Information Technology Reference
In-Depth Information
9.3
Data Validation
9.3.1 Contingent Negative Variation Extraction
For a given participant, the CNV analysis will allow us to determine whether the
relaxation step was effective. CNV extraction has been performed by applying the
event-related potentials (ERPs) method (Rosenblith
1959
). It consists, in the present
experimental design, on averaging the electrical activity recorded in synchrony with
all warning signals (S1: beep) until the response stimulus (S2: square). Such
average allows event-related brain activity components, re
ecting stimulus pro-
cessing, to emerge from the overall cortical electrical activity, unrelated to the task
performed. Thus, in our paradigm, a negative de
ection of the averaged waveform,
called CNV, is obtained (Walter et al.
1964
). This attentional component has the
property of decreasing in amplitude when the participant is less alert, either because
he is distracted (Tecce
1979
), is deprived of sleep (Naitoh et al.
1971
), or is falling
asleep (Timsit-Berthier et al.
1981
). This fundamental result is shown in Fig.
9.5
.In
this Figure, the CNV is plotted as a dotted line for an alert participant and as a solid
line for a participant which is less alert. The amplitude of the CNV is proportional
to the alertness of the subject.
That is why, although the instruction given to the participant during CNV
acquisition was to press the space bar as quickly as possible after the square
appearance, the reaction time is not investigated in this study. However, the way the
participant prepares to perform the task is observed.
The comparison of the amplitude of the CNV between tasks performed in steps 2
and 5 is used to determine whether the alertness of a participant has changed. It
allows us to know if he is actually relaxed. Only the positive cases, for which the
amplitude of the CNV has signi
cantly declined, were selected for comparative
'
analysis of their raw EEG
s (stages 1 and 4). Their EEGs were then tagged,
respectively, as
state. An example of a participant kept after
studying his CNV is shown in Fig.
9.6
and an example of a rejected participant is
given in Fig.
9.7
.
In these Figures, the solid curve represents the CNV recorded during step 2 and
the dotted curve represents the CNV recorded in step 5. The solid vertical lines
correspond to warning signals (S1: beep, S2: square). The area between the curve
“
normal
”
or
“
relaxed
”
Fig. 9.5
Representation of
the amplitude variation of the
CNV with respect to the
alertness of a participant
participant less alert
100 300
Time (ms)
participant more alert
