Information Technology Reference
In-Depth Information
hemispheres are compared to decide the activated hemisphere, which can then be
used to design a control interface.
2.7
EEG-based BCI 5 SCP BCI
SCP are low frequency potential shifts in EEG (around 1
2 Hz) and can last several
seconds. It is possible to control SCP using feedback and reinforcement mecha-
nism. Different tasks can be used to control either the positivity or negativity SCP.
For example, cognitive tasks (or even inactive relaxed states) can generate posi-
tivity SCP while negativity SCP can be generated with tasks such as readiness/
planning to move. Hence, it can be used to generate a binary signal, which can be
used as a control mechanism. It is not as popular as the other EEG-based BCIs as it
requires extensive training in order to give good performance.
-
2.8
EEG-based BCI A Brief Comparison of the Paradigms
Comparing the different EEG-based BCIs, it can be seen that each method has its
strengths and weaknesses. For example, motor imagery requires user training and
also the response time is slower (the imaginary movement causes changes in EEG
to show up typically after a few seconds) but this paradigm circumvents a visual
interface and also be can run in the asynchronous mode, thereby allowing the user
to turn the system ON/OFF and also use the control mechanism. Mental thoughts
are similar in this regard but with the brain rapidly changing over time, such EEG-
based BCIs will require frequent retraining.
SSVEP is very robust and requires only a single active channel but require users
to gaze at
flashing blocks, which is only practical for short periods of time (typically
a few minutes). There is also the risk of triggering epilepsy if the
flashing frequency
is set to be too low. P300 VEP also suffers from this risk, though of a lesser degree.
Of all the EEG-based BCIs, SCP requires the most extensive training and is less
appealing for this reason but gives good performance.
2.9
Application 1 Biometrics (Password, PIN Generation)
The common biometric is
fingerprint but in recent years, others such as DNA, hand
geometry, palm print, face (optical and infrared), iris, retina, signature, ear shape,
odour, keystroke entry pattern, gait, voice, etc., have been proposed. But all these
biometrics can be compromised at some stage but biometrics based on BCI is more
fraud resistant as thoughts cannot be forged!
The P300 BCI paradigm can be used to generate a sequence of passwords (or
personal identi
cation number, PIN) that can be used in ATMs and computer logins
(Gupta et al. 2012 ).
Instead of entering the password using a keypad,
the
 
Search WWH ::




Custom Search