Dissociation Between Attention and Consciousness During a Novel Task: an ERP Study
While consciousness and attention seem to be tightly connected, recent evidence has suggested that one of these processes can be present in the absence of the other. Recent researches showed that observers can pay attention to an invisible (unconscious) stimulus, and that a stimulus can be clearl...
Gespeichert in:
| Datum: | 2015 |
|---|---|
| Hauptverfasser: | , , |
| Format: | Artikel |
| Sprache: | English |
| Veröffentlicht: |
Інститут фізіології ім. О.О. Богомольця НАН України
2015
|
| Schriftenreihe: | Нейрофизиология |
| Online Zugang: | http://dspace.nbuv.gov.ua/handle/123456789/148181 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| Назва журналу: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| Zitieren: | Dissociation Between Attention and Consciousness During a Novel Task: an ERP Study / R. Davoodi, M.H. Moradi, A. Yoonessi // Нейрофизиология. — 2015. — Т. 47, № 2. — С. 169-180. — Бібліогр.: 32 назв. — англ. |
Institution
Digital Library of Periodicals of National Academy of Sciences of Ukraine| id |
irk-123456789-148181 |
|---|---|
| record_format |
dspace |
| spelling |
irk-123456789-1481812019-02-18T01:23:20Z Dissociation Between Attention and Consciousness During a Novel Task: an ERP Study Davoodi, R. Moradi, M.H. Yoonessi, A. While consciousness and attention seem to be tightly connected, recent evidence has suggested that one of these processes can be present in the absence of the other. Recent researches showed that observers can pay attention to an invisible (unconscious) stimulus, and that a stimulus can be clearly perceived (seen) in the absence of attention. We have proposed a novel psychophysical task to explore neural correlates of top-down attention and consciousness. The task is meant to confirm that these two processes can occur independently of each other. EEG was recorded during realizations of the task and target-locked event-related potentials (ERPs) for masked and unmasked conditions were constructed. Time features corresponding to the P100, N150, and P300 components were extracted for each channel separately. Utilizing these features, we employed some common classifiers for classification of the fourfold state. Our task could separate attention and consciousness successfully through their neural correlates. The results indicate that some of the mentioned components changed when attention or consciousness occurs. By comparing difference waves in each condition separately, our results introduce new ERP correlates of attention and consciousness. We also revealed that parieto-occipital areas are the most relevant areas for dissociation between attention and consciousness. To our knowledge, this is the first time that these correlates are introduced in a separable mode, and that the classification accuracies are reported for this purpose. Як вважають, увага та усвідомлення тісно поєднані, проте результати недавніх досліджень дають підстави думати, що один із цих процесів може реалізуватися за відсутності другого. У новітніх експериментах виявилося, що спостерігачі можуть приділяти увагу невидимому (неусвідомлюваному) стимулу і що стимул може чітко розпізнаватися (диференціюватися) за відсутності уваги. Ми запропонували нове психофізіологічне завдання для дослідження нервових корелятів змін рівнів уваги та усвідомлення. Цільові ЕЕГпотенціали, пов’язані з подією (ППП), відводили в умовах пред’явлення замаскованих та незамаскованих візуальних патернів. Часові характеристики компонентів P100, N150 та P300 визначалися роздільно для відведень по кожному ЕЕГканалу. З урахуванням цих характеристик були використані певні загальні класифікатори для параметрів, спостережуваних у всіх чотирьох стимуляційних станах. У нашому тесті ефекти уваги та усвідомлення могли бути успішно розділені відповідно до їх нервових ЕЕГ-корелятів. Згідно з отриманими результатами, згадані вище компоненти ППП змінюються залежно від того, як проявляється увага або усвідомлення. Роздільне порівняння відмінностей між хвилями ППП для кожної з умов дозволило виявити нові кореляти уваги та усвідомлення, що відбиваються в ППП. Показано також, що тім’яно-потиличні кортикальні зони є структурами, найбільшою мірою пов’язаними з дисоціацією ефектів уваги та усвідомлення. Як ми вважаємо, ці кореляти вперше представлені з використанням методики, котра дозволяє їх розрізнити, і наведені дані щодо точності такої диференціації. 2015 Article Dissociation Between Attention and Consciousness During a Novel Task: an ERP Study / R. Davoodi, M.H. Moradi, A. Yoonessi // Нейрофизиология. — 2015. — Т. 47, № 2. — С. 169-180. — Бібліогр.: 32 назв. — англ. 0028-2561 http://dspace.nbuv.gov.ua/handle/123456789/148181 612.014:612.821.2:159.922 en Нейрофизиология Інститут фізіології ім. О.О. Богомольця НАН України |
| institution |
Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| collection |
DSpace DC |
| language |
English |
| description |
While consciousness and attention seem to be tightly connected, recent evidence has suggested
that one of these processes can be present in the absence of the other. Recent researches
showed that observers can pay attention to an invisible (unconscious) stimulus, and that a
stimulus can be clearly perceived (seen) in the absence of attention. We have proposed a novel
psychophysical task to explore neural correlates of top-down attention and consciousness. The
task is meant to confirm that these two processes can occur independently of each other. EEG
was recorded during realizations of the task and target-locked event-related potentials (ERPs)
for masked and unmasked conditions were constructed. Time features corresponding to the
P100, N150, and P300 components were extracted for each channel separately. Utilizing these
features, we employed some common classifiers for classification of the fourfold state. Our
task could separate attention and consciousness successfully through their neural correlates.
