Brain Topography, Volume 13, Number 3, 2001
ISBET 2000: The 11th World Congress of the International
Society for Brain Electromagnetic Topography,
November 16-19, 2000
Organizer: Maurer and Dierks
In contrast, the parietal P300 source strength increased af-
Low-Resolution Brain Electromagnetic Tomography
ter acute (day 1), subacute and superimposed infusion
(LORETA) of Cognitive Event-Related Potentials in
(day 15) of Actovegin as compared to placebo.
Age-Associated Memory Impairment and Effects of
Nootropic Drug Therapy. - P. Anderer*, H.V.
Some Advances in MEG and EEG Source Imaging
Semlitsch*, B. Saletu*1 and R.D. Pascual-Marqui+
Methods. - S. Baillet (Cognitive Neuroscience and
(*Department of Psychiatry, University of Vienna,
Brain Imaging Laboratory, Hôpital de la Salpêtrièr,
Austria and +The KEY Institute for Brain-Mind Re-
search, Department of Psychiatry, University Hospital,
Most of the approaches to MEG and EEG source esti-
mation can be reduced to 1) parametric techniques con-
Cognitive event-related potentials were recorded in
sisting in the localization of discrete equivalent source
an auditory two-tone paradigm in 43 patients aged 51-79
models like current dipoles and 2) imaging techniques
years with the diagnosis of age-associated memory im-
that produce estimates of the cortical current density,
pairment (AAMI). Eighteen patients participated in a dou-
usually given anatomical information from the MRI.
ble-blind, placebo-controlled study on the effects of
Parametric approaches to MEG/EEG source localization
Actovergin®, a protein-free metabolically active
can rapidly generate a physical model of neural current
hemoderivative improving oxygen and glucose utiliza-
generators using a limited number of parameters. How-
tion. Low-resolution brain electromagnetic tomography
ever, physiological interpretation of these models is of-
(LORETA) revealed medial (frontal and parietal) and lat-
ten difficult. Image-based approaches to the inverse
eral (dorso- and ventrolateral prefrontal, middle/superior
problem provide a direct estimate of cortical current gen-
temporal, posterior superior temporal/inferior parietal)
erators, but computationally expensive non-linear meth-
P300 sources. Significant reductions in LORETA source
ods are required to produce focal sources. We will review
strength in AAMI were mainly found medially frontally,
some of the basic principles related to both of these ap-
right dorsolaterally prefrontally and right inferiorly
proaches and will present some recent advances for a
parietally. Since these anatomically highly interconnected
better characterization of the sources of the MEG/EEG.
brain regions in the right hemisphere are part of a network
This includes the quantification of the spatial extension
associated with sustained attention, the results speak for a
of the cortical areas involved by cortical remapping
decline in attentional resource capacity in AAMI patients.
through a multi-step approach involving discrete mod-els for the sources and their cortical counterparts usingnon-linear imaging constraints. We will also present
Copyright 2001 Human Sciences Press, Inc.
some application of multi-resolution cortical imaging
techniques. We will finally discuss about the challenge of
tency did not differ between HD and control children.
comparing and possibly combining an f-MRI analysis
LORETA source solutions revealed posterior P300
sources which were attenuated in HD children.
Conclusions: Covert attention to cues preceding po-
Slow Cortical Potential Manifestations of Anticipatory
tential targets is already impaired in HD, determines
Fear and Attention. - K.B.E. Böcker*, J.M.P. Baas*, M.N.
performance with subsequent targets, and activates the
Verbaten* and H.M. Huizenga+ (*Utrecht University,
posterior brain. These findings extend previous results
Department of Psychopharmacology, The Nether-
obtained for children with Attention Deficit Hyperactiv-
lands, +University of Amsterdam, Department of Psy-
ity Disorder and demonstrate the usefulness of clinical
chology, The Netherlands).
multicenter studies using cognitive brain mapping.
The Contingent Negative Variation (CNV) is usu-
ally recorded in experimental tasks which induce motor
Evoked and Induced EEG Responses Under Median
preparation. The non-motor CNV, or Stimulus Preceding
Nerve Stimulation. - A. Brovelli*, E. Daprati*, R.J.
Negativity (SPN), is most readily observed in S1-S2 para-
Naranjo*, A. Muzur*, R. Budai+ and P.P. Battaglini*^ (*
digms involving the presentation of S2 stimuli with af-
SISSA-ISAS, Cognitive Neuroscience Sector, Trieste,
fective valence. The SPN indeed manifested itself in a
Italy, +Neurology Unit, Dept. of Neuroscience, OSMM,
previous study which employed the anticipatory fear
Udine, ^B.R.A.I.N. Centre for Neuroscience, Dept. of
paradigm.The present study intended to compare antici-
Physiology and Pathology, University of Trieste, Italy).
patory fear (S2 = shock) and anticipatory attention (S2 =
Two main categories of EEG responses can be identi-
to-be-counted weak vibration). Experimental conditions
fied: evoked and induced. The present study aims at in-
were otherwise equal. The S1-S2 interval was 1.2 - 1.4 s.
vestigating the relationship between evoked and
Preliminary results (N=7) show that the SPN (measured
induced responses after median nerve stimulation. Scalp
at FCz) was significant in both the anticipatory fear (-4.8
EEG was recorded using a 32-channel system on 4 volun-
± 2.0 µV) and the anticipatory attention (-2.5 ± 1.0 µV)
teers under medial nerve stimulation and integrated
conditions. However, the latter difference between the
with anatomical MRI data. Methods based on wavelet
two anticipatory conditions was not statistically signifi-
and Hilbert transform allow the computation of a
cant. Details on the topographical distribution of the SPN
time-frequency (TF) representation of the instantaneous
in both conditions will be presented.
amplitude of a signal. TF representations of the instanta-neous amplitude averaged across trials and the TF repre-
Multicenter P300 Brain Mapping of Inattention to Cues
sentation of the instantaneous amplitude of averaged
i n H y p e r k i n e t i c C h i l d r e n . - D . B r a n d e i s , T .
evoked potentials were computed so to extract the
Banaschewski*, L. Baving~, P. Georgiewa+, B. Blanz+,
evoked + induced and evoked responses, respectively.
M.H. Schmidt~, A. Warnke^, H.C. Steinhausen, A.
An induced mean amplitude increase in the beta band
Rothenberger* and P. Scheuerpflug^ (Departments of
was found approximately 500 msec after stimulation
Child and Adolescent Psychiatry Universities of
over the controlateral somatosensory area, whereas an
Zürich, *Göttingen, +Jena, ^Würzburg, ~ Central Insti-
alpha band decrease was found over extended regions.
tute for Mental Health, Mannheim).
Preliminary results are presented and discussed accord-
Objective: Multicentric 19-channel EP brain mapping
data was recorded from children with and without
hyperkinetic disorders (HD) during a cued continuous per-formance test (CPT) to examine attentional P300 effects for
EEG Source Identification Related to a Frequency
deficits in HD children and for consistency across clinics.
Band: 3D Located Fluvoxamine Effects During REM
Method: CPT performance and P300 maps to cues
Sleep. - A. Coatanhay, L. Soufflet, L. Staner, M.
and distractors from 148 children were tested for differ-
Toussaint, P.H. Boeijinga and J.-P. Macher(FORENAP
ences between age- and sex-matched groups with and
- Centre Hospitalier - Rouffach - France).
without HD (N=57 each), and between clinics (N=5).
We have developed a methodology, based on the
Results: Consistency of performance, P300 mea-
LORETA (Low Resolution Tomography) algorithm, to
sures, and amplitude increases plus topographic
solve inverse EEG problems related to a frequency band.
changes for P300 maps to cues compared to distractors
We have studied the EEG source localisation during
was established across most clinics. HD children missed
REM sleep (paradoxical sleep) for four frequency bands:
more targets, made more false alarms, and had larger N1
delta (0.5-3.5Hz), theta (4-7.5Hz), alpha (8-12.5Hz) and
and smaller P3b map amplitudes after cues than control
children. Cue-P300 amplitude correlated positively with
Our study compared EEG source localisation for
detecting subsequent targets. P300 topography and la-
each frequency band before and after one week of treat-
ment with the antidepressant fluvoxamine (100 mg).
ally identified with eyeblink and other artifacts. One
Subjects were 9 young (25.8 ±6.6 years old) healthy vol-
component clearly represented the spike and sharp wave
unteers. EEG was recorded using 21 electrodes based on
activity. The EEG was reconstructed with only this com-
ponent present, and subjected to single dipole modeling.
EEG activity was reduced during REM sleep after
The residual variance over the entire epoch was lower
the fluvoxamine treatment in all frequency bands. In ad-
than that of any single, unaveraged spike or sharp wave
dition, tomographic data evidenced that modification
selected from the original recording. Thus ICA, with no a
seems different between high (theta, alpha, beta) and low
priori dipole assumptions, was able to extract a very di-
(delta) frequencies. The reduction was more significant
pole-like spatiotemporal component from raw EEG.
(Student test p<0.05) in the right inferior parietal lobe forlow frequencies and in the central occipital cortex for
Frequency Domain EEG Source Locations During Arm
Levitation Under Hypnosis: A Pilot Study. - P.L. Faber,
P. Wohlgemuth, L.R.R. Gianotti and D. Lehmann (The
Topographic EEG Alterations in Cognitive Impair-
KEY Institute for Brain-Mind Research, University
ment. - T. Dierks*+, V. Jelic^, L. Frölich+, L.O.
Hospital of Psychiatry, Zurich, Switzerland).
Wahlund^ and K.Maurer+ (*Dept. of Psychiatric
27-channel EEG was recorded from 4 right-handed
Neurophysiology, University Hospital of Clinical Psy-
subjects during 2 conditions: left arm raised under hyp-
chiatry Bern, Switzerland, +Dept. of Psychiatry, Univ.
notic suggestion, and left arm raised on command before
of Frankfurt, Germany, ^Alzheimer Research Center,
and after hypnosis. Locations of intracerebral 3-D source
Karolinska Institute, Stockholm, Sweden).
gravity centers of seven independent EEG frequency
Many EEG parameters like peak frequency, the
bands were computed (using FFT-Dipole- Approxima-
amount of slow wave and alpha activity are abnormal
tion) for each 2-second analysis epoch (total 324), and av-
and correlate with the severity of cognitive functions in
eraged over epochs for each condition and subject.
Alzheimer Disease (AD). However, due to technical limi-
Exploratory statistics (two-tailed paired t-tests, N=4 sub-
tations conventional frequency analysis does not allow to
jects) showed significant differences of source locations
draw any conclusions about spatial pattern of cerebral
between conditions for delta (1.5-6Hz), alpha1
pathophysiology. We have quantified spatial EEG alter-
(8.5-10Hz), beta1 (12.5-18Hz) and beta2 (18.5-21Hz). Un-
ations in around 200 patients suffering from Alzheimer
der hypnosis, delta was more posterior, alpha1 more left,
disease and cognitive impairment using novel analysis
beta1 more anterior, and beta2 more right. Disregarding
lateralizations (no right arm data!), the results indicate
(FFT-approximation and adaptive segmentation). In AD
that during hypnotic arm levitation, along the ante-
patients, a shift of alpha- and beta-EEG generators and
rior-posterior brain axis, inhibitory activity (delta)
center of gravity of the microstate topography towards
shifted to more posterior areas but excitatory activity
frontal brain regions was observed. The amount of this
shift correlated with the degree of dementia. The
Grant #670806 from Institut für Grenzgebiete der
anteriorization of EEG-generators and microstates reflect
the pathological spatial pattern of glucose metabolism inAD. A similar spatial pattern of EEG-alterations could be
Cannabis-Induced Topographic Changes in Pre/Post
found in asymptomatic family members with genetic AD.
