Information Technology Reference
In-Depth Information
LS-SVM classi
er has been studied on EEG dataset publicly available by the
University of Bonn, Germany. Experimental results have been included to show the
effectiveness of the proposed method in comparison to other existing methods.
1 Introduction
Human brain is a highly complex system. The epilepsy is a common neurological
disorder of human brain. It affects at least 50 million people of the world (Ngugi
et al. 2010 ). The annual occurrence of epilepsy, 48 per 100,000 populations in
developed counteries was reported in Hirtz et al. ( 2007 ). The prevalence of epilepsy
is higher in low and middle income countries than developed countries (Thurman
et al. 2011 ). At least 50 % of the epileptic cases start developing at childhood or
adolescence (World Health Organization 2014 ). Occurrence of epilepsy can also be
noticed in elderly people, which may require special considerations in treatment
(Ramsay et al. 2004 ). If the patient with epilepsy are treated properly, then 70
80 %
of them can lead to normal lives (World Health Organization 2014 ). Therefore,
study of epilepsy is an important research area in the
-
field of the biomedical
engineering.
The electroencephalogram (EEG) signals are very useful to measure the elec-
trical activity of the human brain. The EEG signals are commonly analyzed by
experts in order to assess the states of the brain. The EEG based measures are very
helpful for diagnosis of neurological disorders specially epilepsy. Presence of
spikes in EEG signals is main indication of epileptic seizure activity in the brain
(Ray 1994 ; Mukhopadhyay and Ray 1998 ). Automatic detection of epileptic seizure
by analyzing EEG signals using advanced signal processing techniques is very
useful for diagnosis of epilepsy (Iasemidis et al. 2003 ).
The EEG recording using electrodes from the scalp is the result of
firing of
neurons within the brain (Schomer and da Silva 2005 ). The EEG signals recorded
from scalp exhibit oscillations of different range of frequencies, corresponding to
the electrical activities of the neurons in the brain. These oscillations of different
frequency ranges present in the EEG signals are related with the different states of
the functioning of human brain. A typical voltage range of the amplitude of EEG
signal for a adult human is about 10
100
V when measured from the scalp
-
μ
(Aurlien et al. 2004 ).
Conventional scalp EEG signal recording is performed by placing electrodes
over the scalp. The conductive gel or paste is utilized in order to make proper
electrical connection between scalp and electrode. The electrode cap is generally
used for recording of EEG signals. There are some standards that have been de
ned
for electrode placement over the scalp, out of which international 10
20 system
(Andrzejak et al. 2001 ) is commonly used for EEG signal recording of healthy
persons. The intracranial EEG signals obtained using intracranial electrodes which
also requires depth electrodes for recording of ictal and seizure-free EEG signals for
-
Search WWH ::




Custom Search