Biomedical Engineering Reference
In-Depth Information
curves the grand-averages for the illusory figures (IF) and con-
trol figures (CF); the control figures have the inducers rotated
so no illusory contours are perceived. The activation curves are
shown only for the cases when an IF effect is present, i.e. when
ANOVA has shown that the response for IF and CF was signifi-
cantly different. The activation curves below the middle row show
grand-averages for the three different tasks: Classification task
when subjects had to indicate the presence of an (illusory) shape
(illusory shape discrimination); when the subjects had to respond
randomly every time an image appeared, irrespective of the pres-
ence or absence of illusory figures (random response); and when
subjects simply viewed the stimuli passively (passive viewing). The
activation curves are shown only for the cases when a task effect
is present, i.e. when ANOVA has shown that the response for dif-
ferent tasks was significantly different. These results showed for
the first time a distinct difference between processing when stim-
uli were presented to the center and periphery of the visual field.
In each and every task, stimuli presented at the center produce
an IF effect first in V1/V2 (
∼
100 ms) and then in the LOC and
FG (130-160 ms); this IF effect thus appears to proceed inde-
pendently of attention. For center presentation, attention effects
show up late, after 200 ms in the FG activation. For presentation
in the periphery, no IF effects are seen early, they are seen only
after 200 ms mainly in the FG. Peripheral stimuli produce early
attentional effects first in V1/V2 (80-100 ms), then in LOC (80-
150 ms) and finally in the FG (120-180 ms).
In summary, the results presented show that very precise
information can be extracted in both the temporal and spatial
domain. Very simple techniques, almost like reading the raw sig-
nal are enough for strong and superficial generators (
Fig. 8.1)
.
Full tomographic techniques can be used to obtain real-time
information across the brain in single trials or from a small set
of trials as shown in
Fig. 8.2
. The ability to extract reliable infor-
mation from a small number of trials is exploited in the last exam-
ple.
Figure 8.3
shows when and where illusory figure process-
ing takes place and how task demands modulate the allocation of
attention in each area. These results were obtained from a detailed
MFT analysis of average MEG data from seven subjects, using
only a small number of trials in each average.
5. Notes
5.1. Disadvantages
The need for shielding and use of liquid helium makes MEG an
expensive technology both in terms of the cost of hardware and
the operating costs. Another disadvantage of MEG is the need
for the subject to stay motionless while data are collected. MEG
is insensitive to radial currents; so generators close to the center
