Biomedical Engineering Reference
In-Depth Information
paradigm. Deviations in amplitude and topopgraphy of components can lead to in-
ferences about the nature and locus of brain dysfunction.
One of the most commonly considered ERP component is the P300. In fact, the
“300” is a naming convention rather than the latency value since this positive de-
flection may appear between 300 and 600 ms after the stimuli of different modalities
(e.g., visual, auditory, sensory). P300 is thought to reflect processes involved in stim-
ulus evaluation or categorization. It is usually elicited using the oddball paradigm in
which low-probability target items are inter-mixed with high-probability non-target
(or “standard”) items. The topography of the P300 recorded by scalp electrodes re-
veals a maximum over the midline centro-parietal regions. The most direct way to
localize functional activation may be obtained by means of iEEG signal analysis. In
particular the P300 components were extensively studied with the aim of identifica-
tion of distributed networks responsible for their generation [Sinai et al., 2009]. A
number of studies addressed in the above reference explored the neural substrates
of a variety of ERPs elicited by functional activation of motor and sensory cortices.
The literature on experimental paradigms for evoking and the interpretation of am-
plitudes, latencies, and topographies of late ERPs such as P300 (and its subdivision
into P3a and P3b), N200, N2pc, N170, P200, N400, P600 is very rich and reviewing
it is far beyond the scope of this topic. However, from the signal analysis point of
view the problems and techniques for dealing with the ERPs are common to all of
them and are described in the following paragraphs.
The main difficulty in analysis of ERPs is the fact that, in comparison to spon-
taneous activity of the brain, the changes provoked by the stimuli are often very
subtle; they are by an order of magnitude smaller than the ongoing EEG. The aim
of the analysis is separation of the activity of brain not related to the stimuli (called
the ongoing or background activity) which is treated as noise, and the activity related
to the stimulus which is the signal of interest. Thus in technical terms the problem
in quantification of ERPs is the poor signal to noise ratio. The choice of methods to
improve the signal to noise ratio depends on the assumptions that one makes about
the relation between the signal, the stimulus, and noise. In the following sections we
discuss methods appropriate for analysis of responses that are phase-locked to the
stimulus ( Sect. 4.1.7.1) and these that are non-phase-locked ( Sect. 4.1.7.3) .
4.1.7.1
Analysis of phase locked responses
4.1.7.1.1 Time averaging The basic technique used in the analysis of EP is
the time averaging of multiple responses time locked to the stimulus. The first ap-
proach of this type, which resulted in visualization of the EPs was reported by Daw-
son [Dawson, 1947]. In his solution a number of time locked responses were pho-
tographically superimposed. Today, the averaging of the digitized epochs of EEG,
aligned to the stimulus is easily done by means of a computer.
The time averaging method is based on three assumptions:
The brain response is time locked (invariably delayed) in respect to the stimu-
lus
 
Search WWH ::




Custom Search