EEG Activity During Rest and Music Perception. - J.
Fachner*, D. Aldridge* and E. David+ (*Institute for
Dipole Modeling of an Epileptiform Component Ex-
Music Therapy, UW/H, +Institute of Physiology,
tracted by Independent Component Analysis of 128
Channel EEG. - K.J. Eriksen*, K.E. Karnofski+ and C.
In scientific literature cannabis is found to change or
Davey+ (*Legacy Good Samaritan Hosp, Portland, OR,
enhance time-, space-, body- and movement perception,
+Electrical Geodesics, Inc, Eugene, OR).
emotion, imagery and association patterns. Presented
Independent Component Analysis (ICA) is a tech-
here is an explorative study on Cannabis and Music Per-
nique for separating mixed, statistically independent
ception, conducted in a qualitative and quantitative way
spatiotemporal signal components. We wished to exam-
in a habituated setting. EEG-Brainmapping Data (rest;
ine whether it could remove background "brain noise"
pre/post listening; 28 EEG traces; smoked Cannabis φ 20
from unaveraged epileptiform activity as a preprocess-
mg D9 THC) were averaged and treated with a T-test and
ing step prior to source analysis. ICA was applied to 128
channel EEG data obtained from a patient with right
Compared to pre-THC-rest and pre-THC-Music in
fronto-temporal spikes and confounding eyeblinks. We
the post-THC-Music-EEG a rise of Alpha percentage and
obtained 128 components, some of which could be visu-
power was observed in the parietal cortex on four sub-
jects, while other frequencies decreased in power. De-
guage processing was followed by a bifrontal
creased amplitudes could represent a decreased
hypooxygenation, which was interpreted as a relative
cell-firing mode caused by cannbinoidreceptor mecha-
prefrontal hypoperfusion due to a blood volume redistri-
nism. Comparing pre/post music EEGs, differences
bution in favor of language-related temporal brain areas.
(p<0.01) were found in the right frontotemporal cortex
A visual control task did not affect the prefrontal concen-
on Theta and on Alpha in the left occipital cortex.
trations of O2Hb and HHb at all. An investigation with a
Changes in temporal and occipital areas and in-
language-generation task (verbal fluency test) showed
creasing a-signal strength in parietal association cortex
an oxygenation overshoot in left prefrontal brain areas in
seem to represent a neural correlate of altered music per-
healthy controls consistent with the location of Broca’s
ception and hyperfocusing on the musical time-space.
area, and a loss of this hemispheric asymmetry in pa-
Alpha amplitude changes remind on "reverse Alpha"
tients with Alzheimer‘s dementia. All these findings
findings in studies with gifted individuals.
point to the ability of the non-invasive NIRS-method toassess task-specific patterns of physiological and
Event-Related LORETA- and FMR-Imaging of Motor
Response Inhibition Converge to Anterior Cingulum
Activations. - A.J. Fallgatter, M.J. Herrmann, K.P.
Electrophysiological Measures Related to Impulsivity
Lesch, B. Hamelbeck and A.J. Bartsch (Dept. of Psychia-
and Serotonergic Neurotransmission. - A.J. Fallgatter
try, University of Wuerzburg, Germany).
and M.J. Herrmann (Dept. of Psychiatry, University of
Several investigations concluded that the anterior
cingulate gyrus (ACG) of the prefrontal cortex is involved
Impulsivity is both, a very important feature in sev-
in response inhibition. In our present study, this particular
eral psychiatric conditions related to a low central
brain function was elicited experimentally by a cued ver-
serotonergic neurotransmission like aggressive behavior
sion of the Continuous Performance Test (CPT). The CPT
and suicidality, and a core symptom of frontal lobe syn-
contains both, the execution (Go-condition) and inhibition
dromes of various etiologies. ERP-parameters of re-
(NoGo-condition) of prepared motor responses. Source
sponse control elicited with the execution (Go-condition)
analyses of event-related potentials by Low Resolution
and the inhibition (NoGo-condition) of an anticipated
Electromagnetic Tomography (LORETA) revealed signif-
motor response within the Continuous Performance Test
icant electrical ACG-hyperactivity during response inhi-
might be suitable candidates for an objective measure-
bitions. In addition, BOLD-contrast sensitive echo planar
ment of impulsivity. The location of the brain electrical
imaging (mosaic EPI on 1.5 T Vision scanner: TE = 60 ms,
field in the NoGo-condition has been shown to be associ-
RT = 2083 ms, FA = 78°, 17 interleaved slices, distance fac-
ated with the presence of at least one short allele of the se-
tor 0.25, 4 x 4 x 5 mm voxel size) provided the first
rotonin transporter promoter and, therefore, with a
event-related functional MRI data of healthy volunteers
diminished serotonergic neurotransmission. The topo-
performing the CPT. General linear modelling by Statisti-
graphical ERP-parameters and the I7-impulsivity score
cal Parametric Mapping (SPM 99) disclosed significant
were determined in 22 healthy subjects. Impulsivity was
BOLD-responses within the ACG for the NoGo-condition.
correlated with both, a more anterior location of the Go-
Taken together, these findings indicate that simple motor
(r = 0.58, p < 0.01) and the NoGo-centroid (r = 0.53, p =
response inhibitions drive neuroelectrical and -metabolic
0.01). Therefore, these ERP-parameters may be candi-
dates for the measurement of the brain electrical basis ofimpulsive behavior.
In Vivo Measurement of Human Brain Function with
NIRS. - A.J. Fallgatter*, W.K. Strik+, Th.J. Mueller+ and
Electrophysiological Indications for an Altered
M.J. Herrmann* (*Dept. of Psychiatry and Psychother-
Prefrontal Response Control in Schizophrenias. - A.J.
apy, University of Wuerzburg, Germany, +University
Fallgatter*, W.K. Strik+, Th.J. Mueller+ and M.J.
Hospital of Clinical Psychiatry, Bern, Switzerland).
Herrmann* (*Dept. of Psychiatry and Psychotherapy,
Near-Infrared Spectroscopy (NIRS) is a new
University of Wuerzburg, Germany, +University Hos-
non-invasive optical method which allows an in vivo
pital of Clinical Psychiatry, Bern, Switzerland).
measurement of the concentration changes of oxy-
Prefrontal brain areas, in particular the anterior
(O2Hb) and deoxyhemoglobin (HHb) in brain tissue. In
cingulate gyrus (ACG), are supposed to be involved in the
healthy controls NIRS revealed a bifrontal activation pat-
pathophysiology of schizophrenic diseases. Recently, the
tern during performance of a Wisconsin Card Sorting
NoGo-anteriorisation (NGA) has been introduced as a
Test, and a right prefrontal activation during the Contin-
new electrophysiological measure of response control and
uous Performance Test. This hemispheric asymmetry
ACG-function. The NGA consists of a more anterior loca-
was reduced in schizophrenic patients. Moreover, lan-
tion of the brain electrical field in the event-related poten-
tials of the NoGo- compared to the Go-condition elicited
tial gradients from the inside to the outside of the head.
with a cued Continuous Performance Test (CPT). In 19
Thus, there is an optimum distribution of nodes to the
chronic schizophrenic patients the NGA was significantly
different boundaries, resulting in improved accuracy as
reduced in comparison with age- and sexmatched healthy
compared to standard uniform distributions for a given,
controls (0.07 ± 0.74 vs. 0.90 ± 0.37 electrode distances; t =
-4.36, p = 0.0001). On a qualitative level, 10 of 19 schizo-
Spherical and realistically shaped surfaces were ap-
phrenic subjects had no NGA, which was present in every
proximated by 500, 1000, 2000, and 3000 nodes each lead-
single healthy control (Chi-square (df = 1) 6.81, p < 0.01).
ing to 64 different BEM-models with 1500 to 9000 nodes in
These results are interpreted as an electrophysiological ex-
total. 81 electrodes were placed on an extended 10/20 sys-
pression of disturbed ACG-function in schizophrenic pa-
tem grid. Potential distributions of test-dipoles at 4000
tients. Subsequent studies will clarify the relationship of
random positions were calculated. Dipoles were then fit-
this electrophysiological parameter with distinct
ted using 192 different BEM-models and mislocalizations
psychopathological features, e. g. motor symptoms in
were evaluated to find the optimum node distribution.
catatonic schizophrenias or flattening of affect and drivein hebephrenic schizophrenias.
Spatial Analysis of Brain Electrical Activity in Psychi-
atric Disorders. - S. Galderisi (Department of Psychia-
The Feasibility of Simultaneously FMRI and EEG Ex-
try, University of Naples SUN, Naples, Italy).
periments. - A. Federspiel, T.J. Müller, T. Dierks, T.
Most investigations involving spatial analysis of
Koenig, D. Hubl and W.K. Strik (University Hospital of
brain electrical activity in psychiatric disorders, in anal-
Clinical Psychiatry, Department of Psychiatric
ogy with empirical work carried out in neurological pa-
Neurophysiology, Bern, Switzerland).
tients, have been looking for dysfunctional regions in
Understanding the functioning of information pro-
different psychopathological categories, trying to confirm
cesses of the human brain requires the merge of knowledge
findings provided by other brain imaging techniques. Ex-
obtained from all possible fields involved in this topic. The
amples of this approach include hypofrontality or abnor-
physiological responses related to brain activation can be
malities of temporal regions in schizophrenia and frontal
dysfunctions in obsessive compulsive disorder.
level-dependent) fMRI. The spatial resolution obtained
It will be argued that this approach does not fully ex-
with this method is sufficient. The determination of the dy-
ploit the potential resources of spatial analysis of brain
namics of the brain activation is limited by the poor tempo-
electrical activity, especially those related to its high tem-
ral resolution, however. Electrophysiologic responses are
poral resolution, i.e. information on spatiotemporal dy-
best measured by means of EEG with temporal resolution
in the subsecond range. The outcome of the combined fMRI
Psychopathological phenomena involve abnormali-
and EEG recording is thought to complement the picture of
ties of the complex and dynamic interaction of several cog-
physiologic processes. In the present study we present re-
nitive processes, often show a waxing and waning nature
sults of the simultaneously fMRI and EEG measurment of
and sometimes may be induced by drug administration or
visual evoked potential (VEP). In order to test the goodnes
environmental manipulation. They seem to be better ac-
of the EEG recording as a function of the magnetic environ-
counted for by abnormalities of functional connectivity
ment we repeaded the measurement under four different
within largely distributed cortical networks, rather than
conditions. Our results showed no changes in the ampli-
the dysfunction of individual specialized regions.
tude and in the latency of the P100 (136 ms) during the gra-
The study of the dynamic changes of spatial config-
dient boost as compared to the absence of magnetic field.
urations in time and their relations to task demands
The LORETA analysis of the positive P100 revealed the
and/or state changes is more appropriate to promote our
dominant source location in the region of V5 (MT).
understanding of neural mechanisms underlying psy-chiatric disorders and treatment interventions.