The results indicate that some of the mentioned components changed when attention or
consciousness occurs. By comparing difference waves in each condition separately, our
results introduce new ERP correlates of attention and consciousness. We also revealed that
parieto-occipital areas are the most relevant areas for dissociation between attention and
consciousness. To our knowledge, this is the first time that these correlates are introduced in
a separable mode, and that the classification accuracies are reported for this purpose. |
| format |
Article |
| author |
Davoodi, R. Moradi, M.H. Yoonessi, A. |
| spellingShingle |
Davoodi, R. Moradi, M.H. Yoonessi, A. Dissociation Between Attention and Consciousness During a Novel Task: an ERP Study Нейрофизиология |
| author_facet |
Davoodi, R. Moradi, M.H. Yoonessi, A. |
| author_sort |
Davoodi, R. |
| title |
Dissociation Between Attention and Consciousness During a Novel Task: an ERP Study |
| title_short |
Dissociation Between Attention and Consciousness During a Novel Task: an ERP Study |
| title_full |
Dissociation Between Attention and Consciousness During a Novel Task: an ERP Study |
| title_fullStr |
Dissociation Between Attention and Consciousness During a Novel Task: an ERP Study |
| title_full_unstemmed |
Dissociation Between Attention and Consciousness During a Novel Task: an ERP Study |
| title_sort |
dissociation between attention and consciousness during a novel task: an erp study |
| publisher |
Інститут фізіології ім. О.О. Богомольця НАН України |
| publishDate |
2015 |
| url |
http://dspace.nbuv.gov.ua/handle/123456789/148181 |
| citation_txt |
Dissociation Between Attention and Consciousness During a Novel Task: an ERP Study / R. Davoodi, M.H. Moradi, A. Yoonessi // Нейрофизиология. — 2015. — Т. 47, № 2. — С. 169-180. — Бібліогр.: 32 назв. — англ. |
| series |
Нейрофизиология |
| work_keys_str_mv |
AT davoodir dissociationbetweenattentionandconsciousnessduringanoveltaskanerpstudy AT moradimh dissociationbetweenattentionandconsciousnessduringanoveltaskanerpstudy AT yoonessia dissociationbetweenattentionandconsciousnessduringanoveltaskanerpstudy |
| first_indexed |
2025-07-12T18:32:22Z |
| last_indexed |
2025-07-12T18:32:22Z |
| _version_ |
1837467065215090688 |
| fulltext |
NEUROPHYSIOLOGY / НЕЙРОФИЗИОЛОГИЯ.—2015.—T. 47, № 2 169
UDC 612.014:612.821.2:159.922
R. DAVOODI1, M. H. MORADI1, and A. YOONESSI2
DISSOCIATION BETWEEN ATTENTION AND CONSCIOUSNESS DURING A NOVEL
TASK: AN ERP STUDY
Received January 15, 2014
While consciousness and attention seem to be tightly connected, recent evidence has suggested
that one of these processes can be present in the absence of the other. Recent researches
showed that observers can pay attention to an invisible (unconscious) stimulus, and that a
stimulus can be clearly perceived (seen) in the absence of attention. We have proposed a novel
psychophysical task to explore neural correlates of top-down attention and consciousness. The
task is meant to confirm that these two processes can occur independently of each other. EEG
was recorded during realizations of the task and target-locked event-related potentials (ERPs)
for masked and unmasked conditions were constructed. Time features corresponding to the
P100, N150, and P300 components were extracted for each channel separately. Utilizing these
features, we employed some common classifiers for classification of the fourfold state. Our
task could separate attention and consciousness successfully through their neural correlates.
The results indicate that some of the mentioned components changed when attention or
consciousness occurs. By comparing difference waves in each condition separately, our
results introduce new ERP correlates of attention and consciousness. We also revealed that
parieto-occipital areas are the most relevant areas for dissociation between attention and
consciousness. To our knowledge, this is the first time that these correlates are introduced in
a separable mode, and that the classification accuracies are reported for this purpose.
Keywords. top-down attention; consciousness; ERP; psychophysical task.
1 Department of Biomedical Engineering, Amirkabir University of Technology
(Tehran Polytechnic), Tehran, Iran
2 Iranian National Center for Addiction Studies, Institute for Cognitive Science
Studies, and School of Advanced Technologies in Medicine, Tehran University
of Medical Sciences, Tehran, Iran,
Correspondence should be addressed to:
R. Davoody (r.davoodi@aut.ac.ir),
M. H. Moradi (mhmoradi@aut.ac.ir), or
A. Yoonessi (yoonessi@gmail.com)
INTRODUCTION
Within the last part of the past century, the interest of
researchers on the influence of top-down attention and
consciousness on perception has steadily increased.
The relevant discussions have raised the question of
the relationship between attention and consciousness.
Several studies have used various types of attention and
consciousness, and the obtained results were different
and sometimes opposite [1-4]. Both above terms refer
to complex concepts, and this implies the importance
of clarifying the definition of both concepts [5].
Some scholars claim that both attention and
consciousness are irrefrangibly connected [6-9].
Some authors argue that consciousness is necessary
for attention, while others consider attention as a
prerequisite of consciousness [2, 3, 10-12]. For
instance, Naccache et al. [9] stated that attention is
necessary but not sufficient for consciousness. On the
other hand, some recent studies showed that observers
can pay attention to an invisible stimulus [13, 14], and
that a stimulus can be clearly seen when attention is
absolutely or nearly absent [15-17].
It is essential to consider the distinction between
types of attention and those of awareness in a
discussion of the relationship between attention
and consciousness. In the recent literature on
neural correlates of consciousness (NCC), two
concepts of consciousness have been distinguished.
These are phenomenal consciousness and access
consciousness [18-20]. The former is defined as the
case where qualitative experiences, such as simple
color sensations, are present. In the visual modality,
phenomenal consciousness covers the entire subjective
visual field in a similar manner as iconic memory
does [11]. Neural mechanisms of the contents of
phenomenal visual consciousness most likely reside
in the cortical extrastriate visual areas, especially
NEUROPHYSIOLOGY / НЕЙРОФИЗИОЛОГИЯ.—2015.—T. 47, № 2170
R. DAVOODI, M. H. MORADI, and A. YOONESSI
along the ventral visual stream [11]. Reflective,
or access, consciousness, by contrast, consists of
only those relatively few contents of phenomenal
consciousness that have been selected in an attentive
mode for further cognitive processing in working
memory [11]. Considering these definitions, it seems
that phenomenal consciousness is a prerequisite of
access consciousness. The contents of reflective
consciousness can be reported verbally or otherwise
expressed through any voluntary output mechanisms;
these contents can be named, conceptualized,
propositionally thought about, and encoded into
long-term memory. Neural correlates of reflective
consciousness involve modality-independent frontal
areas important for the control of top-down attention
and working memory; these correlate are closely
connected to voluntary motor output mechanisms and
access to declarative memory.
Some lines of research distinguish the relation
between different types of attention and consciousness
and do not generalize their results in acceptance or
rejection of the relationship between these two
phenomena. Hsu et al. [21] stated that “voluntary and
involuntary spatial attentions interact differently with
awareness.” While the classical view claims that strong
evidence supports the existence of a link between
spatial attention and consciousness, there is evidence
that spatial attention can be deployed in the absence of
conscious perception of the attended information [5].
It is crucial to note that all studies reviewed by Koch
and Tsuchiya [18] searched endogenous (or top-down)
forms of spatial attention and awareness.
Some recent s tudies (e .g. , [22]) in favor
of afterimage effects argue that attention and
consciousness can exert opposing effects on visual
perception. Boxtel et al. [4] claimed that attention
reduces the complexity of the incoming input so that
the brain can process it online and in real time. This
could be the function of the dorsal visual stream for
action. The fronto-parietal area has been implicated in
the control of attention, which is a part of the dorsal
vision-for-action pathway and, in contrast to the
ventral vision-for-perception pathway, has been linked
to consciousness. These two streams interact deeply
under most circumstances, but they can be dissociated
under some conditions [23]. Many examples of
attention without consciousness may be thought
of as normal functioning of the dorsal attention-
orienting system without proper functioning of the
ventral system. So, an object may attract attention
without giving rise to consciousness via this pathway.
