Brain cognitive-specific functional states studied with quantitative EEG

1
Brain cognitive-specific functional states
studied with quantitative EEG

• Bechtereva N.P., Danko S.G.
• Institute of the Human Brain, Russian Academy of
  Sciences, Sankt-Petersburg, Russia.

2
Initially functional state of the central
nervous system was defined and treated as a
background to the behavioural activities of
humans and animals (e.g. Sokolov, 1975)
Functional states regulated by modulating system
of the brain are a mandatory component of any
activity and behavior (Danilova, 2001 - from
Textbook on Psychophysiology, ed Y.Alexandrov ).
3
A well-known difference with conditioned reflex
activity is the great discontinuity and
polimorphism of the changes in the EEG during
tests that are strictly psychological. although
local changes can be noted during psychological
tests, the development of these local changes is
always a part of the general modification of
bioelectric activity. Thus, EEG changes during
mental activity reflect not so much the active
state of individual structures as they reflect
general changes in the brain, which evidently
optimize the conditions for activity.
From Bechtereva N.P. The neurophysiological
aspects of human mental activity. Leningrad,
Meditsina, 1971. (in Russian) Bechtereva N.P.
The neurophysiological aspects of human mental
activity. NYOxford Univ. Press, 1978
4
The main ideas feeding the studies in question
are
1. Cognitive activities include not only
short-term transitions of the information
processed but are based also on specific
interactions of network brain systems underlying
basic psychic functions such as vigilance,
attention, memory.
2.There are some necessary integrations of the
brain systems which result in certain functional
states of brain characteristic to certain
cognitive activities
3. There should be certain manifestations in EEG
processes related to a functional state supposed.
5
The term quantitative
electroencephalography qEEG should be
understood so that both digital processing of EEG
raw data and statistical evaluation of observed
differences are mandatory.
When using term state-related qEEG
mean values of EEG parameters are assumed to be
constant during certain functional states. The
states under considerations here are differed
according to cognitive/emotional activities to be
controlled. And the states should be durable
enough to allow statistical estimations of mean
values with acceptable accuracy and they can be
discriminated from transitory processes and acts
on this base.
6
The volunteers who undergone computer
EEG registration presumably in different
functional states were of both sexes, aged 17-27,
university students and graduates, right-handed,
without acute problems with health and without
CNS injuries or diseases in their past.
EEG was recorded from 19 sites
(International 10-20 system), monopolar with
reference to earlobes.
EEG records were processed on intervals 1-3s
duration corresponding to different states under
investigation, 20-40 intervals for each state
were taken for processing
Mean values of EEG power and coherence
for each subject in each state were calculated in
frequency bands ? (1,5-3,5Hz), ? (4-7Hz), ?1
(7,5-9,5Hz), ?2 (10-12,5Hz), ?1 (13-18Hz), ?2
(18,5-30Hz). In the last investigations low ?
(30-40Hz) was added.
7
Statistical analysis of the values was
conducted not as testing of certain hypotheses
but as screening of the all data available using
multi-way Anova and within subjects design.
Frequency bands, states and zones were
considered as effective factors.
Greenhouse-Geisser correction was applied for
evaluation of main effects and their interactions
significance.
Statistical significances of differences
between means in certain frequency bands, states
and zones were tested with post hoc tests
(multiple comparisons). Tukey HSD test was used
as more conservative and less sensitive, Fisher
LSD test as less conservative and more
sensitive.
8
Commonly used paradigm of psychophysiological
investigation is based on comparisons of test
situations as close as possible except for an
element being investigated.
Thats a very productive paradigm indeed
especially if one is sure about physiological
interpretation of the parameters controlled. That
is the case in rCBF tomography studies but
unfortunately not so with EEG
In the situation its desirable to anticipate
the results related to cognitive-specific FS with
a result of qEEG application to well-known
physiological states quiet rest with eyes open
and eyes closed
9
State-related statistically significant EEG power
changes Contrast eyes closed eyes open
D
T
A1
A2
B1
B2
G
N57
EC-EO

Tukey HSD
Red colors - values higher in the first of the
compared states, blue lower.
10
State-related statistically significant EEG power
changes Contrast eyes closed eyes open
D
T
A1
A2
B1
B2
G
N57
EC-EO

Tukey HSD
1. Occipital and parietal zones are most
involved. 2. The changes are
unidirectional in all frequency bands
considered. 3. So decrease of rCBF in
the visual cortex (Gebhardt et al., 2001,2002)
corresponds to EEG synchronization in every of
the frequency band
11
State-related statistically significant EEG power
changes Contrasts memorizing(MM) eyes open
rest(EO) memory retrieval(MR) EO MM-MR
N57

Red colors - values higher in the first of the
compared states, blue lower.
Tukey HSD
12
State-related statistically significant EEG power
changes Contrasts memorizing(MM) eyes open
rest(EO) memory retrieval(MR) EO MM-MR
N57

1. Both MM-EO and MR-EO patterns are
rather similar desynchronization in alpha and
synchronization in beta2 and gamma looks like
classical non-specific activation. 2.
But in MM-MR no difference in alpha and prominent
difference in beta2 and gamma non-specific
activation should be omitted. 3. In MM-EO
and MR-EO the changes in alpha 2 and gamma are
rather widespread.
Red colors - values higher in the first of the
compared states, blue lower.
13
State-related statistically significant EEG power
changes Contrasts positive induced emotions(P)
time count(C) negative induced emotion(N)
C P-N
N16