Optimized Node Distributions for Realistically
Shaped Boundary Element Method Volume Conductor
EEG Intracerebral Source Locations, LORETA Brain
Models. - M. Fuchs, M. Wagner and J. Kastner
Electric Tomography and MMSE in Alzheimer’s Dis-
(Neuroscan Labs, Lutterothstr. 28e, 20255 Hamburg,
ease. - L.R.R. Gianotti*+, D. Lehmann+, R.D.
Pascual-Marqui+ and U. Schreiter Gasser* (*Division
The Boundary Element Method (BEM) approxi-
of Geriatric Psychiatry, +The KEY Institute for
mates the different compartments of volume conductor
Brain-Mind Research, University Hospital of Psychia-
models by closed triangle meshes with a limited number
try, Zurich, Switzerland
of nodes. The shielding effect of the weakly conducting
Brain electric activity (19-channel eyes-closed resting
skull layer of the human head leads to decreasing poten-
EEG) and Minimal Mental State Examination (MMSE) was
recorded in nine patients suffering from Alzheimer type
tex of a subject such that the assumed dipoles possess
dementia (DAT) in two sessions, before and three months
rather orthogonal orientation to each other. The retinotopic
after onset of rivastigmine medication. For seven EEG fre-
map of the subject provides the information about the cor-
quency bands (delta, theta, alpha1&2, beta1, 2&3), the
responding regions in the visual field which have to be
stimulated in order to produce activity at the selected di-
FFT-Dipole-Approximation) and the 3-dimensional distri-
pole positions. We report about first results using a simple
butions of current density (via Low Resolution Electromag-
binding experiment consisting of dots moving in the same
netic Tomography "LORETA") were computed.
or opposite direction presented at two identified regions in
Comparing the two sessions, there was a significant corre-
the visual field. The outlined multi-modal approach may
lation between improved MMSE scores and more anterior
provide a new research tool for non-invasively studying
location of beta2 (r=0.69; p<0.05) and beta3 (r=0.78; p<0.05)
the binding problem in the human brain and to learn about
sources. Similar results were found computing gravity cen-
the possibilities and limitations of high-resolution
ters of the LORETA result volumes: beta2 (r=0.66; p<0.05)
spatio-temporal fMRI/EEG/MEG mapping.
and beta3 (r=0.82; p<0.05). Voxel-by-voxel LORETA im-ages showed increased activation in the right middle/su-
The Influence of Brain Anisotropy on Human EEG and
perior temporal gyrus for beta2 and beta3 with improved
MEG. - J. Haueisen*, D. S. Tuch+, C. Ramon^ and P.
MMSE, i.e., in areas known to be impaired in DAT
S c h i m p f ~ ( * B i o m a g n e t i c C e n t e r , F r i e d r i c h-
(Desgranges et al., 1998; deToledo-Morrell et al., 1997).
Schiller-University, Jena, Germany, +Massachusetts
General Hospital - NMR Center, Charlestown, MA,
A New Approach to Study Temporal Coding in Early
USA, ^Department of Electrical Engineering, Univer-
Visual Areas of the Human Brain. -R. Goebel (Dept. of
sity of Washington, Seattle, WA, USA, ~Department of
Neurocognition, Faculty of Psychology, Maastricht
Computer Science, Eastern Washington University,
Cheney, WA, USA).
Cognitive functions such as perception, memory
The electrical conductivity of brain tissue, particu-
and language are based on massively parallel informa-
larly white matter, is known to be anisotropic. However,
tion processing involving many cortical and subcortical
basically all currently used modeling approaches in
brain areas. This raises the question of how this distrib-
source localization neglect anisotropy.
uted information is (re-)integrated, and how features
The influence of gray and white matter anisotropy
from different entities (such as different visual objects)
on the human EEG and MEG was examined with a high
are kept apart. This problem is known as the "binding
resolution finite element model of a human head. We
problem". A proposed solution of the binding problem
quantified this influence by comparing simulated EEG
assumes that the exact timing of neuronal discharges
and MEG maps using three different types of volume
provides a temporal coding scheme. In this scheme dif-
conductor models: a detailed inhomogeneous model
ferent features of different entities can be separately asso-
with anisotropic conductivity tensors, a detailed
ciated through synchronous oscillatory discharges.
inhomogeneous model with isotropic conductivity val-
There is accumulating electrophysiological evidence
ues, and a three compartment model. The conductivity
from animal research that binding is indeed reflected in
tensor data for gray and white matter were obtained
synchrounous oscillatory activity above 30 Hz (gamma
from magnetic resonance diffusion tensor imaging.
band). Human studies using EEG or MEG have also re-
Anisotropic volume conduction in the brain was
cently shown that gamma band activity increases in spe-
found to have a minor influence on the topology of EEG
cific time windows after the presentation of coherent vs.
and MEG (and hence source localization) but a major in-
incoherent visual stimuli. These studies, however, have
fluence on the amplitude of EEG and MEG (and hence
so far failed to localize and separate the underlying
We propose a new approach to study synchronization
Influence of Conventional and Newer Antidepressive
phenomena in the human brain in a manner very close to
Drug Treatment on Cognitive Evoked Potentials in Pa-
the approach used in electrophysiological animal experi-
tients with Depressive Disorder using LORETA. - W.
ments. This approach requires the integration of
M. Herrmann and J. Gallinat (Laboratory of Clinical
high-resolution fMRI, high-density EEG/MEG recordings
Psychophysiology, Department of Psychiatry,
and appropriate three-dimensional cortex reconstruction
Benjamin Franklin Hospital, Free University of Berlin,
techniques. The main idea is to increase the spatio-temporal
resolution of EEG/MEG by exploiting the results of
The results of a recent investigation using LORETA
fMRI-based retionotopic mapping experiments. We first
(Low Resolution Electric Tomography) are compared
select surface patches in early visual areas on the folded cor-
with classical Pharmaco-EEG results.
In the recent investigation, the N100 and P300 compo-
studies elicited this face-specific ERP with the presenta-
nent was recorded using the auditory "oddball" paradigm
tion of smiling faces. As differences in amplitudes ac-
in 25 drug-free monopolar depressive patients before and
cording to the emotional valence of the stimuli were
after 4 weeks of antidepressive monotherapy and 25 nor-
found as early as 100 ms after stimulus presentation
mal controls matched with respect to age and sex. Pretreat-
(Pizzagalli et al., 1999), this ERP might also be affected by
ment amplitudes of the N100 (Cz) and frontal P300 (Fz) in
the emotional valence of the stimuli. In order to investi-
depressive patients were smaller as compared to controls.
gate the effect of facial expressions, event related poten-
N100 amplitude (Cz) was negatively correlated with the
tials (ERPs) elicited with sad, happy and neutral faces
Hamilton depression score (HAMD) and P300 amplitude
were compared to those elicited with buildings in 16
(Pz) correlated negatively (trend) with the HAMD.
healthy subjects. ERP-analysis confirmed a face specific
With classical evaluation of N100 and P300, no change
P160 component at vertex. In contrast to our hypothesis
of either component after 4 weeks of treatment with con-
we did not find any differences in amplitudes between
ventional or newer antidepressants was observed in con-
sad, happy or neutral facial expressions in that time seg-
trast to results reported by Gangadhar et al. (1993).
ment. The results indicate, that face perception seems to
However, the N100 and P300 are a composition of
be processed independently from the decoding of differ-
multiple coactive generators. Interesting aspects can be
demonstrated using LORETA to follow the course of ac-tion in each location separately.
Electrophysiological Cue-Reactivity in High and Low
Social Drinkers. - M.J. Herrmann, A.J. Wiesbeck, H.-G.
Right Hemispheric Shift for the Processing of Negative
Weijer, J. Böning and A.J. Fallgatter (Psychiatric
Facial Expressions. - M.J. Herrmann*, H. Ellgring+ and
Neurophysiology, Department of Psychiatry and Psy-
A.J. Fallgatter* (*Psychiatric Neurophysiology, Depart-
chotherapy, University Hospital of Wuerzburg, Ger-
ment of Psychiatry and Psychotherapy, University
Hospital of Wuerzburg, Germany , +Institute of Psy-
A cue-reactivity has been described for different
chology, University of Wuerzburg).
physiological variables including electrophysiological
The processing of facial expressions was investi-
measures, like event-related potentials (Herrmann et al.,
gated using event-related potentials (ERP). Sad, happy
in press) in alcohol-dependent patients. However, it can
and neutral facial expressions were presented for 500 ms
be observed that some healthy subjects display such an
to 16 healthy, right-handed subjects. The ERPs elicited
electrophysiological cue-reactivity as well. The aim of
with the facial expressions were recorded from 21 leads,
this study, therefore, was to assess the effect of drinking
and both the amplitudes and the topography were calcu-
behaviours of healthy subjects on the cue-reactivity as
lated in time segments determined based on the Global
measured with event-related potentials (ERPs).
Field Power. Neutral facial expressions elicited higher
In 15 high and 15 low social male drinkers ERPs
amplitudes as compared to happy and sad faces in an
were registered with 21 scalp electrodes while 8 alco-
early time segment lasting from 148.4 - 265.4 ms after
hol-related and 8 neutral pictures were presented. Stim-
stimulus presentation. A significant topographical effect
uli were presented for 500 ms each with an interstimulus
was found in a later time segment from 269.5-371.1 ms.
The centre of gravity of the positive brain electrical fields
A stronger cue-reactivity consisting of significantly
was located over right-hemispheric brain regions in all
higher amplitudes in the ERPs elicited with alcohol-related
three conditions, with the most right-hemispheric posi-
as compared to unrelated pictures was found in high social
tion for sad faces. These results indicate that different
drinkers. This result indicates that the electrophysiological
neuronal assemblies are involved in the processing of fa-
cue reactivity previously found in alcohol-dependent
cial expressions, depending on their emotional content.
drinkers also occurs in social drinkers and that this cue re-activity is associated with the alcohol drinking customs.
The Face-Specific Event-Related Potential (P160) is Not
Affected by the Facial Expression. - M.J. Herrmann*, H.
Electrophysiological Cue-Reactivity in Alco-
E l l g r i n g + a n d A . J . F a l l g a t t e r * ( * P s y c h i a t r i c
hol-Dependent Patients.- M.J. Herrmann, H.-G.