Similarly, consciousness without attention may occur
because of some ventral functioning without help of
attentive amplification from the dorsal systems [22].
Now, we propose simple and admissible definitions
for both concepts. By attention, here we refer to top-
down perceptual attention, which a subject would
attend to as he/she intends, and not stemming from
vigilance, or arousal, or any involuntary attention, as
was stated by Boxtel et al. [22]. By consciousness,
we refer to the content of consciousness (sometimes
also referred to as awareness), and not to states of
consciousness (e.g., wakefulness, dreamless sleep,
or coma) as was assumed by some authors [19, 24].
In modern literature, the effects of attention are often
objectively defined and measured as a reduced reaction
time and improved performance. In a similar way,
awareness of an object is determined as a subjective
report in combination with objective forced-choice
performance. With these measures in place, a variety
of methods has been used to manipulate attention
(cueing, divided attention, etc.) and consciousness
(masking, crowding, and binocular rivalry) [5].
A typical way of making a stimulus either
unavailable or available for visual consciousness is to
present a mask after it with varying asynchronously
presented pictures of objects or non-objects between
a pattern mask that was presented before (forward
mask) and after (backward mask) the stimulus. The
participants were asked to decide whether the stimulus
represented a target object or not [11].
Recording of event-related potentials (ERPs) is
one of the most informative methods of noninvasive
monitoring of the state of the brain. Therefore,
evaluation of ERPs are attractive for neuroscience
and, particularly, for attention-consciousness studies
[19]. In the relevant works [11, 25-28], it was
found that that an enhancement of the P100 wave
correlates with visual consciousness. These studies
were faced with the problem of possible interference
between attention and arousal, as they manipulated
consciousness using methods that are vulnerable to
attention. This is usually acknowledged in the studies
where the P100 component was reported to correlate
with consciousness.
The purpose of our study was to demonstrate the
distinction between phenomenal consciousness and
top-down attention in a visual task. To do this, a novel
psychophysical task was designed, and attention and
consciousness were manipulated separately. As a
result, a fourfold state was built. In order to overcome
some shortcomings existing in previous studies, our
NEUROPHYSIOLOGY / НЕЙРОФИЗИОЛОГИЯ.—2015.—T. 47, № 2 171
DISSOCIATION BETWEEN ATTENTION AND CONSCIOUSNESS DURING A NOVEL TASK: AN ERP STUDY
task tried to eliminate the adverse effects of attending
to special targets (e.g., face images); these effects
accompany a peripheral task during attention (e.g.,
computing the number of targets). The former problem
was resolved by applying a mixture of target types
instead of one type, and the latter was assumed to be
resolved by combining two phases of the attention
time (attention only and attention together with
computation). The ERP waveforms were constructed
and relative ERP component features were extracted.
Statistical analyses (ANOVA) were done, and three
classifiers were employed in order to quantify the
results. The last step was channel prioritization
according to the classification accuracy.
METHODS
Subjects. Forty-eight volunteers (age 19 to 26 years)
were involved in the study and paid for their
participation. All had normal or corrected-to-normal
visual acuity.
Stimuli and Procedure. The stimuli were presented
on a CRT monitor with a gray background. The
program was written in MATLAB software using a
psychtoolbox.
We used a set of 100 images consisting of three
categories of the objects (fruits, scenes, and human
faces). Mask images were also created using phase-
scrambled versions of the same images. The
experiment consisted of four blocks, each consisting
of two trials. Each image was shown in the first
trial for only 20 msec followed by a 980-msec-long
presentation of the mask of the same image. In the
second trial, each image was shown for 1.0 sec. We
used these two trials as the two states of consciousness,
the first in an “unconscious” state and the second in a
“conscious” state. Examples of masked and unmasked
sequences for the unconscious and conscious states are
shown in Fig. 1.
Four blocks were used in which the attention was
manipulated. In the first and second blocks in all trials
(masked and unmasked), the target was a human face
for which the subjects were asked to attend. In the
third and fourth blocks, the target was a fruit. In the
first and third blocks, the subjects were required to
attend to the target, while in the second and fourth
blocks the subjects were asked to “count” the numbers
of the targets in both trials.
Recording and Analysis. EEG was recorded us-
ing eight active electrodes attached to an elec-
tro-cap electrode system (g.Tec.) according to in-
ternational 10/20 system sites O1/O2, PO8/PO7,
P6/P5, and F3/F4. The left earlobe was used as a ref-
erence, and lead Cz was a grounding electrode. An
electrode placed over the right eye was used for mon-
itoring vertical eye movements and blinks, and two
electrodes on the right side of the right eye and on
the left side of the left eye were used for monitor-
ing horizontal eye movements. EEG was filtered us-
ing a bandpass of 0.1-60 Hz, with a sampling rate of
256 sec–1. A 50 Hz notch filter was used. The imped-
100 sec
Target 20 msec
Mask 980 msec
Nontarget
20msec
100 sec
1 sec
Target
Nontarget
F i g. 1. Examples of a masked stimulus sequence for the unconscious state (A) and an unmasked stimulus sequence for the conscious state
(B). Consciousness was manipulated by masking (long vs. short stimulus onset asynchrony, SOA). The target stimuli were face or fruit
images; neutral stimuli were also presented. The participants were asked to be attentive and to respond to specified targets. In A) A trial was
manipulated with respect to consciousness and was repeated in all blocks belonging to the unconscious state; in B) a trial had no mask and,
thus, belonged to the conscious state.
Р и с. 1. Приклади замаскованої послідовності стимулів для неусвідомленого стану (А) та немаскованої послідовності для
усвідомленого стану (В).
A B
NEUROPHYSIOLOGY / НЕЙРОФИЗИОЛОГИЯ.—2015.—T. 47, № 2172
R. DAVOODI, M. H. MORADI, and A. YOONESSI
ance of the electrodes was below 5 kΩ. The baseline
corresponded to the activity within a −200 to 0 msec
interval preceding the stimulus onset. Trials with ar-
tifacts (>60 µV) in any of the electrodes were reject-
ed off-line. EEG for three subjects was eliminated
because of significant noise and other technical prob-
lems. A preprocessing block diagram is shown in Fig.
2.
ERP Extraction. The data acquired were analyzed
using MATLAB 2008a software (Math Works, USA).
In all trials, the epochs were extracted for each
stimulus separately from 200 msec before the stimulus
onset to 1.0 sec after the latter.
