Dry andWater-Based EEG Electrodes in SSVEP-Based BCI Applications - Biomedical Engineering Systems and Technologies - page 3

1. Cumbersome and inconvenient procedures to prepare the user before BCI opera-
tion, low comfort during the BCI operation, and issues in detaching the system at
the end of the usage.
2. Lower accuracy of the BCI command classification algorithms, especially when
deployed outside lab conditions, leading to a lower information transfer rate (ITR).
3. Long time required for the user to adapt and learn to use the BCI, including the
time required for the BCI to learn specific user parameters, i.e., long calibration
procedure.
4. Unpleasant and intrusive interaction with a BCI system that results in users being
aversive to the use of BCIs.
5. High number of users that cannot learn to use the BCI, i.e., a so-called BCI
illiteracy.
While the first problem is common to all BCIs, except that the number and positioning
of electrodes used in a particular system can differ, the latter ones have different impact
depending on the BCI modality. In addressing these problems, we consider the SSVEP
BCI as a promising solution because, when compared to other BCIs, it can provide high
level of detection accuracy (i.e., high ITR), requires short calibration time, and has low
BCI illiteracy [2, 3].
The steady state visual evoked potential refers to the response of the cerebral cortex
to a repetitive visual stimulus (RVS) oscillating at a constant stimulation frequency. The
SSVEP manifests as peaks at the stimulation frequency and/or harmonics in the power
spectral density (PSD) of EEG signals [4]. Because of their proximity to the primary
visual cortex, the occipital EEG sites exhibit a stronger SSVEP response. SSVEP based
BCIs operate by presenting the user with a set of repetitive visual stimuli (RVSi). In
most of current implementations, the RVSi are distinguished from each other by their
stimulation frequency [5-8]. The SSVEP corresponding to the RVS receiving user's
attention is more prominent and can be detected in the ongoing (i.e., background) EEG.
Each RVS is associated with an action or a command which is executed by the BCI
when the corresponding SSVEP response is detected.
The majority of current SSVEP-based BCIs use stimulation frequencies in the 4 to
30 Hz frequency range [9]. RVS at these frequencies, as compared to higher frequencies,
have several disadvantages that underpin the fourth problem defined previously: they
are prone to visual fatigue which decreases the SSVEP strength, they entail a higher
risk of photic induced epileptic seizure [10], and they overlap with the frequency bands
of spontaneous brain activity. Higher stimulation frequencies are, thus, preferable for
the sake of safety and comfort of the BCI user [3].
The major aspect addressed in this paper is the first problem of the inconvenient
and uncomfortable preparation, usage, and detachment of a BCI. This problem stems
from the fact that EEG recording procedures remained very similar to that of the early
EEG days. The EEG is recorded using Ag/AgCl electrodes that are in contact with the
scalp through electrolytic gel [11]. The electrolyte serves two purposes, i) it bridges
the ionic current flow from the scalp and the electron flow in the Ag/AgCl electrode
and ii) 'glues' the electrode to the scalp. To further improve signal quality, the scalp is
frequently cleaned and, especially in clinical applications, skin on the scalp is abraded.