Neurophysiology, Department of Psychiatry and Psy-
Weijers, G.A. Wiesbeck, J. Böning, A.J. Fallgatter(Psy-
chotherapy, University Hospital of Wuerzburg, Ger-
chiatric Neurophysiology, Department of Psychiatry
many, +Institute of Psychology, University of
a n d P s y c h o t h e r a p y , U n i v e r s i t y H o s p i t a l o f
The face-specific event-related potential (ERP) oc-
In alcohol-dependent patients a cue-reactivity can be
curring 160 ms after stimulus presentation has been de-
observed for subjective and physiological measures.
scribed by different research groups. However, most
Conditioning models of alcohol addiction suggest that
stimuli associated with previous drug use (cues) may ini-
phy (LORETA) from spontaneous EEG, separately for
tiate reward craving and relapse in a definite subgroup of
patients. Event-related potentials (ERPs) might be useful
The study was carried out on 12 patients with de-
to reveal the brain functional substrates of cue-reactivity.
mentia. Nicotine was administered transdermally for 4
In 19 male alcohol-dependent patients (44.2 ± 8.5 years)
weeks. The patients were examined before and 4 weeks
and 19 male healthy controls (42.5 ± 12.5 years)
after administration of nicotine. Eighteen electrodes EEG
event-related-potentials were registered with 21 scalp
during resting was recorded before and after administra-
electrodes while 15 alcohol-related and 15 neutral words
tion of nicotine. Three-dimensional images of cortical
were presented. Significantly higher amplitudes in the
electric neuronal activity, corresponding to the genera-
ERPs after alcohol-related compared to unrelated words
tors of different EEG frequency bands were computed by
(cue-reactivity) was found in alcohol-dependent patients,
but not in controls, at the electrode location Pz (F(1,36) =
Significant effects were found, depending on brain
5.2, p < 0.05). The results support the concept that
regions and EEG frequency band. The sources of EEG
cue-reactivity in alcoholism is based on a neurobiological
delta and theta activity attenuated at temporo-occipital
substrate Future investigations will show whether this
area after administration of nicotine. On the other hand,
cue-reactivity can be applied for the evaluation of a
the alpha 1 and 2 sources shift to more posterior locations
cue-exposure therapy and the assessment of the risk of a
resulted in nicotine administration. A significant de-
relapse in individual alcohol-dependent patients.
crease in limbic lobe occurred in beta1-3 range.
A Visual-Semantic System Commonly Accessed by
Global Field Synchronization (GFS) in Cognitive Im-
Animals and Tools Categories as Revealed by
pairment.- D. Hubl*, T. Koenig*, V. Jelic+, L.O.
Event-Related Potentials.- J.A. Hinojosa*, M.
Wahlund+ and T. Dierks* (*University Hospital of
Martín-Loeches*+, F. Muñoz*, P. Casado*, C.
Clinical Psychiatry, Bern, Switzerland, +Alzheimer Re-
Fernández-Frías+ and M.A. Pozo* (*Brain Mapping
search Center, Department of Clinical Neuroscience,
Unit, Universidad Complutense, Madrid, Spain,
Karolinska Institute, Huddinge, Schweden).
+Dpto. de Psicobiología, Universidad Complutense,
Cortical functioning is necessarily associated with a
well function neuronal network system. In the
Two different approaches aim to explain the organiza-
neurodegenerative course of Alzheimers dementia the net-
tion of category-related knowledge in the semantic system.
work is deteriorated and the coherent work of neurons dis-
One approach assumes a categorical organization in which
rupted. Glucose metabolism (GluM) measured with PET is
different categories are represented in segregated brain re-
used as a diagnostic tool for Alzheimer dementia, and cor-
gions. Another perspective postulates a non-categorical or-
relates with the severity of the disease. This might make it
ganization in which several categories place different
difficult to differentiate in individual cases between pa-
demands on shared brain subsystems. This study aims to
tients and healthy persons. Similar is true for commonly
investigate categorization processes by means of the Recog-
used EEG parameters like Alpha- or Theta-Amplitude.
nition Potential (RP), an electrical brain response peaking at
Global field synchronization (GFS) is presumed to measure
around 250 ms that reflects semantic processing and is orig-
the global coherence of electrical brain activity. Here we in-
inated in areas subserving visual-semantic processing. Re-
vestigated the influence of the disease severity on GFS and
sults show that animals and tools categories similarly
the relation between GFS and cerebral glucose metabolism.
access those areas generating the RP even though they were
PET and GFS were accessed in 48 patients with Alzheimer
supposed to differ in their processing demands. These find-
dementia and 17 healthy controls. We could demonstrate
ings are taken as supporting the non-categorical view, al-
that the synchronisation in alpha band is significantly re-
though introducing subtle variations since our data
duced and allows thereby a differentiation between con-
provide evidence of the existence of a brain subsystem spe-
trols and AD patients similar to GluM results. However, in
cialized in the processing of perceptual-semantic attributes
contrast to GluM GFS did not correlate with severity of cog-
that is equally accessed regardless of the semantic category.
nitive impairment. This may indicate, that global EEG syn-chronisation may become a more specific tool for
A LORETA Evaluation of Nicotine’s Effects on EEG in
Alzheimer diagnostics compared to glucose PET.
Dementia.- K. Hirata, H. Tanaka, M Arai, M. Harada, T.
Suga, J. Kubo, T. Nagashima and K. Yamazaki (Depart-
Source and Conductivity Estimation in a Simultaneous
ment of Neurology, Dokkyo University School of Med-
MEG/EEG Analysis. - H.M. Huizenga, T.L. van Zuijen,
icine, Tochigi, Japan).
D.J. Heslenfeld and P.C.M. Molenaar (University of
We evaluate the efficacy of nicotine as a cognitive
Amsterdam, Amsterdam The Netherlands).
enhancer with low-resolution electromagnetic tomogra-
A method is described to jointly estimate source and
conductivity parameters and the covariance matrix of
cess. Second, the sources are estimated given this noise
these parameters in a simultaneous MEG and EEG analy-
model, which can be done very efficiently due to the
sis. The method allows for unequal noise variances and
nonzero noise covariances between sensors. This method
Simulation results show that the source estimates
has several advantages. First, it profits from the preferred
are far more precise than those obtained from a standard
directions of MEG and EEG, respectively transverse and
analysis neglecting the noise covariance.
longitudinal. Second, it yields more precise estimates of
An application to real EEG data shows that the
deep sources than MEG alone. Third, the location of su-
Kronecker description is valid, that the spatial covariance
perficial sources profits from the tight MEG field,
can be described by a simple distance dependent model,
whereas the moment of superficial sources profits from
and that the temporal covariance can be described by an
EEG. Fourth, in contrast to MEG, radial moment can be
autoregressive moving average model of limited order.
estimated, which also increases the precision of the otherparameters. And fifth, the conductivity parameters can
Long-Term Memory Traces Facilitate Short-Term
be estimated. It is shown that conductivity estimation
Memory Trace Formation. - M. Huotilainen, A. Kujala
improves precision greatly compared to using incorrect
and R. Näätänen (Cognitive Brain Research Unit, De-
conductivity values. This is even the case if the data are
partment of Psychology, University of Helsinki, Fin-
generated in a realistic head, and analyzed in concentric
spheres. In that case conductivity estimation decreases
Long-term memory traces of phoneme categories
the location error from 1.74 to .32 cm, and the moment er-
have an effect on short-term memory (Näätänen et al.
1997), as shown by the enhancement of the mismatchnegativity (MMN), an auditory event-related response to
Optimal Sensor Positioning for Electromagnetic
a change in an ongoing sound stream, when the repeated
Source Analysis. - H.M. Huizenga, D. J. Heslenfeld and
"standard" stimulus was replaced with a na-
P. C.M. Molenaar (University of Amsterdam, Amster-
tive-language prototype phoneme compared to a
dam The Netherlands).
non-prototype phoneme. The present study aimed at de-
An analytic method is presented to compute optimal
termining whether this facilitation affects the number of
EEG/MEG sensor positions. The method consists of the
stimulus repetitions needed to form an adequate mem-
following stages. First, a source hypothesis is specified,
ory trace used in change detection. We recorded MMNs
this hypothesis can be derived for example from a pilot.
to infrequent stimuli occurring in trains of prototype and
Second, the covariance matrix C of the hypothesized
non-prototype phonemes, matched in their physical dis-
source parameters is calculated. Third, the overall pa-
tances to the standards, and in trains of sinusoidal tones.
rameter variance, as indexed by the determinant of C, is
We found that the shortest trains of standards were
minimized as a function of sensor positions. This yields
needed to produce an MMN when native-language pro-
totype phonemes were used as stimuli, suggesting that
Simulation results and an application to a real EEG
the automatic change-detection system can use informa-
dataset indicate that the proposed method improves pre-
tion from the long-term memory traces.
cision to a high extent. Source parameter standard errorshalved, and it became possible to separate two simulta-
Source Localization of Brain Electric Activity During
neously active sources, which was not possible on a stan-
Hypnotically Induced Anxiety and Relaxation. - T.
I s o t a n i * , H . T a n a k a + , D . L e h m a n n ^ , R . D .
Pascual-Marqui^, K. Kochi^, N. Saito*, T. Yagyu*, T.
Maximum Likelihood Spatiotemporal Source Analy-
K i n o s h i t a * a n d K . S a s a d a ~ ( * D e p a r t m e n t o f
sis. - H.M. Huizenga, L.W. Waldorp and R.P.P.
Neuropsychiatry, Kansai Medical University, Osaka,
Grasman (University of Amsterdam, Amsterdam The
Japan, +Department of Neurology, Dokkyo University
School of Medicine, Tochigi, Japan, ^The KEY Insti-
tute for Brain-Mind Research, University Hospital of
spatiotemporal noise covariance in a source analysis. The
PsychiatrY, Zurich, Switzerland, ~Nara Self-Control
essential feature is that the estimation problem is split
Clinic and Institute of Psychology, Nara, Japan).
into two parts. First, a model is fitted to the observed
The engagement of different brain regions which im-
noise covariance matrix. This model is a Kronecker prod-
plement subjectively experienced emotional states is not
uct of a spatial and a temporal matrix. The spatial matrix
clarified. We studied brain electric activity during hypnot-
describes the spatial covariances by distance dependent
ically induced anxiety and relaxation in eleven
functions. The temporal matrix describes the temporal
right-handed normals (26.5±7.6 years). After light hypno-
covariances by an autoregressive moving average pro-
sis, anxiety and then relaxation was suggested with a stan-
dardized text. 19-channel, eyes-closed EEG (20
Subjects had to discriminate gratings of different
seconds/subject) was analyzed using FFT approximation
spatial frequency and orientation flashed sequentially
and Low Resolution Electromagnetic Tomography
and randomized in the left or right visual field. Task rele-
(LORETA). Global tests revealed the strongest difference
vant stimuli had an occurrence probability of 20%. The
between EEG source gravity center locations during the
EEG of 16 healthy adults with normal vision was re-
two emotional states for excitatory beta2 frequency
corded from 30 channels regularly placed between the
(18.5-21 Hz). Post-hoc tests showed that the sources were
located more right during anxiety than relaxation.
The main effects observed with relevant stimuli was
LORETA specified that anxiety showed maximally stron-
activation in the delta band and a reduction of the alpha
ger activity than relaxation in right Brodmann area 10, and
band over the contralateral occipital regions. For beta ac-
relaxation than anxiety in left 22. Clearly, the two induced
tivity (between 13 and 22 Hz) there was a different topo-
emotional states were associated with different neural
graphical pattern: it was significantly reduced over
populations activity. Our results agree with reports on
central and parietal areas for cortical processing of task
brain activity shifted to the right (fronto-temporal) during
negative compared with positive emotions.
Supported by Deutsche Forschungsgemeinschaft,
Different Psychophysical Responses Correlated with
Topographical Differences During Binocular Rivalry
Mental Rotation of Alphabetic Characters and Human
of Faces. - A. Jedynak, E. Eger, T. Iwaki and W.
Hands: A High-Density Event-Related Potential Study.
