Bivariable Analysis of EEG Signals - Quantitative EEG Analysis Methods and Clinical Applications - page 109

CHAPTER 4
Bivariable Analysis of EEG Signals
Rodrigo Quian Quiroga
The chapters thus far have described quantitative tools that can be used to extract
information from single EEG channels. In this chapter we describe measures of syn-
chronization between different EEG recordings sites. The concept of synchroniza-
tion goes back to the observation of the interaction between two pendulum clocks
by the Dutch physicist Christiaan Huygens in the seventeenth century. Since the
times of Huygens, the phenomenon of synchronization has been largely studied,
especially for the case of oscillatory systems [1].
Before getting into technical details of how to measure synchronization, we first
consider why it is important to measure synchronization between EEG channels.
There are several reasons. First, synchronization measures can let us assess the level
of functional connectivity between two areas. It should be stressed that functional
connectivity is not necessarily the same as anatomical connectivity, since anatomi-
cal connections between two areas may be active only in some particular situa-
tions—and the general interest in neuroscience is to find out which situations lead to
these connectivity patterns. Second, synchronization may have clinical relevance for
the identification of different brain states or pathological activities. In particular, it
is well established that epilepsy involves an abnormal synchronization of brain
areas [2]. Third, related to the issue of functional connectivity, synchronization
measures may show communication between different brain areas. This may be
important to establish how information is transmitted across the brain or to find out
how neurons in different areas interact to give rise to full percepts and behavior. In
particular, it has been argued that perception involves massive parallel processing of
distant brain areas, and the binding of different features into a single percept is
achieved through the interaction of these areas [3, 4].
Even if outside the scope of this topic, it is worth mentioning that perhaps the
most interesting use of synchronization measures in neuroscience is to study how
neurons encode information. There are basically two views. On the one hand, neu-
rons may transmit information through precise synchronous firing; on the other
hand, the only relevant information of the neuronal firing may be the average firing
rate. Note that rather than having two extreme opposite views, one can also con-
sider coding schemes in between these two, because the firing rate coding is more
similar to a temporal coding when small time windows are used [5].
109