Target-locked ERPs for masked and unmasked
conditions were extracted. The recorded epochs during
the target stimuli were set as attention states, while
others were considered inattention ones. In a similar
way, the epochs during 20-msec-long stimuli were
assigned as unconscious states, and 1.0-sec ones as
conscious ones. Hence, there were four epoch groups
constructed for each participant that produced these
conditions, (i) attention with consciousness (1-sec-long
target images), (ii) attention without consciousness
(20-msec-long target images), (iii) consciousness
without attention (1-sec-long nontarget images), and
(iv) unconscious with no attention (20-msec-long
nontarget images).
The epochs were averaged for targets and nontargets
in aware and unaware conditions separately. The
grand-average ERPs for all subjects in channels O1
and O2 are shown in Fig. 3.
ERP Components. In the statistical analyses of
ERP data, we focused on the maximum amplitudes
and peak latency within P100 (90-130 msec), N150
(130-300 msec), and P300 (350-700 msec) time
windows [29] in occipital, parietal, and frontal
leads in all four states (i.e., unconscious-inattentive,
conscious-inattentive, unconscious-attentive, and
conscious-attentive) in which the effects specifically
related to masking (i.e., consciousness) and attention
were most clearly observable. These time windows
and electrode positions are similar to the ones used
in previous studies [11] in which visual awareness
negativity (VAN) overlapped the N150 component, and
late positivity (LP) overlapped the P300 wave, whereas
selective negativity [6] was observed as enhanced
negativity in response to targets; it overlapped the N2
component.
Extraction of the P100, N150, and P300 Features.
Because of visible differences in the waveforms
of different classes in the P100, N150, and P300
components, we have defined 16 features related
to the shape of the waveforms in these components
for all eight channels separately [30]. All features
were normalized to a zero mean and a unit standard
deviation according to the relation
.
As a result, we have 16 × 8 features for the P100,
16 × 8 features for the N150, and 16 × 8 features for
the P300 component.
Feature Selection. The task of feature selection is
an important one involved in signal processing, when
there are data with a high dimensionality or features
that may not provide valuable information [31]. So, we
defined a subset of features that describe the data in
an efficient way. During this phase, the feature space
is investigated, and different combinations of subsets
are ranked by the means of class separability criteria.
Sixteen features were extracted for each ERP
component in each channel. As a result there were 16
time features × ERP components × 8 channels. So,
feature selection is essential for classification. There
are various methods for selecting the most proper
features, where the Fisher method is a common way.
The idea of the Fisher method is to select a subset
combination of features in a way that the between-
class variance to within-class variance ratio is
maximized [31]:
where W is the greatest SW
–1SB eigenvector, SB is the
between-class variance, and SW is the sum of within-
class variances.
Here, we have used SPSS 11.5 software for
employing multi-class Fisher criteria.
Classification. After the feature extraction and
selection processes, a suitable classifier must be
employed in order to separate the multiclass data.
There are different classifiers, including linear and
nonlinear ones. As there is no single best-classification
algorithm, we have used here three powerful
classifiers, (i) linear discriminant analyzer (LDA), (ii)
K-nearest neighbor (KNN), and (iii) support vector
machines (SVM), in order to get the best-accuracy
results and also to compare these three in analyzing
our own data.
Discriminant Analyzer (LDA). The idea of this
classifier is to evaluate the W vector in a way that
NEUROPHYSIOLOGY / НЕЙРОФИЗИОЛОГИЯ.—2015.—T. 47, № 2 173
DISSOCIATION BETWEEN ATTENTION AND CONSCIOUSNESS DURING A NOVEL TASK: AN ERP STUDY
F i g. 3. Grand-average difference waveforms in lead O1 (aware-unaware conditions) under attentive condition (A), under inattentive
condition (B), under aware condition (C), and under unaware condition (D). Red dashed lines show the standard deviation.
Under attentive condition, increased positivities in the P100 and P300 are noticeable, while increased negativity in the N150 should be
examined. Under inattentive condition, increased positivity in the P100 is observable, but its amplitude is lower than under attentive
condition. Under aware condition, increased positivity in the P300 and increased negativity in the N150 are observable. Under unaware
condition, increased positivity in the P300 is seen, but its amplitude is lower than that under aware condition; increased negativity in the
N150 seems to be suspicious.
Р и с. 3. Усереднені диференціальні хвилі, що реєструвались у відведенні O1 за умов усвідомлення–неусвідомлення в різних
умовах (A–D).
-200 0 200 400 600 800 1000
-4
-2
0
2
4
6
8
10
12
µV µV
msec
-200 0 200 400 600 800 1000
-4
-2
0
2
4
6
8
10
12
msec
F i g. 2. Four-class grand-average ERP waveforms recorded from leads O1 (A) and O2 (B). Traces corresponmding to the unconscious-
inattentive condition are shown in red, to the unconscious-attentive condition, in blue, to the conscious-inattentive condition, in green, and
to the conscious-attentive condition, in black.
Р и с. 2. Усереднені хвилі пов’язаних з подією потенціалів чотирьох класів, відведені від локусів O1 (A) та O2 (B).
A B
-4
-2
0
2
4
6
8
10
12
msecmsec
-200 0 200 400 600 800 1000
-4
-2
0
2
4
6
8
10
12
µV µV
µVµV
A B
C D
-200 0 200 400 600 800 1000
NEUROPHYSIOLOGY / НЕЙРОФИЗИОЛОГИЯ.—2015.—T. 47, № 2174
R. DAVOODI, M. H. MORADI, and A. YOONESSI
the between-class variance to within class variance
ratio is maximized (Fisher criteria). As a result, g(x)
is calculated from a linear relation; according to a
threshold w0, if g(x) > 0, x belongs to one class, and if
g(x) < 0, x belongs to the other class.
K-Nearest Neighbor (KNN). The KNN is a
method for classifying objects based on closest
training samples in the feature space. The KNN is a
type of instance-based learning where the function
is approximated locally. The k-nearest neighbor
algorithm is among the simplest among all machine
learning algorithms. An object is classified by a
majority vote of its “k neighbors,” with the object
being assigned to the class most common among its k
nearest neighbors while k is a positive integer and is
typically small.
Support Vector Machines (SVM). The SVM is a
powerful classifier that uses a kernel function (i.e.,
f(x)) and transforms data into the feature space. The
SVM finds W in such way that the distance from the
nearest sample in the feature space to W is maximized
[31].
In this study, we used the radial basis function
(RBF) kernel in the OSU SVM Toolbox for multiclass
classification.
Cross-Validation Method. The leave-one-out (LOO)
is one of the most reliable cross-validation methods
where one sample in each set is used for the test, and
others are used for training. This process is repeated
until all data get a label; as a result, the accuracy is
the number of true labeled data in all sets to the total
number of data.
In our study, the LOO cross-validation method was
chosen in order to get robust and reliable results.
RESULTS
Analysis of the Differences between ERP
Waveforms. Differences between the ERP waveforms
in four states (i. e., aware-unaware in the attentive
and inattentive state and attentive-inattentive, and the
conscious and unconscious states) were considered for
all eight recording channels. Two samples observed in
leads O1 and O2 are shown in Figs. 3 and 4.
As is seen in Fig. 3, increased positivities in P100
and P300 waves are noticeable in lead O1, while
increased negativity in the N150 should be examined
under attentive condition. Under inattentive condition,
increased positivity in the P100 is observed, but
its amplitude is lower than that under attention
condition. In aware condition, increased positivity in
the P300 wave and increased negativity in the N150
are observable. Under unaware condition, increased
positivity in the P300 is obvious, but its amplitude is
lower than that under aware condition, and an increase
in negativity in the N150 seems to be suspicious.
It is seen in Fig. 4 that increased positivities in the
P100 and P300 waves are noticeable in channel O2
under attentive condition. Under inattentive condition,
increased P100 positivity is observable, and its
amplitude is higher than that under attentive condition.
Under aware condition, an increase in positivity in
the P300 and a corresponding negativity increase in
the N150 are observable. Under unaware condition,
increased positivity in the P300 should be considered.
In a similar way, increased positivities in the P100
and P300 in lead PO8 are noticeable under attentive
condition. Under inattentive condition, increased
positivity in the P300 component is observable, and
its amplitude is higher than in the case of attentive
condition. Under aware condition, increases in
positivity in the P300 and that in the N150 are
observable. Under unaware condition, an increase in
positivity in the P300 seems doubtful.
In lead PO7 under inattentive condition, increased
positivities in the P100 and P300 are obvious. Under
inattentive condition, increased positivity in the
P300 is observable, and its amplitude is greater than
that under attentive condition. In the case of aware
condition, increased positivity in the P300 and
increased negativity in the N150 are observable. Under
unaware condition, increased positivity in the P300 is
not sufficiently clear.
In lead P6 under attentive condition, all three
correlates (i.e., greater positivity in the P100 and P300
and greater negativity in the N150) are observable.
Under inattentive condition increased positivity in
the P100 is observed, while increased positivity in
the P300 and increase negativity in the N150 are
visible under aware condition. In the case of unaware
condition, clearly increased positivity in the P300 is
seen.
Results for lead P5 were similar to those in leads
P, F4, and F3; under attentive condition all three
correlates (i.e., increased positivity in the P100 and
P300 and increase negativity in the N150) appeared.
Under inattentive condition, higher negativity in
the N150 and increased negativity within the N400
window are visible. Under aware condition, increased
NEUROPHYSIOLOGY / НЕЙРОФИЗИОЛОГИЯ.—2015.—T. 47, № 2 175
DISSOCIATION BETWEEN ATTENTION AND CONSCIOUSNESS DURING A NOVEL TASK: AN ERP STUDY
positivities in the P100 and P300 are noticeable, and
increased negativity in the N100 appeared. Under
unaware condition, clearly increased negativity in the
N150 should be examined.
Statistical Analysis . Event-related potentials
were analyzed separately for stimuli in the attentive
and inattentive visual fields by entering maximum
amplitudes from the occipital (O1, O2), parieto-
temporal (PO7, PO8), parietal (P5, P6), and frontal
electrodes (F3, F4) into masking (masked vs.
unmasked) vs. targethood (target vs. nontarget)
analyses of variance (ANOVAs). Such separate
analyses were considered as justified because
attention vs. masking analyses separately for targets
and nontargets presented to the participants showed
several interactions between attention and masking
within the P100, N150, and P300 time windows,
indicating that attention and consciousness possess
distinct electrophysiological correlates.
Peak amplitudes for the P100, N150, and P300
time windows were extracted, and the ANOVA test
(P < 0.05) was performed separately in both aware
and unaware states for all recording channels. The
respective results for consciousness correlates are
shown in Table 1 and for attention correlates in
Table 2.
Considering the left side of Table 1. correlates
of consciousness under inattentive condition, are
increased positivity within the P100 window for
all leads (P < 0.05) except frontals and increased
negativity for the N150 window for channels F3
and F4. So, the correlates of consciousness in the
inattentive state are increase within the P100 window
in parietal and occipital leads and also increase in
negativity within N150 window in frontal leads.
Regarding to right side of Table 1, it is clear that
-200 0 200 400 600 800 1000
-4
-2
0
2
4
6
8
10
12 uv uv
uvuv
ms ms
ms ms
-200 0 200 400 600 800 1000
-4
-2
0
2
4
6
8
10
12
-200 0 200 400 600 800 1000
-4
-2
0
2
4
6
8
10
12
-200 0 200 400 600 800 1000
-4
-2
0
2
4
6
8
10
12
F i g, 4. Grand-average difference waveforms in lead O2 (aware-unaware conditions)– under different conditions (A-D). designations are
similar to those in Fig. 3. Under attentive condition, increased positivities in the P100 and P300 are noticeable. Under inattentive condition,
increased positivity in the P100 is observable, and its amplitude is larger than that under attentive condition. Under aware condition, increased
positivity in the P300 and increased negativity in the N150 are observable. Under unaware condition, increased positivity in the P300 should
be considered.
Р и с. 4. Усереднені диференціальні хвилі, що реєструвались у відведенні O2 за умов усвідомлення–неусвідомлення в різних
умовах (A–D).
NEUROPHYSIOLOGY / НЕЙРОФИЗИОЛОГИЯ.—2015.—T. 47, № 2176
R. DAVOODI, M. H. MORADI, and A. YOONESSI
correlates of consciousness under attentive condition
look like increased positivity in the P100 and P300
waves in all channels and also increased negativity
within the N150 window just for P5, F4, and F3
channels. Attending to both conditions, it is obvious
that the only correlate of consciousness is increased
positivity in the P100 wave for parietal and occipital
leads, the phenomenon called early consciousness
positivity (ECP).
The analogous method is taken for ERP correlates
of attention. The P300 wave increases in all relevant
channels in both conscious and unconscious
states, except left occipital and frontal leads in the
unconscious state, and also there is an increase in
negativity for the N150 for parietal leads. The other
component variation (i.e., increased positivity in the
P100) is, however, not found in all channels and thus
should be deleted. As a result, the only significant
correlate of attention is increased positivity within the
P300 window, which is called late top-down positivity
(LTP).
Feature Classifications of the ERP Components.