Skrandies (Institute of Physiology, Justus-Liebig Uni-
- B.W. Johnson, Z.C. Thayer, J.P. Hamm, I.J. Kirk and
versity, Giessen, Germany).
M.C. Corballis (Department of Psychology, University
Does different perception of physically identical
of Auckland, Auckland, New Zealand).
stimuli influence evoked potential topography in binoc-
This study investigated brain processes concerned
with the mental rotation of two classes of shapes: alpha-
We presented schematic faces of brief duration syn-
betic characters and human hands. 128-channel ERPs
chronously in the left and right hemifield in a binocular
were measured from 2 groups of normal, right-handed
rivalry condition. In a forced-choice reaction time task, 31
human subjects. In both experiments, visual stimuli were
healthy subjects decided which of two superimposed
presented at varying angles from the upright orientation.
stimuli looked more „face-like". A regular array of 30
One group was required to decide if letters were normal
channels recorded the EEG between the inon and 5% an-
or mirror- reversed, while the other group was required
terior of Fz. Analysis of the subjects’ responses revealed a
to decide if hands were right or left hands. Late (latency
"functional eye dominance" establishing three subject
300-800 ms) parietal negative waves were elicited in both
groups (left, right, no eye dominance). Evoked potential
tasks, and the amplitudes of these waves were
data showed major components with mean latencies of
monotonic functions of the angle of required rotation.
Other components in this time range had task-specific
Subjects’ responses to shortly presented complex vi-
spatial distributions. These data support the interpreta-
sual stimuli enabled us to define an eye dominance dif-
tion that the two rotation tasks engage some common
ferent from conventional eye preference. Averaging
brain mechanisms, particularly parietal cortex; and also
according to stimulus as well as responses of the subject
cortical mechanisms that are idiosyncratic to the mental
allowed a detailed analysis of the evoked brain activity
transformation of each class of shapes.
Supported by Royal Society of New Zealand Mar-
Supported by Deutsche Forschungsgemeinschaft,
Using Independent Component Analysis for Prepro-
Changes in Spectral Topography of Spontaneous EEG
cessing EEG/MEG Data. - J. Kastner, M. Fuchs and M.
During Lateralized Visual Attention Tasks. - A.
Wagner (Neuroscan Labs, Lutterothstr. 28e, 20255
Jedynak, T. Hayakawa and W. Skrandies (Institute of
Physiology, Justus-Liebig University, Giessen, Ger-
For source reconstruction, an appropriate source
model has to be chosen before solving the inverse prob-
We analyzed topographical and global changes of
lem. One common tool is the Singular Value Decomposi-
spectral amplitude distributions induced by cortical pro-
tion (SVD) of the measured data matrix, yielding not only
cessing of visual information in healthy subjects per-
the number of singular values above noise level but also
forming a pattern discrimination task with lateralized
the corresponding spatial patterns and time loadings. The
main drawback of this approach is that the patterns and
loadings are orthogonal to each other, and therefore not
The high-frequency bursts occur with a significantly
representing the true nature of the underlying activity.
longer latency in schizophrenic patients. The source
Thus, an Independent Component Analysis (ICA) is
power showed a tendency towards a higher amplitude in
performed to reveal field patterns that are mutually inde-
schizophrenics, while frequency remained unaffected.
pendent and not essentially orthogonal. The components
The results will be discussed in the context of
found by the algorithm are extracted from the mixing
matrix and sorted by their relevance to the data.
The performance of the ICA is compared to SVD, us-
Temporal Structure of Frequency-Tagged Coherent
ing simulated data with an analytical head model. Calcu-
Multimodal Evoked Neuromagnetic Fields. - T.
lations are performed for strongly and weakly
Kobayashi*, D. Cheyne+^ and T.Cheung^ (*Hokkaido
overlapping time loadings with gaussian noise added.
University, Sapporo, Japan, +Simon Fraser University,
Burnaby, Canada, ^CTF Systems Inc., Port Coquitlam,
Differences in Source Localiation of Electrical and
Pressure Stimulated 96-Channel SEP. - W. Kawohl,
In an attempt to elucidate neocortical dynamics of
T.D. Waberski, R. Gobbelé and H. Buchner (Depart-
conscious perception, we measured and analyzed
ment of Neurology, RWTH Aachen, FRG).
neuromagnetic fields evoked by visual, auditory and
We investigated the determination of the hand area
somatosensory stimuli presented simultaneously. Each
by dipole source estimation following artificial and natu-
sensory stimulus was presented and tagged at a particu-
ral stimuli. In principle there are advantages of both
lar frequency. Four healthy male subjects participated in
methods, pressure stimulation is less invasive, while
the experiment. The neuromagnetic fields were recorded
electrical stimulation can be applied with higher stimu-
for 60 s in each trial using a 68-channel whole-cortex
lus rates. Electrical and pressure-stimulation was per-
formed simultaneously on the thumb and fifth finger of
Coherent neuromagnetic fields over all sensor loca-
six healthy volunteers. The SEPs following electrical
tions were extracted as eigenvectors in single trial data
stimulation show a sharper peak and a higher amplitude
using principal component analysis after applying
compared to the pressure-stimulated potentials. For both
band-pass filters tuned for each tagged frequency. We
stimulus qualities significant differences between the
have devised a new method to retrieve temporal infor-
source localization of the two fingers can be reported
mation of the extracted eigenvectors by the projection of
whereas the sources of the same finger following differ-
original raw data to each eigenvector. It was found that
ent stimuli show no significant difference concerning
the time-frequency/scale structure of the retrieved tem-
poral data could be clearly revealed by short time Fourier
There is a tendency that, despite smaller source
transform and continuous wavelet transform.
strenght, the spatial solution of the hand area is higheraccording to pressure-stimulation. This may be due to
Global Field Synchronization (GFS), A New Way to
stimulation of more different receptor types by electrical
F u n c t i o n a l
C o n n e c t i v i t y
stimuli represented in a larger cortical extension.
q u e n c y - D o m a i n
F i n d i n g s
Neuroleptic-Naïve Schizophrenics. - T. Koenig*+, D.
High Frequency Components of SEP in Schizophrenic
Lehmann+, N. Saito^, T. Kuginuki^, T. Kinoshita^ and
Patients. - W. Kawohl*, T.D. Waberski*, C. Norra+, R.
M. Koukkou* (*University Hospital of Clinical Psychi-
Gobbelé* and H. Buchner* (*Department of Neurol-
atry, Bern, Switzerland, +The KEY Institute for
ogy, RWTH Aachen, FRG, +Department of Psychiatry,
Brain-Mind Research, PUK, Zurich, Switzerland, ^De-
RWTH Aachen, FRG).
partment of Neuropsychiatry, Kansai Medical Univer-
Median nerve SEPs show an oscillatory burst of about
sity, Osaka, Japan).
600 Hz superimposed on the N20 component (Gobbelé et
In order to study global functional connectivity in
al. 1998). This high-frequency component was supposed to
frequency domain EEG, the measure of Global Field Syn-
be generated in thalamo-cortical projection fibers (Gobbelé
chronization (GFS) was developed. After the FFT, the
et al. 1999). Alteration of the thalamic function which are
method computes the principal components of the com-
hypothesized in schizophrenia might be detectable by
plex values over channels at a given frequency. The ratio
analysis of the high-frequency part of SEP. We report dif-
of the two eigenvalues indicates how much brain activity
ferences between high frequency SEP of 12 schizophrenic
was driven by a single phase. Higher ratios indicate more
patients and age and gender matched controls. Median
synchronization, thus higher functional connectivity,
nerve SEP were recorded using 32 channels and evaluated
lower values indicate decreased connectivity. The mea-
by dipole source analysis and wavelet transformation be-
sure is independent of the reference, contains no implicit
fore and after application of a 450 Hz high-pass filter.
source models and avoids multiple testing.
In a first study of resting EEG of two independent
data is composed of orthogonal elements, it does not as-
groups of acute, first-break, neuroloeptic-naïve schizo-
sume stationarity, topographical maps are obtained and
phrenics compared to healthy controls, GFS values were
it allows to include user defined, specific EEG elements
found to be consistently reduced for theta activity
like spike-wave patterns. The method will be introduced
(6.5-8.0Hz). In the context of literature relating theta ac-
formally and examples are given including artificial data
tivity to memory functions, the result is interpreted as a
and multichannel EEG during physiological and patho-
functional disconnection of widespread neurocognitive
processes subserving working memory functions inschizophrenia.
Do Wine Aromas Have Relaxation Effects? - Y. Koga, K.
Nakagawa, Y. Hirayasu, H. Nagai, Y. Nakamura, H.
Developmental Norms of Brain Electric Microstates. -
Tanahashi (Department of Neuropsychiatry, Kyorin
T. Koenig*+^, L. Prichep+, P. Valdes Sosa^, E. Braeker^,
University School of Medicine, Tokyo, Japan).
D. Lehmann~, R. Isenhart+ and E.R. John+ (*University
Effects of aromas on brain activity were evaluated by
Hospital of Clinical Psychiatry, Bern, Switzerland,
using electroencephalography (EEG) in normal subjects.
+Brain Research Laboratories, NYU School of Medi-
EEG was measured for 3.5 minutes after exposure of each
cine, New York, ^Cuban Neuroscience Center Ave. 25
aroma sample; 6 wines, ethanol solution, and distillated
esq. 158, Playa, Havana, Cuba, ~The KEY Institute for
water (DW), with a 5 munites interval. Frequency power
Brain-Mind Research, PUK, Zurich, Switzerland).
spectrum was analyzed on selected EEG waves for 30 sec-
In EEG, time epochs with stable field topography can
onds from each session. Relative power spectrum was
be observed. These so-called brain electric microstates pre-
calculated for 6 frequency bands from each sample rela-
sumably correspond to a sustained coherent activity of ex-
tive to those from DW at 12 electrode sites. Repeated mea-
tensive functional networks. When EEG data is parsed into
surement ANOVA (Sample and Electrodes as within
distinctive microstates, one can classify and quantify differ-
factors) was applied with the relative power on each fre-
ent brain functional states mediated by such networks; the
quency band. Significant main effect of sample was found
time resolution is compatible with rapid information pro-
in all frequency bands, except 11-12 Hz. Follow-up
cessing. We established the age norms for basic microstate
ANOVA suggested that odor from Chardonnay and
parameters such as mean duration and frequency of occur-
Concord significantly increased power for 9-10 and 10-11
rence using an EEG database of 496 subjects between 6 and
Hz bands compared to ethanol solution. In addition etha-
80 years. We found that brain electric microstates evolve in
nol solution did not change the power spectrum com-
a discontinuous and non-linear way. Several putative brain
developmental stages were observed, separated bychanges in microstate parameters. These stage changes oc-
Source-Current Density in the Brain Underlying Ver-
curred at 12, 16 and 21 years of age. The result is discussed
bal and Spatial Cognitive Activity. - Z.J. Koles and J.C.
in the framework of developmental psychology and state
Lind (Department of Biomedical Engineering, Univer-
sity of Alberta, The Clinical Diagnostics and Research
Center, Alberta Hospital, Edmonton, Canada).
