[Comp-neuro] July 28th,
10:00 OCNS 2011 Workshop on non-invasive imaging of
cortical interactions
Alex Ossadtchi
ossadtchi at gmail.com
Fri Jul 1 17:58:43 CEST 2011
Dear colleagues,
I would like to attract your attention to the workshop that is going
to be held as a part of OCNS 2011 meeting.
Please, follow this link for more detailed information listing
speakers and abstracts of the talks
along with the list of relevant papers.
http://www.cnsorg.org/2011/ossadtchi.shtml
For you convenience I am placing some of this information below.
Thanks for your potential interest and sorry for possible cross-posting,
Alex
Non-invasive imaging of non-linear interactions
July 28th, 10:00, one day workshop within CNS*2011
Twentieth Annual Computational Neuroscience Meeting CNS*2011 - July
23-28, 2011 in Stockholm, Sweden
Organizer : Alexei Ossadtchi, St. Petersburg State University, St.
Petersburg, Russia
Titles of talks and abstracts
Network identification and characterization methods from human
electrophysiological data – an overview (Richard Greenblatt)
The characterization of large scale transient networks plays an
increasingly prominent role in the study of the dynamical neural
systems on the centimeter length scale and millisecond time scale. The
properties of these networks may be inferred from structural imaging,
functional imaging and electrophysiological measurements. In this
overview presentation, I will compare a number of linear, bilinear,
and nonlinear methods in both the time and time-frequency domains.
These include MVAR, coherence, phase synchrony, mutual information,
and transfer entropy, which have been used to infer network geometries
and topologies from EMEG and ECoG (local field potential) data.
Domains of applicability and the relative strengths and limitations of
various methods will be considered. Algorithmic descriptions of the
methods will be discussed in the common context of a random variable
estimation framework, and application examples from experimental data
will be presented.
Reconstructing phase dynamics of oscillator networks (Michael Rosenblum)
We consider the problem of reconstruction of phase dynamics of coupled
oscillators from data. The crucial issue of our approach is
distinction between phase estimates, obtained, e.g., via the Hilbert
transform, and hereafter denoted as protophases, and true phases, used
in the theory of coupled oscillators. We present a transformation from
protophase to phase, which allows us to reconstruct the phase dynamics
equations of coupled oscillators in an invariant way, i.e. to a large
extend independently of the observables, used for reconstruction. We
start by the case of two coupled systems and illustrate it with
numerical examples as well as with real data. We present examples,
demonstrating importance of the protophase to phase transformation for
a correct quantification of the strength and directionality of
coupling. We proceed by consideration of small networks of coupled
periodic units. Starting from the multivariate time series, we first
reconstruct genuine phases and then obtain the coupling functions in
terms of these phases. The partial norms of these coupling functions
quantify directed coupling between oscillators. We illustrate the
method by different network motifs for three coupled oscillators and
for random networks of five and nine units. We also discuss effects of
non-pairwise coupling.
Cross –frequency coupling in ECOG – