some participants said that it was easier to focus on
the visual stimulus in high luminance
contrast, while it was difficult in low
luminance contrast because visual stimuli
did not provide enough intensity and
required too much concentration. On the
other hand, some participants disliked
high luminance contrast. One participant
said that in high luminance contrast the
visual stimuli were too bright while low
luminance contrast was more user-friendly and allowed for longer durations of use.
Another participant commented that high
luminance contrast produced some sort of
"burn" effect in the eyes that worsened
with time. These comments advanced the
following research questions: What level
of luminance contrast is most user-friendly for long-term use? Is there a luminance
contrast threshold? Further research is
needed to answer those questions.
C. Effects of Motor Disability

It revealed no significant main effect of
motor disability nor any interaction
effect between motor disability and
other factors. ALS participants exhibited
comparable task performance to the
able-bodied participants in all three
modes, which differs from the previous
research in [2], [3], [40], [41]. Li et al. [2]

IEEE Computational intelligence magazine | august 2016

and Volosyak et al. [4] found that ablebodied participants had better task performance than participants with severe
motor disabilities.
In an SSVEP-based BCI, participants
need only focus on the visual stimulus,
which brings into question whether ALS
participants' motor disabilities could
affect the quality and duration of focus.
Li et al. [2] argued that participants with
motor disability become easily fatigued
in shorter duration of focus, which
affected their performance in comparison to able-bodied participants. Piccione
et al. [59] also found that participants
with motor disabilities were unable to
avoid head and/or eye movement, which
are known to produce artifacts, disturb
classification and harm performance.
However, the hypothesized "fatigue"
from Li et al. [57] was not observed in
this study. No ALS participants requested
extra rest during the experiment, indicating that their motor disability didn't result
in extra fatigue. Also, the hypothesized
"head or eye movement" by Piccione et
al. [59] was not observed in the experiment. ALS participants not utilizing
wheelchairs (6 in the study) retained
most motor abilities and could control
head or eye movement as good as any
able-bodied participant. ALS participants
utilizing wheelchairs (8 in the study)
were unable to move and rested their
heads against the back of the wheelchair.
In this sense, it is possible to argue that
SSVEP-based BCI usage is unaffected by
motor disability. Most BCI researchers
also hold this paradigm [1], [3], [4],
although this postulation has not been
significantly supported with a large sample size. Therefore, the empirical data in
this study should provide greater substantial support in favor of such a paradigm.
Results of this study indicated that
ALS patients could use SSVEP-based
BCI as accurately and efficiently as ablebodied people. It upholds the potential
of BCI as an alternative method of
communication and control for those
with motor disabilities in communication and control. It also sheds light on
the future research and development of
BCI. However, it should be noted that
this study didn't consider the different

Table of Contents for the Digital Edition of Computational Intelligence - August 2016