Topographic Time-Frequency Decomposition of the
The source-current density in the brain underlying 43
EEG. - T. Koenig*+, F. Marti-Lopez* and P.Valdes-Sosa*
channel EEGs recorded from 33 normal female subjects
(*Cuban Neuroscience Center, POB 6880, La Habana,
while at rest and while engaged in the execution of verbal
Cuba, +University Hospital of Clinical Psychiatry,
and spatial cognitive tasks was examined. Source-current
densities were computed using the LORETA algorithm at
Topographic time-frequency decomposition is a
2394 locations in the gray matter of the brain and com-
new computerized EEG analysis combining available
pared in the theta and alpha frequency bands using multi-
techniques from time-domain spatial EEG analysis and
ple pair-wise t-tests. The Type I, or false-positive error rate,
time-frequency decomposition of time series. A new,
was controlled using a randomised hypothesis testing pro-
physiologically and statistically plausible topographic
cedure. Significant differences between the tasks and the
time-frequency representation of multi-channel EEG is
resting condition based on the full cross-spectral matrices
obtained. The EEG is accounted by coefficients of a large
obtained from each subject in each frequency band were
set of user-defined EEG like time-series localized in time
nonspecific. However, a factorization based on the simul-
and frequency; which are optimized for maximal spatial
taneous diagonalization of the two cross-spectral matrices
smoothness and minimal norm. The coefficients are then
produced current density patterns corresponding to the
cluster-analyzed, yielding a small number of field con-
two active tasks that were highly lateralized. In the theta
figurations which vary in intensity as a function of time
band, the patterns of current density were much more lo-
and frequency. The method does not assume that the
Harmonic Structure and Topography of EEG Driving
and mutual information theory created in late eighties
Reaction to Intermittent Photic Stimulation in
and developed in nineties. It has already succeeded in
Children and Adolescents. - V.V. Lazarev*, D.M.
separating eye movement artifacts from human scalp
Simpson+, B.M. Schubsky* and L.C. de Azevedo*
EEG recording. Several characteristic sleep waves can be
(*Inst. Fernandes Figueira, FIOCRUZ, Rio de Janeiro,
recorded during sleep EEG recording. They are used as
Brazil, +Biomedical Engineering Program, Federal Uni-
stage determining factors of sleep staging. We appliedICA method to the sleep EEG for the sleep waves separa-
versity of Rio de Janeiro).
tion. 18 channels scalp digital EEG data were converted
In 15 normal subjects, the correlation (Pearson’s R)
into 18 by n matrix which was used as a original data ma-
between the frequency profiles (EEG amplitude spectra)
trix X. Independent source matrix U was obtained by in-
of the driving reaction to photic stimulation at 3-24 Hz,
dependent component analysis method (U=W*X, where
and harmonics, were calculated in 15 leads. Driving was
W is an 18 by 18 matrix obtained by ICA). ICA was ap-
found in 70-100% of cases in the occipital and in 30-80% in
plied to the original EEG containing sleep waves. Among
central areas, with strongest responses in the alpha, and a
the 18 independent components, components containing
second maximum in the theta band. Profiles for the second
characteristic shape of sleep waves could be identified.
and third harmonics were highly correlated with those of
Each independent component was reconstructed into
the fundamental, when considering the corresponding
original montage by the product of inverse matrix of W
EEG frequencies. Peak frequencies in profiles were close
and U. This result might demonstrate that characteristic
to the individuals’ background alpha rhythm. The profile
sleep waves could be separated from original EEG of un-
amplitudes and their Rs with occipital profiles, decreased
known mixed neural origins by the ICA method.
towards the frontal region, where the maxima in theta andalpha bands were similar. There was highinterhemispheric correlation between profiles. Patients
EEG Gamma Frequency Brain Sources Differ Between
with migraine had higher reactivity in theta and beta
Meditation-Induced, Altered States of Consciousness.
bands, reduced spatial consistency in peak frequencies,
- D. Lehmann*, P.L. Faber*, P. Achermann+, D.
and decreased interhemispheric correlation of profiles.
Jeanmonod*, L.R.R. Gianotti* and D. Pizzagalli^ (*The
KEY Institute for Brain-Mind Research, University
Hospital of Psychiatry, CH-8029 Zurich, +Institute of
Auditory Event-Related Potentials (AERP) in the Par-
Pharmacology, University of Zurich, CH-8057 Zurich,
ents of Schizophrenics. - I. Lebedeva and V. Orlova
Switzerland, ^Affective Neuroscience Lab., Dept. of
(National Mental Health Research Center, Moscow,
Psychology, University of Wisconsin, Madison, WI
The group of middle-age parents of schizophrenics
EEG gamma frequency implements consciousness;
was subdivided by psychiatrist into groups of mentally
meditation alters consciousness. We analyzed
healthy (11 subjects, 5 men) and those with different per-
27-channel EEG from Lama Ole Nydahl during four
sonality disorders (11,4). AERP in the standard oddball
meditations (two minutes each), repeated once: "Buddha
paradigm (target 2000 Hz, probability 0.2; non-target 1000
in front"; "100-syllables Mantra"; "Dissolution"; "Recon-
Hz) were compared between these groups and
stitution". LORETA tomography (Pascual-Marqui) com-
age-matched group of 11 healthy subjects (5 men) without
puted the intracortical three-dimensional distribution of
psychiatric disorders in first and second degree relatives.
the gamma (35-44Hz) generators (2394 voxels). Results:
Statistical analysis revealed significant intergroup differ-
all comparisons between meditations showed statistical
ences (p<0.02-0.05 in parietal zones) only for N100 in ERPs
significance (voxel-by-voxel T-statistics with nonlinear
to non-target stimuli with its decrease in the groups of par-
correction for multiple testing). All 3 maximal activity
ents. Unexpectedly, there were not significant differences
differences between meditations were localized right in-
for N200 and P300, although the latencies of P300 were
ferior posterior for "visualizing" meditation #1; left cen-
tended (p<0.06) to be prolonged in parents with personal-
tral for "language" meditation #2; and right anterior
ity disorders. The findings show the impairment in the in-
superior for "dissolution" meditation #3. For "reconstitu-
formation processing in the parents of schizophrenics
tion" meditation #4, localization was left central superior
similar to that found in patients with schizophrenia.
vs meditation #1 and right posterior inferior vs #2 and #3,conceivably because of linguistic as well as visual com-
Independent Component Analysis (ICA) Can Separate
ponents. Thus, different mentally self-induced altered
Sleep Waves. - I.K. Lee (Department of Neurology,
states of consciousness are associated with different lo-
Inha University Medical College, Inchon, Korea).
calizations of maximal brain electric gamma activity.
Independent Component Analysis (ICA) is a blind
Grant #670806 from Institut für Grenzgebiete der
source separation method using unsupervised learning
Coherence Changes in the Scalp EEG Corresponding to
EEG/MEG can help to understand the spatio-temporal
Verbal and Spatial Cognitive Activity. - J.C. Lind and
distribution of brain activity during cognitive tasks.
Z.J. Koles (Clinical Diagnostics and Research Center,
Alberta Hospital Edmonton and the Department of
Magnetoencephalographic Studies of the Auditory
Biomedical Engineering, University of Alberta, Ed-
Pathway. - B. Lütkenhöner (Institute of Experimental
Audiology, University of Münster).
The differences in coherence patterns in 43 channel
Animal experiments as well as invasive recordings
EEGs recorded from 33 normal female subjects while en-
in humans revealed a high complexity of the functional
gaged in the execution of verbal and spatial cognitive
organization of the auditory cortex. Compared to that,
tasks was examined. Coherencies between all of the sites
the common interpretation of the auditory evoked field
(903 combinations) were compared in the theta and al-
(AEF), based mainly on the peak structure of the re-
pha bands using multiple pair-wise t-tests. The Type I, or
corded data, appears to be overly simple. Detailed anal-
false-positive error rate, was controlled using a random-
yses of a variety of experiments revealed indeed a more
ized hypothesis testing procedure. Significant differ-
complex structure of the underlying sources. In general,
ences between the tasks based on the full cross-spectral
a high intraindividual consistency of the spatio-temporal
matrices obtained from each subject in each frequency
AEF pattern contrasted with a high interindividual vari-
band were few (less than 6). However, a factorization
ability, presumably reflecting the anatomical variability
based on the simultaneous diagonalization of the two
of the auditory cortex. Individual details are generally
cross-spectral matrices elicited numerous (more than
sacrificed when trying to identify features common to all
100) electrode pairs where the coherencies were signifi-
investigated subjects. While such a procedure has the
cantly different (p<0.05). The patterns of differences in
advantage of resulting in handy conclusions, it entails
the magnitudes of the coherencies were more lateralized
the risk that statistically significant results are obtained
in the alpha than in the theta band. Phase differences in
for meaningless parameters. For this reason it is more
the two tasks were as large as 30 degrees.
sensible to investigate single subjects with high precisionand to put the observed AEF down to the underlying cor-
C o m b i n i n g F u n c t i o n a l B r a i n I m a g i n g a n d
tical (and subcortical) structures on an individual basis.
Electrophysiology in the Study of Higher Visual Func-
Spatial information obtained by functional magnetic res-
tion. - D.E.J. Linden*, T. Dierks+, L. Muckli^, J. Waltz^,
onance imaging (fMRI) could help to achieve this chal-
D. Prvulovic~, M. Scherg# and R. Goebel% (*Dept. of
lenging goal, but only if it can be ensured that basically
Neurology, Johann Wolfgang Goethe-Universität,
the same events as observed in the AEF are studied.
Frankfurt, Germany, +University Hospital of Clinical
Psychiatry, Bern, Switzerland, ^Max Planck Institute
Ongoing EEG Activity During Explicit Learning of
for Brain Research, Frankfurt, Germany, ~Dept. of Psy-
Words and Faces. - M. Mölle*, L. Marshall*, H.L. Fehm+
chiatry, Johann Wolfgang Goethe-Universität, Frank-
and J. Born* (*Department of Neuroendocrinology,
G e r m a n y ,
N e u r o l o g y ,
+Department of Internal Medicine, Medical University
Ruprecht-Karls-Universität, Heidelberg, Germany,
of Lübeck, Lübeck, Germany).
%Dept. of Psychology, Maastricht University, Nether-
Activation in different oscillating neural systems
during implicit and explicit learning of words and faces
The investigation of the spatio-temporal dynamics of
was investigated by measuring the ongoing EEG activity.
cognitive functions in humans requires the combination
During implicit learning power in lower frequency bands
of non-invasive functional imaging techniques at high
(delta, theta) was higher and parietal alpha activity was
spatial and temporal resolution. We performed parallel
lower than during explicit learning. These differences
fMRI and MEG experiments with visual motion stimuli,
were attributed to increased visual scanning during im-
which provided insights into the generators of the early
plicit learning. Comparison of trials with efficient vs. poor
visual evoked potentials, while a joint fMRI and EEG
memory performance at later retrieval testing indicated
study of visual and auditory oddball paradigms revealed
that explicit encoding is associated with upper alpha
an overlapping system for the generation of the late cog-
desynchronization and theta synchronization. The com-
nitive potentials (P 300) in both domains. Visual
bined measure of theta synchronization and upper alpha
short-term memory was investigated with fMRI and EEG
desynchronization was selectively enhanced over left
using a delayed matching-to-sample paradigm. FMRI
fronto-temporal cortical regions during explicit learning
showed a prominent delay activity in the prefrontal cor-
of words and over right parietal regions during learning
tex, with a corresponding increase in high frequency ac-
of faces (p<0.001). This striking topographical dissociation
tivity at the respective frontal sites revealed by EEG. We
between materials for the combined measure of theta and
propose that the parallel application of fMRI and
upper alpha EEG activity indicates that the mode of ex-
plicit learning essentially relies on a joint activation within
the time course of the global field strength (GFP). Topo-
two neuronal circuitries: the thalamo-cortical circuitry
graphical analysis was computed with the descriptors of
presumed to subserve search processes within long-term
memory, and the hippocampo-cortical circuitry subserv-
Two major ERP-components have been detected:
ing the actual encoding into memory.