Figure 5 i l lustrates the results for different
classifications. As is shown, N150 component features
have the best classification accuracies compared to
two other groups (72.22% vs. 64.44 and 66.11%).
This means that this ERP component has the strongest
capability to separate the respective four classes.
T a b l e 1. ANOVA Results for Consciousness Correlates
Т а б л и ц я 1. Результати використання ANOVA для корелятів усвідомлення
Lead
Consciousness Correlates
Inattentive Attentive
P100 P300 N150 P100 P300 N150
F(1,88) P F(1,88) P F(1,88) P F(1,88) P F(1,88) P F(1,88) P
O1 27.59 0.000 0.94 0.330 1 0.300 18.2 0.000 147.7 0.000 1.45 0.230
O2 29.29 0.000 0.37 0.500 0.6 0.440 20.26 0.000 162.21 0.000 0.19 0.660
PO8 28.87 0.000 0.75 0.380 1.83 0.170 16.6 0.000 177.49 0.000 0.18 0.670
PO7 24.2 0.000 0.39 0.530 0.82 0.360 11.56 0.001 206 0.000 1.64 0.204
P6 12.7 0.000 0.35 0.550 0.01 0.927 5.98 0.016 175.49 0.000 2.96 0.080
P5 7.08 0.009 3.25 0.070 1.65 0.200 5.62 0.020 123.51 0.000 6.68 0.010
F4 1.05 0.300 2.12 0.150 55.24 0.000 12.64 0.000 38.37 0.000 21.75 0.000
F3 2.78 0.099 5.1 0.026 62.06 0.000 10.06 0.002 46.86 0.000 22.1 0.000
Footnote: Cases with P < 0.05 are shown dashed
T a b l e 2. ANOVA Results for Attention Correlates
Т а б л и ц я 2. Результати використання ANOVA для корелятів уваги
Leads
Attention Correlates
Unconsciousness Consciousness
P100 P300 N150 P100 P300 N150
F(1,88) P F(1,88) P F(1,88) P F(1,88) P F(1,88) P F(1,88) P
O1 1.56 0.200 7.84 0.006 3.64 0.059 1.73 0.190 159.8 0.000 12.24 0.000
O2 0.89 0.340 2.41 0.120 0.59 0.440 1.24 0.260 140.34 0.000 2.8 0.098
PO8 4.42 0.038 6.87 0.010 3.55 0.060 1.73 0.190 159.65 0.000 8.86 0.003
PO7 2.2 0.140 7.26 0.008 0.9 0.346 0.05 0.828 211.28 0.000 7.5 0.007
P6 3.17 0.070 9.34 0.003 4.41 0.030 0.42 0.510 209.67 0.000 13.78 0.000
P5 3.08 0.0825 7.63 0.007 3.97 0.049 2.08 0.150 210.07 0.000 8.88 0.003
F4 0.03 0.860 1.17 0.280 9.63 0.002 16.5 0.000 59.28 0.000 2.5 0.120
F3 0.02 0.899 0.07 0.780 6.69 0.011 17.65 0.000 76.22 0.000 0.44 0.500
Footnote: Cases with P < 0.05 are shown dashed
NEUROPHYSIOLOGY / НЕЙРОФИЗИОЛОГИЯ.—2015.—T. 47, № 2 177
DISSOCIATION BETWEEN ATTENTION AND CONSCIOUSNESS DURING A NOVEL TASK: AN ERP STUDY
Eleven features were selected during the feature
selection process from all feature groups and were
used for total fourfold classification; the results
are shown in the last column of Fig. 5. It shows the
classification results when all three feature groups
are used (i.e., P100, N150, and P300). As we can see,
a combination of all feature groups provides better
results (81.67%). This implies that although one
of these feature sets is the best, the other sets have
features with meaningful differences in fourfold states
of attention and consciousness and can improve the
results. As we can see, the best results are obtained
from the SVM classifier.
Channel Prioritization and Classification.
Another classification performed in this study was
channel prioritization. We have used all feature groups
for each channel separately, and the classification
accuracy results are shown in Fig. 6. As can be seen in
this figure, the best channels are PO7 and PO8, while
F4 is the worst. By comparing the between-channel
results and total classification results, it is clear that
all channels bring some useful information for class
dissociation. Another proof for this fact is that features
that have been selected during the feature selection
process belong to all channels.
DISCUSSION
In summary, the task proposed by our group has the
following advantages and remarks in comparison to
the previous works.
We could manipulate top-down attention and
consciousness in two separate modes and produce
LDA 66.11%
P100
0
10
20
30
40
50
60
70
80
90
%
P300 AllN100-N200
71.67% 80.56%64.44%
66.11% 72.22% 81.67%64.44%
63.33% 72.22%
72.22%
PO8
56
58
60
62
64
66
68
70
72
74
%
PO7 O1 O2 F3 F4 F6 F5
71.11%
72.78%
72.78% 69.44% 66.67% 64.44% 64.44%67.78%
62.22% 66.11%66.11%
62.78% 65.56%65.56%
67.78%
68.33%72.78% 70% 70%
70%
68.33% 65.56% 68.89%
76.67%55%KNN
SVM
F i g. 5. Classification results for LDA,
SVM, and KNN classifiers for the P100,
N150, P300 and all feature groups.
Р и с. 5. Результати класифікації з
використанням класифікаторів LDA,
SVM та KNN для компонентів P100, N150
та P300 при всіх умовах.
LDA
KNN
SVM
F i g. 6. Classification results for LDA, SVM, and KNN classifiers for all feature groups in different channels.
Р и с. 6. Результати класифікації з використанням класифікаторів LDA, SVM та KNN для всіх умов при відведеннях по різних
каналах.
NEUROPHYSIOLOGY / НЕЙРОФИЗИОЛОГИЯ.—2015.—T. 47, № 2178
R. DAVOODI, M. H. MORADI, and A. YOONESSI
four conditions. As is emphasized by Boxtel et al. [4],
manipulation of attention and consciousness should be
done independently. Thus, we have for the first time
employed in our study different target and nontarget
image types, and the latter was done by utilization of
masking.
Top-down attention was manipulated by considering
three types of target stimuli (human faces, fruits, and
scenes), and consciousness was manipulated using
masked and unmasked stimuli.
Using both types of human face and fruit images as
targets eliminated the undesirable effects of attending
to one special type of targets (e.g., face targets).
Another approach was considering different
electrophysiological correlates of attention and
consciousness obtained separately. Correlates were
introduced for each concept in the presence and
absence of the other by considering difference
waveforms:
ANOVA results revealed that correlates of
consciousness under inattentive condition, namely
increased positivity within the P100 window for all
channels except frontals and increased negativity
within the N150 window for frontal leads. Correlates
of consciousness under attentive condition increased
positivities in the P100 and P300 waves for all
channels and also increased negativity within the N150
window just for P5, F4, and F3 channels. Attending
to both conditions, it is obvious that correlates of
consciousness increased P100 positivity, called early
consciousness positivity (ECP), in parietal, occipital
and parieto-occipital channels and increased negativity
within the N150 window just for frontal leads.
