ABSTRACT
Effective algorithms for automatic seizure detection and prediction can have
a far-reaching impact on diagnosis and treatment of epilepsy. However, pri-
marily due to a relatively low understanding of the mechanisms underlying
the problem, most existing methods suffer from the drawback of low accuracy
which leads to higher false alarms and missed detections (Iasemidis, 2003).
Moreover, to obtain a reliable estimate of the efficacy of the epilepsy detec-
tion parameters and algorithms, they should be tested on a relatively large
number of datasets. Due to the lack of reliable standardized data, most of
the EEG analysis reported in the literature is performed on a small number
of datasets which reduces the statistical significance of the conclusions. Such
algorithms often demonstrate good accuracy for selected EEG segments but
are not robust enough to adjust to EEG variations commonly encountered in
a hospital setting.