59-148 msec and 152-586 msec after stimulus presenta-tion. Latencies between faces and landmarks ERPs did
A c t i v a t i o n
L a n d m a r k s
not dffer significantly. Neither there were amplitude’s
LORETA-ERP-Study. - Th. J. Müller, A. Federspiel, T.
Dierks and W.K. Strik (University Hospital of Clinical
By contrast to studies in healthy controls, we did
Psychiatry, Bern, Switzerland).
not find face-specific activation and amplitudes seemed
fMRI-studies support the hypothesis of the existence
to be higher. Rather, there were no differences between
of specialized brain regions involved during face recogni-
ERPs of the complex visual stimuli. These results may
tion that are mostly found in the right fusiform area. The
give evidence that disturbance of orientation or mis-
aim of the present study was the investigation of the time
judging of faces is due to dysfunction of complex object
course of the brain electrical response to the presentation
of complex visual stimuli based on LORETA-solutions. 25subjects (39.4±9.4 yrs.) saw pictures of faces or buildings.
Functional Differences in the Processing of Concrete
Components of 21-channel-ERPs were assessed on the
and Abstract Words Differing in Their Degree of
basis of the time course of the global field strength. The lo-
Imageability. - M. Martín-Loeches*+, J.A. Hinojosa*, F.
calization of the electrical sources in a normalized
Muñoz*, P. Casado*, C. Fernández-Frías+ and M.A.
Talairach space was calculated. T-levels > 3.65 were set to
P o z o * ( * B r a i n M a p p i n g U n i t , U n i v e r s i d a d
be significant at a 5%-level. Between 51-138 msec more
Complutense, Madrid, Spain, +Dpto. de Psicobiología,
face-activation at the left fusiform gyrus (t=6.3) was ob-
Universidad Complutense, Madrid, Spain).
tained. Between 139-178 msec significant activations due
Whereas several investigators assume a unitary mo-
to faces were found at right fusiform gyrus (t=6.9) and
dality-independent semantic system, others consider that
posterior gyrus cinguli (t=7.2). Further hyperactivation
it is segregated into modality-specific regions. Previous
due to face-stimuli was found in the time window be-
research has investigated this question by using concrete
tween 179-295 msec with the right fusiform gyrus being
and abstract words. However, an important variable such
significant at a t-level of 6.3 and the left at t=5.6.
as the degree of imageability of the stimuli was not con-
LORETA-results give evidence that right fusiform gyrus
sidered in these experiments. In this study we attempted
face-activation, as found in several functional studies, is
to elucidate how concrete and abstract words highly dif-
attributed to a time window of 138-295 msec.
fering in imageability are processed as reflected by abrain electrical response, the Recognition Potential (RP),
Loss of Face-Specific Activation in Acute Schizo-
that indexes semantic processing. Our results show that
phrenic Epsiodes: an ERP-Study. - Th.J. Müller*, A.J.
those brain regions originating the RP became more acti-
Fallgatter+, G. Berger^, A. Federspiel*, A. Heidrich+
vated for concrete (more imageable) words, although ab-
P s y c h i a t r i c
stract materials activated these areas to a greater extent
Neurophysiology, University Hospital of Clinical Psy-
than other non-meaningful stimuli. This indicates that at
chiatry, Switzerland, +Dept. of Psychiatry, Clinical
least part of the semantic system is specialized in the pro-
Neurophysiology, University of Würzburg, Germany,
cessing of concrete-imageable words, although there is
^MH-SKY Park Ville, Australia).
not a full functional segregation since these areas are also
Detection and processing of visual stimuli (faces or
importantly accessed by abstract words.
landmarks) is fundamental in getting along with envi-ronmental and social interactions. These interactions are
Spatio-Temporal Correlation Between Brain Electrical
typically altered in patients with acute episodes of
Activity and Cerebral Blood Flow in Brain Activation. -
schizophrenia. The aim of the present study was to ex-
K. Nagata*, S. Matsuoka+ and K. Yaguchi^ (*Depart-
plore neurophysiological processing differences in com-
ment of Neurology, Research Institute for Brain and
plex visual objects in this certain sample.
Blood Vessels, Akita Japan, +Showa Hospital,
14 patients (33.9±10.9 yrs., 6 male) with acute psy-
Shimonoseki, Japan, ^Faculty of Education and Hu-
chotic disorder according DSM-IV (298.8, 12; 295.7, 2;
man Studies, Akita University, Akita, Japan).
PANSS: 63.7±14.5) saw pictures of faces or buildings.
Topographic mapping of EEG and ERP carries accu-
Event-related potentials (ERP) were recorded with 21
rate temporal resolution, although the spatial resolution
channels according the International-10-20-System.
is limited even with multichannel recordings. Con-
Components of the ERPs were assessed on the basis of
versely, positron emission tomography (PET) enables a
quantitative measure of hemodynamic and metabolic
current distribution, remaining consistent with
variables with a precise spatial resolution, but the tempo-
electrophysiology. From a mathematical viewpoint,
ral resolution is not better than 60-90 sec due to its theo-
LORETA yields a spatial low-pass filtered version of cor-
retical limitations. The combination of topographic EEG
tical current density. If there are no high frequency peaks
mapping and PET provides a useful neurophysiological
in the spatial spectrum of the current density, then
informations with good spatio-temporal resolution by
LORETA has low localization error. Quantitative neuro-
supplementing the shortcomings of each modality.
anatomical validation for LORETA is presented based on
Simultaneous recording of ERP and PET was carried
measurements with visual and auditory stimulation, and
out in 12 young male volunteers during the continuous
in a verbal fluency task experiment.
performance task (CPT) paradigms in which the subjectswere requested to discriminate pictures of famous per-
The Effect of Selected Prestimulus Activity on the Au-
sons’ faces ["face"], simple geometric figures ["figure"], or
ditory P300 Paradigm: Application to Schizoprenia. -
Japanese words consisting of two Hiragana characters
G.W. Price (Wolston Park Hospital, Queensland Cen-
["word"]. By the intravenous bolus injection of the oxy-
tre for Schizophrenia Research, Australia).
gen-15 labeled water, cerebral blood flow (CBF) was mea-
Interactive recording (iERP) is a method that can be
sured repeatedly and the results were analyzed by the
used to investigate the effect of an EEG pattern on the Event
three-dimensional statistical method. Based on
Related Potential (ERP). It involves real time processing of
32-channel EEG data, ERPs were averaged and the results
the background EEG, and the application of stimuli in re-
based on the global field power were analyzed statisti-
sponse to the recognition of a pre-determined pattern.
cally. When the CBF images for "face" were compared
An auditory oddball paradigm was implemented
with those for "figure" or "word", a significant focal in-
with Interactive recording in order to study the effect of
crease in CBF was seen in the parahippocampal gyri,
this form of recording in a population with treatment re-
fusiform gyri and lingual gyri in a right-side dominant
fractory schizophrenia. The pattern used had previously
fashion. On the ERP results, three major positive peaks
been found to be associated with an increase in ERP sig-
were detected at 112, 164 and 244 msec. The peak second
nal to noise ratio in single sweeps. The iERP was com-
major peak was seen in the parieto-occipital regions and
pared with the standard ERP from a matched recording
the subtraction of the peak amplitude showed a marked
without pre stimulus EEG testing. The results from 24
laterality between the "face" and "word". As previously
subjects showed that the P300 amplitude derived from
suggested in the literature, the ERP peak activity at 164
interactive recording, was significantly greater than that
msec may carry the functional process of facial recogni-
from a standard recording. The amplitude from both re-
tion and be closely related with the bilateral mesial tem-
cordings was significantly less in the schizophrenia
poral and basal occipital activation on PET images.
LORETA: Possibilities and Limitations. - R.D.
Selective Impairment of Parietal Cortex Activation in
Pascual-Marqui*, M. Esslen*, K. Kochi*, D. Hell+, D.
Alzheimer Patients Duirng Visuospatial Tasks - A
Lehmann*, K. Hirata^, H. Tanaka^, B. Saletu~ and P.
Functional Magnetic Resonance Imaging Study. - D.
Anderer~ (*The KEY Institute for Brain-Mind Re-
Prvulovic*, D. Hubl*~, A.T. Sack*, L. Frölich*, M.
search, University Hospital of Psychiatry, Zurich; Swit-
Jandl*, H. Lanfermann+, F.Zanella+, K. Maurer*, R.
zerland, +University Hospital of Psychiatry, Zurich;
Goebel#, T. Dierks*~, D.E.J. Linden*^ (*+^Departments
Switzerland, ^Department of Neurology, Dokkyo Uni-
of Psychiatry and Psychotherapy, Neuroradiology,
versity School of Medicine, Tochigi; Switzerland, ~De-
Neurology, University of Frankfurt, ~University Hos-
partment of Psychiatry, University of Vienna, Austria).
pital of Clinical Psychiatry, Bern, #Department of Psy-
Multichannel EEG measurements do not contain suf-
chology, University of Maastricht).
ficient information for determining uniquely the 3D elec-
Impairment of visuocognitive abilities is a common
tric neuronal activity distribution. Therefore, infinitely
phenomenon in Alzheimer’s disease (AD). A couple of
many solutions exist for this inverse problem. Many par-
neuropathological and functional imaging studies
ticular solutions have been published, each one character-
proved the crucial role of the posterior parietal cortex in
ized by some assumptions. The sole basis of "low
spatial processing and transformation of visual objects.
resolution electromagnetic tomography" (LORETA) is an
The aim of the following study was to determine whether
electrophysiological one: scalp EEG is due to the instanta-
the activation of parietal regions during visuospatial
neous occurrence of highly synchronized post-synaptic
tasks is differentially impaired by the neuropathological
potentials in clusters of cortical pyramidal neurons. It will
processes of AD and could therefore serve as a new
be shown that the mathematical implementation of "syn-
non-invasive diagnostic marker. 14 AD patients and 14
chronization" in LORETA over-blurs the estimated 3D
age matched controls were investigated with fMRI while
performing an angle discrimination task. The AD pa-
a decrease in alpha activity, a slowing of the delta/theta
tients showed a poorer performance on the task and re-
centroid as well as an acceleration of the alpha and beta
vealed a significantly lower activation in the right
centroid, reflecting a decrease in vigilance. Correlation
superior parietal lobule compared to controls. These re-
maps showed significant relationships between these
sults correspond to neuropathological findings in AD,
EEG measures and the Hamilton depression score
demonstrating a differential damage of cortical areas of
(HAMD): the worse the vigilance, the more pronounced
the visual system. This study shows that fMRI may be a
the depression. In the alpha-asymmetry index, de-
useful tool for the diagnosis and monitoring of AD.
pressed patients showed less power over the right thanleft frontal lobe, whereas normal controls exhibited the
New Insights into the Functional Relevance of Parietal
opposite. The frontal asymmetry index was significantly
Cortex Activation for Visuospatial Functions: Com-
correlated with HAMD, which suggests a right frontal
bined Evidence from fMRI and rTMS. - A.T. Sack*, D.
hyper- and left frontal hyperactivation in depression.
Hub*#, D. Prvulovic*, E. Formisano%, M. Jandl*, F.E.
Sedative antidepressants of the imipramine/
Zanella+, K. Maurer*, R. Goebel%, T. Dierks# and D.E.J.
amitriptyline type attenuated total power, decreased ab-
Linden*^~ (*Depts. of Psychiatry, Div. of Clinical
solute delta/theta and specifically alpha power, in-
Neurophysiology, Universität, Frankfurt, Germany,
creased relative delta/theta and decreased relative alpha
+Neuroradiology, Universität, Frankfurt, Germany,
and to some extent also relative beta power and slowed
^Neurology, Universität, Frankfurt, Germany,
the total centroid. This is partly in contrast to
~Max-Planck-Institut für Hirnforschung, Frankfurt,
non-sedative antidepressants, which attenuated total
Germany, #University Hospital of Clinical Psychiatry,
power and absolute delta/theta power, but increased ab-
Bern, Switzerland, %Department for Psychology,
solute and specifically relative beta power, slowed the
Universität Maastricht, The Nederlands).
delta/theta centroid and accelerated the alpha and beta
We combined functional magnetic resonance imaging
centroid, as well as the total centroid.
(fMRI) and repetitive transcranial magnetic stimulation
Finally, identifying cortical target regions of
(rTMS) to investigate the functional relevance of parietal
pyschotropic drugs by low-resolution brain electromag-
cortex activation during the performance of visuospatial
netic tomography (LORETA), we recently described that -
tasks. FMRI provides information about the correlation be-
at pharmacodynamic tmax - 20 mg citalopram induced
tween local transient changes in neuronal activation and
the most pronounced changes in the beta-3 and beta-2
specific cognitive functions. Information on the functional
bands, showing a bilateral increase, predominantly in the
relevance of this activation was obtained by using rTMS to
frontal (F) and temporal (T), but also parietal (P) and limbic
induce temporary regional deactivations as an independ-
(L) lobes and sublobularly (S), further an increase in beta-1
ent variable and investigate its effect on the performance of
and alpha-2 over the right F and T, but also L lobes and S as
specific cognitive functions as dependent variables within
well as a delta augmentation in the left F, T and P lobes.
a controlled experimental design. We investigated cogni-
The results of a predominantly right hemispheric in-
tive tasks that were performed on the same visual material
crease in alpha-2 and beta-1 activity after 20 mg citalopram
but differed in the demand on visuospatial functions. For
are in agreement with the post-mortem neurochemical
the visuospatial tasks we found a selective enhancement of
studies of Arato et al. (1991), implying a higher 5HT turn-
fMRI signal in the superior parietal lobule (SPL) and a se-
over in the right non-dominant "affective hemisphere"
lective impairment of performance after rTMS to this re-
than in the left "cognitive hemisphere", which had already
gion. We could thus show that the parietal cortex is
earlier been reported by Flor-Henry (1985).
functionally important for the execution of spatial judge-ments on visually presented material.
Functional Imaging Improves When EEG/MEG is
Combined with MRI/fMRI. - M. Scherg (Sektion
Neurophysiological Aspects of Depression and Anti-
Biomagnetismus, Dept. of Neurology, University Hos-
depressants: EEG Mapping and Tomography
pital of Heidelberg).
( L O R E T A ) . - B . S a l e t u * , P . A n d e r e r * , G . M .
EEG or MEG data rarely permit to localize the source
Saletu-Zyhlarz*, R.D. Pascual-Marqui+ (*Department
activities of all active brain regions. Resolution can be im-
of Psychiatry, University of Vienna, Austria, +The KEY
proved using anatomical and functional information from
Institute for Brain-Mind Research, Zürich, Switzer-
MRI and fMRI. For any active brain region, a spatial filter
can be established by a linear combination of the record-
Recent EEG mapping studies demonstrated that
ing channels. However, when the number of active areas
major depression was characterized by a decrease in ab-
increases, the spatial filter must be sharpened to separate
solute power in all frequency bands, a tendency towards
the activities of all areas. Functional-anatomical informa-
an augmentation of relative delta/theta and beta power,
tion helps to confine source space by seeding of dipole
sources into active brain regions or by priori weighting
representations of 3D objects were identical or mirror im-
over space (MRT/fMRT) and time (EEG/MEG) to focus
aged within a rotation. Brain electrical activity was re-
corded from 64 scalp sites while a diffuse 13Hz flicker
We recorded tactile somatosensory evoked fields
superimposed over the visual fields elicited the SSVEP.
(SEF) separately from tactile and motion tasks in fMRI.
During the period of mental rotation, increasing angular
Source activities of the contralateral primary and second-
disparity was associated with increased SSVEP amplitude
ary areas were separated by multiple dipole modeling but
at parietal, left temporal and prefrontal sites. Increased
not by cortical current mapping. fMRI-based weighting
angular disparity was also associated with increased
improved separation considerably. By combining MRI-
SSVEP phase-lag at prefrontal and left temporal sites.
and fMRI data of a patient with epileptic myoclonus of the
During mental rotation occipito-prefrontal and
hand, the multiple source model of EEG and MEG demon-
parieto-prefrontal SSVEP partial coherence increased
strated propagation from somatosensory cortex to motor,
with increasing angular disparity. Our data suggests that
premotor and contralateral motor cortex. Thus, by com-
the SSVEP may offer useful information on the functional
bining functional images with high spatial (MRT/fMRT)
relationship between various brain regions during the
and temporal (EEG/MEG) resolution we gain more in-
creation and manipulation of mental images.
sight where and when cortical processing occurs.
Effects of Temporal Gaps Between Successive Fixation
Various Clinical Applications of the Deviation Ration
Targets on Discrimination Performance and visually
Topography (DRT). - F. Shichijo*, Y. Kaji+, M.
Evoked Brain Activity. - W. Skrandies* and E.
Akutagawa^, O. Takimoto*, S. Nagahiro*, H.
Anagnostou+ (*Institute of Physiology, Justus-Liebig
Nakayama~, Q. Zhang~ and Y. Kinouchi~ (*Department
University, Giessen, +Department of Neurology, Lud-
o f N e u r o l o g i c a l S u r g e r y , T h e U n i v e r s i t y o f
wig Maximilians University, Munich).
+Faculty of Engineering, Tokushima
We studied how a specific oculomotor task influ-
Bunri University, ^School of Medical Sciences, the
ences afferent visual processing. Twenty subjects had to
University of Tokushima, ~Faculty of Engineering, the
perform visually guided saccades, and between the off-
University of Tokushima, Tokushima, Japan).
set of the fixation light and the onset of the new target
The deviation ratio topography (DRT) is our original
there was a either a 200 ms temporal gap or no gap. Dur-
software devised in 1986 to detect the dynamic EEG varia-
ing saccades visual stimuli with different orientations
tions. With DRT, not only the changes in frequency bands
were presented, and discrimination performance and
but also the spatial and sequential changes of EEG activi-
topographic VEP activity was analyzed.
ties are visually evaluated statistically. The idea of DRT
In the gap condition, more correct responses oc-
was devised under the EEG constancy that the individual
curred. However, global field power, evoked component
EEG topography patterns show almost the same patterns
latency and topography of the VEP were not affected by
in the same condition. The DRT mapping is made of the
the gap. This suggests that parieto-occipital areas need
logarithmic ratio between the control data (C) and mea-
not be the most important sites of sensorimotor integra-
sured data (M) that is recorded on the same electrode un-
tion. From perceptual data it is clear that intrasaccadic vi-
der the different conditions. The formula is as follows:
sual processing is influenced by presaccadic events.
logarithmic deviation ratio = log (M/C). The clinical appli-
Under the short time constraints prevalent in the
cations of the DRT were as follows: (1) the intraoperative
saccadic task, fixation target cues are not only used for
monitoring, (2) the EEG monitoring during the physiologi-
motor planning, but also determine the visibility of vi-
cal or mental tasks, and (3) the evaluation of EEG changes
sual patterns presented during the eye movement.
after medication or operations. The DRT system for Win-
Supported by Deutsche Forschungsgemeinschaft,
dows has been developed in our institutes.
Changes in Steady State Visually Evoked Potential
The Influence of Visual Field Location and Frequency
(SSVEP) Amplitude, Phase and Coherence During
on Visual Perception and Brain Activity Elicited by
Mental Rotation. - R.B. Silberstein, F.M. Danieli and
Motion-in-Depth Stimuli. - W. Skrandies, A. Jedynak
P.L. Nunez (Brain Sciences Institute, Swinburne Uni-
and Ch. Kirr (Institute of Physiology, Justus-Liebig
versity of Technology. Melbourne, Australia).
University, 35392 Giessen, Germany).
In this study, we used scalp SSVEP amplitude topog-
We investigated the effect of visual field location and
raphy, SSVEP phase topography, and SSVEP coherence to
motion frequency on psychophysical thresholds and brain
investigate the organization of brain activity during the
activity elicited by dynamic random-dot stereograms.
Shepard & Metzler (1971) mental rotation task.
Stereoscopic checkerboards were presented at ran-
Twenty-two right-handed males indicated whether two
dom centrally, or in the left or right visual field with hori-
zontal disparity changing between 2 and 16 Hz. Ten
gions EEGs of Alzheimer patients and controls with age
subjects had to indicate whether and where stimuli had
associated memory impairment were analyzed with lin-
moved in depth. Simultaneously, electrical activity was
ear and nonlinear measures of functional connectivity.
recorded from 30 electrodes over the occipital and parietal
EEGs were recorded in 10 subjects (age 73.6 (SD 6.7); 5
brain areas. Steady-state VEPs were quantified by FFT.
males; 5 females) with age associated memory impair-
Largest sensitivity occurred with central stimuli
ment (AAMI) and 28 patients (age 73.7 (SD 6.6); 14 males;
(mean 4.5 Hz) while laterally motion-in-depth could not
14 females) with Alzheimer’s disease all referred to an
be perceived at frequencies exceeding 2.5 Hz. There was
outpatient clinic for memory disorders. EEGs were ana-
no difference between the left and right visual field.
Curriculum Vitae Richard L. Taylor, M.D., F.A.A.N. DATE OF BIRTH: March 23, 1949 PLACE OF BIRTH: Atlantic City, New Jersey BUSINESS ADDRESS: Taylor Medical Group 22 West Road - Suite 101 Towson, Maryland 21204-2388 Telephone 410-823-3600 Facsimile 410-823-3605 Web Site: www.taylormedicalgroup.com PROFESSION: Neurology Electroencephalography (EEG) Electrodiagno
Aesthetic Dermatology S K I N & A L L E R G Y N E W S • J u l y 2 0 0 8 Photo at left shows a patient before treatment / with the OLDG ENTER photopneumatic device. Photo at ICHAEL right shows . M improvement of the patient’s acne B Y S H A R O N W O R C E S T E R after receiving TESTY L