Biomedical Engineering Reference
In-Depth Information
Fig. 7.1
Comparison of
several brain imaging
techniques, in terms of their
spatial resolution (
vertical
axis
), time resolution
(
horizontal axis
), and
invasiveness (
color-code
)
between different sources as in EEG. Chapter 6 deals with diffusion MRI data, a type
of MR imaging which bring complementary structural connectivity information to
EEG and MEG.
This chapter is devoted to the analysis of MEG and EEG signals, in order to
extract meaningful information from them. The signals measured by MEG and
EEG belong to one of two categories: spontaneous, i.e., endogeneously produced
by the brain, or evoked, i.e., triggered by an outside stimulus. In the latter case, the
EEG is referred to as Evoked Response Potential (ERP). Since the seminal work of
Lehmann et al. on microstates [
16
], much effort is being devoted in the community
in order to be able to analyze single-trial measurements, or to segment continuous
strands of data into pieces within which the signals enjoy similar properties.
7.1.3
Bioelectromagnetic Signal Analysis
In EEG and MEG signals, the information of interest has a very low signal-to-noise
ratio (SNR), because of high ongoing brain activity, not necessarily related to the
object of the study. In fact, the notion of “noiseless signal” does not really exist. In
evoked studies, a way to improve the SNR is to measure brain activity across many
repetitions of the same experiment. Each of the repeated measurements is referred
to as a trial, and multitrial analyses allow to deal with the global dataset.
A simple procedure to apply to a multitrial dataset is to translate each trial in
time so that zero corresponds to the onset of the stimulation, and to average the
measurements across trials. The resulting data is called Averaged Evoked Response
Potentials in the case of EEG. Averaged ERP typically consist of a series of waves,
whose latencies and amplitudes can be extracted.
Search WWH ::
Custom Search