Biomedical Engineering Reference
In-Depth Information
the central visual field. For these two reasons, different SSVEP patterns can be pro-
duced by gazing at one of a number of frequency-coded stimuli. This is the basic
principle of an SSVEP-based BCI.
As shown in Figure 8.2, in a typical SSVEP-based BCI setup, 12 virtual keyboard
buttons appear on a screen and flash at different frequency, while the user gazes at a
button labeled with the desired number/letter. The system determines the frequency
of the SSVEP over visual cortex by means of spectral analysis and looks up the pre-
defined table to decide which number/letter the user wants to select. In the example
paradigm shown in Figure 8.2, when the BCI user directs his attention or gaze at the
digit button “1” flashing at 13 Hz, a 13-Hz rhythmic component will appear in the
EEG signal recorded over the occipital area of scalp, and can be detected by proper
spectral analysis. Thus, the predefined command “1” will be executed. Although
other flashing buttons may cause interference, because of the central magnification
effect, 13-Hz components are very likely to dominate the power spectrum, com-
pared with the flashing frequencies of other buttons. In this paradigm, the rhythmic
SSVEP is modulated by the BCI user's gaze direction (attention) and the conveyed
information is encoded in the frequency contents of occipital EEG.
With careful optimization of the system, an average information transfer rate
(ITR) of more than 40 bits per second can be achieved [30, 31], which is relatively
higher than most other BCI paradigms [2]. Besides a high information transfer rate,
the recognized advantages of SSVEP-based BCI include easy system configuration,
little user training, and robustness of system performance. This is the reason why it
has received remarkably increased attention in BCI research [14, 15, 28-35].
10
Virtual number pad
5
13 Hz
0
1 5
4.5 cm
2.7 cm
0
0.5
1
1.5
2
2.5
3
3.5
4
Time(s)
4
3
15 Hz
2
8cm
1
0
10
12
14
16
18
20
Frequency (Hz)
10
5
0
1 5
0
0.5
1
1.5
2
2.5
3
3.5
4
Time(s)
6
4
2
0
10
12
14
16
18
20
Frequency (Hz)
Figure 8.2
Principle of SSVEP-based BCI.
Search WWH ::




Custom Search