Techniques of EEG Recording and Preprocessing - Quantitative EEG Analysis Methods and Clinical Applications - page 23

CHAPTER 2
Techniques of EEG Recording and
Preprocessing

Ingmar Gutberlet and Stefan Debener 1
Tzyy-Ping Jung and Scott Makeig 2

This chapter summarizes the key features of EEG signals, event-related potentials,
and event-related oscillations, and then more recent EEG hardware developments
are discussed. In particular, different electrode types such as passive and active elec-
trodes, as well as different electrode cap systems and layouts suitable for high-den-
sity EEG recordings, are introduced and their potential benefits and pitfalls
mentioned. The third part of this chapter focuses on prominent exogenous and
endogenous EEG artifacts and on different procedures and techniques of EEG arti-
fact rejection and removal. Specifically, in the final part of this chapter, independent
component analysis (ICA) is introduced. ICA can be used for EEG artifact correc-
tion and for the spatiotemporal linear decomposition of otherwise mixed neural sig-
natures. In combination with state-of-the-art recording hardware, the advanced
analysis of high-density EEG recordings provides access to the neural signatures
underlying human cognitive processing.

2.1 Properties of the EEG

Recently the dominant role of EEG and MEG in understanding the human
brain-behavior relationship has been recognized again. In contrast to functional
magnetic resonance imaging (fMRI), EEG and MEG techniques monitor large-scale
human brain activity patterns noninvasively and with millisecond precision, which
is crucial for understanding the neural foundations of cognitive functions. The fol-
lowing section summarizes some of the assumptions and properties of the EEG sig-
nal in event-related brain research before the hardware necessary for EEG recording
and preprocessing steps is discussed.

2.1.1 Event-Related Potentials
The common way of analyzing event-related EEG signals is the calculation of
event-related potentials (ERPs). This is done by repeatedly presenting an event of
interest, such as a visual stimulus on a computer screen, and analyzing the small

1. These authors contributed to Sections 2.1, 2.2, and 2.3.

2. These authors contributed to Section 2.4.

23