A similar approach was employed for ERP correlates
of attention. The P300 increases in all relevant
channels in both conscious and unconscious states
except left occipital and frontals in the unconscious
state, and also there is an increase in negativity for the
N150 in parietal leads. Other component variation (i.e.
increased positivity for P100) did not demonstrate the
same results for all channels and thus is deleted.
As a result, the only correlates of attention are
increased positivity within the P300 window, which
is called late top-down positivity (LTP) and increased
negativity for the N150 in parietal leads.
Employing three state-of-the-art classifiers, we
quantified our results, i.e., features corresponding to
all three time components were extracted, gathered,
and used for classification after the feature selection
process. The best result was obtained by SVM
(81.67%). This high-accuracy result reveals the power
of the task and also the respective time features.
Features corresponding to each time window were
classified separately, and the highest-accuracy result
belonged to the features corresponding to the N150
window. As a result, this time window should be
considered the most meaningful one.
Another processing was done for choosing the more
relevant channels. In this respect, features for each
channel were classified singly and channels PO7 and
PO8 were introduced as the most appropriate ones.
The findings from this experiment support our
hypothesis, namely, that one should distinguish more
carefully between different types of attention, as well
as different forms of consciousness, when describing
the relationship between attention and consciousness.
While bottom-up attention seems to be strictly
connected to awareness, top-down attention is not
necessary for consciousness. In other words, there
are conditions under which top-down attention or
awareness can occur in the near absence of the other.
Our results provide additional insight into the
relationship between the above-mentioned two
phenomena. Due to the novelty of our task and type
of processing used, exact comparison cannot be
done, but there are considerable similarities (and
also differences) between our results and those of the
respective previous works [11, 18, 19, 22, 27, 32].
Although there were P300 and N150 variations in
the attentive field similar to [11, 33], we could report
only the P100 for parietal, occipital, and parieto-
occipital channels and increased negativity within the
N150 window just for frontal leads as ERP correlates
of consciousness because we have considered both
attentive and inattentive fields. The affected areas
for consciousness were parietal, occipital, parieto-
occipital, and frontal areas. Again, it was expected
to have the N100-200 component as a correlate of
attention (according to [11]),but in our study, the
N150 component was valid for only parietal areas,
and also the P300 component was valid for parietal
and parieto-occipital and right occipital ones. So, as
was discussed in [4], “attention to a stimulus or an
attribute of this stimulus is neither strictly necessary
nor sufficient for the stimulus or its attribute to be
consciously perceived.”
A n o t h e r a p p r o a c h c o n s i d e r s d i f f e r e n t
electrophysiological correlates of attention and
consciousness obtained separately. These ERP
correlates were obtained from difference waves (aware/
unaware) in the presence and absence of attention for
correlates of consciousness and (attentive/inattentive)
NEUROPHYSIOLOGY / НЕЙРОФИЗИОЛОГИЯ.—2015.—T. 47, № 2 179
DISSOCIATION BETWEEN ATTENTION AND CONSCIOUSNESS DURING A NOVEL TASK: AN ERP STUDY
in the presence and absence of consciousness for
correlates of attention.
Here, we have tried to quantify the distinction by
means of classification the ERP component features
extracted from ERP signals in four states, which are
unconsciousness with no attention, consciousness
without attention, attention without consciousness,
and consciousness with attention. These features are
related to the P100, N150, and P300 components,
which are extracted according to waveform differences
and statistical results. The SVM, KNN, and LDA
classifiers have been used for this purpose, and the
results revealed that all three P100, N150, and P300
waves are meaningful components when distinguishing
between these two processes, and N150 component
features are stronger than those of the other two
groups. This is the first time that these quantification
results are mentioned. Employing such simple features
corroborates our assumption on the inherent separation
between the above two phenomena.
Another approach is related to meaningful channels.
The results revealed that channels PO7 and PO8 are
the best leads for distinguishing between attention and
consciousness manifestations, while all channels have
meaningful ERP component features, which can help
in the classification.
Acknowledgment The authors thank all participants for
their contribution to this study.
All procedures were in accordance with the ethical standards
of the responsible committees on human experimentation
(institutional and national) and with the Helsinki Declaration
of 1975, as revised in 2000 (5). Informed consent was obtained
from all subjects included in the study. All participants were
volunteers; they were informed in detail on the pattern of the
experiment.
The authors of this study, R. Davoodi, M. H. Moradi, and
A. Yoonessi, confirm that the research and publication of
the results were not associated with any conflicts regarding
commercial or financial relations, relations with organizations
and/or individuals who may have been related to the study, and
interrelations between co-authors of the article.
Р. Давуді1, М. Х. Мораді1, А. Йунессі2
ДИФЕРЕНЦІАЦІЯ ЕФЕКТІВ УВАГИ ТА
УСВІДОМЛЕННЯ В НОВОМУ ТЕСТІ: ДОСЛІДЖЕННЯ
З ВИКОРИСТАННЯМ ПОТЕНЦІАЛІВ, ПОВ’ЯЗАНИХ З
ПОДІЄЮ
1 Технологічний університет Аміркабір (Тегеранська
Політехніка), Тегеран (Іран).
2 Іранський національний центр з вивчення аддикцій,
Інститут досліджень когнітивної сфери та Школа передо-
вих технологій в медицині при Тегеранському університеті
медичних наук, Тегеран (Іран).
Р е з ю м е
Як вважають, увага та усвідомлення тісно поєднані, проте
результати недавніх досліджень дають підстави думати, що
один із цих процесів може реалізуватися за відсутності дру-
гого. У новітніх експериментах виявилося, що спостерігачі
можуть приділяти увагу невидимому (неусвідомлюваному)
стимулу і що стимул може чітко розпізнаватися
(диференціюватися) за відсутності уваги. Ми запропонували
нове психофізіологічне завдання для дослідження нервових
корелятів змін рівнів уваги та усвідомлення. Цільові ЕЕГ-
потенціали, пов’язані з подією (ППП), відводили в умовах
пред’явлення замаскованих та незамаскованих візуальних
патернів. Часові характеристики компонентів P100, N150 та
P300 визначалися роздільно для відведень по кожному ЕЕГ-
каналу. З урахуванням цих характеристик були використані
певні загальні класифікатори для параметрів, спостере-
жуваних у всіх чотирьох стимуляційних станах. У нашо-
му тесті ефекти уваги та усвідомлення могли бути успішно
розділені відповідно до їх нервових ЕЕГ-корелятів. Згідно
з отриманими результатами, згадані вище компоненти ППП
змінюються залежно від того, як проявляється увага або
усвідомлення. Роздільне порівняння відмінностей між хви-
лями ППП для кожної з умов дозволило виявити нові ко-
реляти уваги та усвідомлення, що відбиваються в ППП.
Показано також, що тім’яно-потиличні кортикальні зони є
структурами, найбільшою мірою пов’язаними з дисоціацією
ефектів уваги та усвідомлення. Як ми вважаємо, ці кореля-
ти вперше представлені з використанням методики, котра
дозволяє їх розрізнити, і наведені дані щодо точності такої
диференціації.
REFERENCES
1. P. M. Merikle and S. Joordens, “Parallels between perception
without attention and perception without awareness,”
Conscious. Cognit., 6, 219-236 (1997).
2. V. A. F. Lamme, “Why visual attention and awareness are
different,” Trends Cognit. Sci., 7, 12-18 (2003);
3. B. Bahrami, N. Lavie, and G. Rees, “Attentional load
modulates responses of human primary visual cortex to
invisible stimuli,” Curr. Biol., 17, 509-513 (2007).
4. J. J. A. V. Boxtel, N. Tsuchiya, and C. Koch, “Consciousness
and attention: on sufficiency and necessity,” Front. Psychol.,
1, 217 (2010).
5. A. B. Chica and P. Bartolomeo, “Attentional routes to
conscious perception,” Front. Psychol., 3, 1 (2012).
NEUROPHYSIOLOGY / НЕЙРОФИЗИОЛОГИЯ.—2015.—T. 47, № 2180
R. DAVOODI, M. H. MORADI, and A. YOONESSI
6. M. I. Posner, “Attention: the mechanisms of consciousness,”
Proc. Nat. Acad. Sci. USA, 91, 7398-7403 (1994).
7. J. K. O’Regan and A. Noë, “A sensorimotor account of vision
and visual consciousness,” Behav. Brain Sci., 24, 939-972
(2001).
8. L. Naccache, E. Blandin, and S. Dehaene, “Unconscious
masked priming depends on temporal attention,” Psychol. Sci.,
13, 416-424 (2002).
9. M. A. Cohen, P. Cavanagh, M. M. Chun, and K. Nakayama,
“The attentional requirements of consciousness,” Trends
Cognit. Sci., 16(8), 411-417 (2012).
10. N. Block, “Two neural correlates of consciousness,” Trends
Cognit. Sci., 9, 46-52 (2005).
11. N. Tsuchiya, N. Block, and C. Koch, “Top-down attention and
consciousness: comment on Cohen et al.,” Trends Cognit. Sci.,
16(8), 527 (2012).
12. H. Railo, M. Koivisto, and A. Revonsuo, “Tracking the
processes behind conscious perception: A review of event-
related potential correlates of visual consciousness,” Consc.
Cognit., 20, 972-983 (2011).
13. R. W. Kentridge, C. A. Heywood, and L. Weiskrantz,
“Spatial attention speeds discrimination without awareness in
blindsight,” Neuropsychologia, 42, 831-835 (2004).
14. R. Kanai, N. Tsuchiya, and F. A. J. Verstraten, “The scope and
limits of top-down attention in unconscious visual processing,”
Curr. Biol., 16, 2332-2336 (2006).
15. A. Mack and I. Rock, Inattentional Blindness, Cambrige, MIT
Press, 1998.
16. F. F. Li, R. VanRullen, C. Koch, and P. Perona, “Rapid natural
scene categorization in the near absence of attention,” Proc.
Nat. Acad. Sci. USA, 99, 9596-9601 (2002).
17. A. R. E. Vandenbroucke, I. G. Sligte, and V. A. F. La-
mme, “Manipulations of attention dissociate fragile
visual short-term memory from visual working memory,”
Neuropsychologia, 49, 1559-1568 (2011).
18. C. Koch and N. Tsuchiya, “Attention and consciousness:
two distinct brain processes,” Trends Cognit. Sci., 11, 16-22
(2007).
19. F. Mormann and C. Koch, “Neural correlates of consciousness,”
Scholarpedia, 2, 1740 (2007).
20. U. Kriegel and I. P. Thagard, “Consciousness: Phenomenal
consciousness, access consciousness, and scientific practice,”
in: Handbook of Philosophy of Psychology and Cognitive
Science, P. Thagard (ed.), Amsterdam, North-Holland (2006),
pp. 195-217.
21. S. M. Hsu, N. George, V. Wyart, and C. Tallon-Baudry,
“Voluntary and involuntary spatial attentions interact
differently with awareness,” Neuropsychologia, 49, 2465-2474
(2011).
22. J. J. A. V. Boxtel, N. Tsuchiya, and C. Koch, “Opposing effects
of attention and consciousness on afterimages,” Proc. Nat.
Acad. Sci. USA, 107, 8883-8888 (2010).
23. A. Schurger, A. Cowey, J. D. Cohen, et al., “Distinct and
independent correlates of attention and awareness in a
hemianopic patient,” Neuropsychologia, 46, 2189-2197 (2008).
24. A. Zeman, Consciousness, Wiley Online Library (2006),
doi: 10.1002/0470018860.s00732.
25. M. Koivisto and A. Revonsuo, “Electrophysiological correlates
of visual consciousness and selective attention,” NeuroReport,
18, 753-756 (2007).
26. M. Koivisto and A. Revonsuo, “Prechange event-related
potentials predict change blindness in various attention
conditions,” NeuroReport, 16, 869-873 (2005).
27. M. Koivisto, A. Revonsuo, and M. Lehtonen, “Independence
of visual awareness from the scope of attention: An
electrophysiological study,” Cerebral Cortex, 16, 415-424
(2006).
28. M. Koivisto and A. Revonsuo, “The role of selective attention
in visual awareness of stimulus features: Electrophysiological
studies,” Cogn., Affect., Behav. Neurosci., 8, 195-210 (2008).
29. S. J. Luck and E. S. Kappenman, The Oxford Handbook of
Event-Related Potential Components, Oxford, Oxford Univ.
Press (2011).
30. I. Kalatzis, N. Piliouras, E. Ventouras, et al., “Design and
implementation of an SVM-based computer classification
system for discriminating depressive patients from healthy
controls using the P600 component of ERP signals,” Comput.
Methods Programs Biomed., 75, 11-22 (2004).
31. C. M Bishop, Pattern Recognition and Machine Learning,
Springer, New York (2006).
32. C. Koch and N. Tsuchiya, “Attention and consciousness:
Related yet different,” Trends Cognit. Sci., 16, 103-105 (2012).
|