P-C
N-C
Red colors - values higher in the first of the
compared states, blue lower.
P-N
Fisher LSD
14
State-related statistically significant EEG power
changes Contrasts positive induced emotions(P)
time count(C) negative induced emotion(N)
C P-N
N16

1. Both P-C and N-C patterns are
unidirectional in all the frequency bands
concerned. 2. But in N-C the changes are
much less pronounced and that is confirmed in P-N
contrast
Red colors - values higher in the first of the
compared states, blue lower.
15
State-related statistically significant EEG power
changes Contrast complicated verbal task
(search for a word changing the proverb sense)
non-complicated verbal task (remembering of the
certain word missing in the proverb)
N67

Gamma
Alfa1
Alfa2
Beta1
Beta2
Theta
Delta
Tukey HSD test
Red colors - values higher in the first of the
compared states, blue lower.
16
State-related statistically significant EEG power
changes Contrast complicated verbal task
(search for a word changing the proverb sense)
non-complicated verbal task (remembering of the
certain word missing in the proverb)
N67

Gamma
Alfa1
Alfa2
Beta1
Beta2
Theta
Delta
1. Even in a situation when main and
control tasks are intentionally close one using
srqEEG can reveal significant changes between the
states corresponding to the tasks performed
2. Here the changes are unidirectional and
concentrated in gamma and beta2 bands.
Tukey HSD test
Red colors - values higher in the first of the
compared states, blue lower.
17

• Conclusions
• Brain functional states cognitive-specific states
  included can be separated using srqEEG with good
  statistical reliability.
• In the situations investigated the differences
  were widespread in space indeed and were present
  in a variety of frequency bands considered.
• Both unidirectional and bidirectional
  statistical significant differences could be
  observed.
• 4. So potentially EEG power values in several
  frequency bands can serve as a tool for
  multiscale (not only regional and general cortex
  activation/deactivation) rating of brain
  functional states.

18
We greatly appreciate contributions of
N.V.Shemiakina, M.G.Starchenko, M.L.Solovjeva and
Y.A.Boitsova in the studies presented
Thank you for your kind attention!
19
Statistically significant EEG differences between
states of creative and non-creative thinking
with various tests


1.Sense changing word

2.Storymaking
V e r b a l

3.Logical chains



4.Original definitions

5.Drawing pictures from a restricted set of
elements
20
Statistically significant EEG differences between
states of creative and non-creative thinking
with various tests


1.Sense changing word

2.Storymaking
Posterior psychological analysis
highlighted a number of factors presumably
responsible for the scatter of EEG results with
the verbal tests time frames, involvement of
remote associations and of mental imagery,
emotion-producing tension etc. It looks like the
factor of creative load is not so effective in
modulating EEG power in the frequency bands d - ß
if compared with aforementioned ones.
V e r b a l

3.Logical chains



4.Original definitions

5.??????? ?? ?????? ?????????
21
Variants of statistically significant EEG
differences between states of creative and
non-creative thinking in 3 groups of subjects


Gr.1N22
Gr.2N25
Gr.3N20
Red colors - values higher in the first of the
compared states, blue lower.
22
Emotional components remarkably influence EEG
power during creative thinking
Alfa1
Alfa2
Beta1
Beta2
Theta
Delta



C-NC
(CPE)-NC
(CNE)-NC
Red colors - values higher in the first of the
compared states, blue lower.
23
A simple task (search for grammar mistakes in a
presentation) added to the creative task
influences EEG power statistical significant
differences between states of creative and
non-creative thinking
Red colors - values higher in the first of the
compared states, blue lower.
24
Statistically significant EEG differences between
states of creative (search for a word changing
the proverb sense) and non-creative (remembering
of the certain missing word) thinking
Tukey HSD test
Gamma

EEG power



Red colors - values higher in the first of the
compared states, blue lower.
EEG coherence
25
Statistically significant EEG differences between
states of creative (search for a word changing
the proverb sense) and non-creative (remembering
of the certain missing word) thinking
Tukey HSD test
Gamma

The results of the test can be guardedly
treated as argumentative of activation in memory
scanning system (? synchronization) and in memory
control zones (ß synchronization in temporal
areas) in creative thinking state.
EEG power



EEG coherence
26
Statistically significant EEG differences between
states of creative and non-creative thinking
with two statistical post hoc tests used
EEG power
Alfa1
Alfa2
Beta1
Gamma
Delta
Theta
Beta2
EEG coherence
Red colors - values higher in the first of the
compared states, blue lower.
27
Statistically significant EEG differences between
states of creative and non-creative thinking
with two statistical post hoc tests used
EEG power
Alfa1
Alfa2
Beta1
Gamma
Delta
Theta
Beta2
Number of 70 in a group seems enough for
reliable results with conservative test Tukey.
For EEG power the results reflect characteristic
patterns recognizable in subgroups. For EEG
coherence the conservative test in such a group
doesnt reveal characteristic patterns.
EEG coherence
Red colors - values higher in the first of the
compared states, blue lower.
28
(No Transcript)
29
Today we consider quantitative EEG as a
helpful instrument for versatile screening of
numerous groups of subjects in evaluation of
hypotheses, psychological tests and research
methods related to basic mechanisms of cortex
functioning.