Epilepsy Detection and Monitoring - Quantitative EEG Analysis Methods and Clinical Applications - page 154

EEG data recorded from particular areas in the neocortex and hippocampus and has
been unsuccessful for other areas. Unfortunately, these areas can vary from seizure
to seizure even in the same patient. The method is therefore very sensitive to the elec-
trode sites chosen. However, when the correct sites were chosen, the preictal transi-
tion was seen in more than 91% of the seizures. On average, this led to a prediction
rate of 80.77% and an average warning time of 63 minutes [28]. Sadly, this method
has been plagued by problems related to finding the critical electrode sites because
their predictive capacity changes from seizure to seizure.

6.5 Multivariate Measures

Multivariate measures take more than one channel of EEG into account simulta-
neously. This is used to consider the interactions between the channels and how they
correlate rather than looking at channels individually. This is useful if there is some
interaction (e.g., synchronization) between different regions of the brain leading up
to a seizure. Of the techniques discussed in the following sections, the simple syn-
chronization measure and the lag synchronization measure fall under a subset of the
multivariate measures, known as bivariate measures. Bivariate measures only con-
sider two channels at a time and define how those two channels correlate. The
remaining metrics take every EEG channel into account simultaneously. They do
this by using a dimensionality reduction technique called principal component anal-
ysis (PCA). PCA takes a dataset in a multidimensional space and linearly transforms
the original dataset to a lower dimensional space using the most prominent dimen-
sions from the original dataset. PCA is used as a seizure detection technique itself
[36]. It is also used as a tool to extract the most important dimensions from a data
matrix containing pairwise correlation information for all EEG channels, as is the
case with the correlation structure.

6.5.1 Simple Synchronization Measure
Several studies have shown that areas of the brain synchronize with one other during
certain events. During seizures abnormally large amounts of highly synchronous
activity are seen, and it has been suggested this activity may begin hours before the
initiation of a seizure.
One multivariate method that has been used to calculate the synchronization
between two EEG channels is a technique suggested by Quiroga et al. [37]. It first
defines certain “events” for a pair of signals. Once the events have been defined in
the signals, this method then counts the number of times the events in the two signals
occur within a specified amount of time ( ) of each other [37]. It then divides this

count by a normalizing term equivalent to the maximum number of events that
could be synchronized in the signals.
For two discrete EEG channels x i and y i , i

τ

1, …, N , where N is the number of
points making up the EEG signal for the segment considered, event times are defined
to be t i x and t i y ( i

=

1, …, m y ). An event can be defined to be anything;
however, events should be chosen so that the events appear simultaneously across
the signals when they are synchronized. Quiroga et al. [37] define an event to be a

= 1, … , m x ; j =