![](bg2.png)
event related synchronisation (ERS). The ERD or
ERS occurs contra laterally to the intended
movements. For example, for right motor
movements, the ERD is observed at the left
hemisphere of the brain and for the left motor
movements, the ERD is found at the right
hemisphere. Classifications of right or left motor
movements are usually made based on the ERD of
the signal.
In a normal ERD detection procedure (Kalcher
and Pfurtscheller, 1995), EEG signals are usually
filtered in a narrow band, squared, low pass filtered
and averaged over trial. However, it is
disadvantageous to use this method because useful
information will be lost from the averaging over
multi-trials.
In this paper we have implemented the adaptive
recursive bandpass filter (Gharieb and Cichocki,
2001) to detect imaginary motor movements and to
classify them according to left or right movements
without observing the ERD and averaging over multi
trials. Implementation of this method reduces the
chances of information lost and increases the
classification rate. This method has been
successfully implemented on the BCI Competition
2003 dataset IIIb, which consists of 140 labelled and
unlabelled trials respectively, further information of
the data set, is explained in the next section.
The target signal, which contains mu rhythm, is
first pre-processed by implementing a band pass
filter. Then, the adaptive recursive bandpass filter is
used to estimate the dominant signal, which
represents the motor movements. The employed
adaptive recursive filter is used to trail the centre
frequency of the dominant EEG signal. The filter
requires only one coefficient to be updated in order
to adjust the centre frequency of the filter bandpass
to be approximated with that of the input signal
(Gharieb and Cichocki, 2001). The time function of
the coefficient represents the distinct feature for each
signal and represents either left or right imaginary
motor movements.
2 METHODOLOGY
2.1 Data Set
This method was trialed on the BCI Competition
2003, dataset IIIb (BCI Competition II, 2003). The data
set was provided by the Department of Medical
Informatics, Institute for Biomedical Engineering,
University of Graz. The signals were obtained from
a 25-year-old female relaxing on a chair with
armrest. The task was to control a feedback bar by
means of imagining left hand of right hand
movements. The data was acquired from the EEG
channels C3, Cz, and C4 (figure 1), which was band
pass filtered for a frequency range of 0.5 to 30Hz
and sampled at 128 Hz. The experiment consists of 7
runs with 40 trials each. All runs were conducted on
the same day with several minutes break in between.
The data has a total of 280 trials, which consists of
140 labelled and 140 unlabeled trials with an equal
number of left hand and right hand movements.
Each trial consists of duration of 9 seconds. At the
3
rd
second a visual cue, an arrow pointing left or
right is presented to indicate left or right motor
movements is to be imagined.
32
321C3 Cz C4
1
5 cm
Figure 1: Electrode positions.
0123456789
sec
Trigger
Beep
Feedback period with Cue
Figure 2: The timing scheme.
2.2 Signal Analysis
The trials were divided into two groups according to
right or left motor imaginary. Signals from channel
C3 and C4 of each group are first pre-processed by
means of band pass filtering. A band pass filter
using 7
th
order Butterworth filter where the pass
band is 9Hz with less than 1 dB of ripple and the
stop band is 11Hz with at least 6 dB of attenuation.
Signals from channel Cz is ignored because it
contains very little significant discriminative
features (Lemm et al., 2004). After band passing the
signals, we could observe that the signal is densely
SIGMAP 2008 - International Conference on Signal Processing and Multimedia Applications
114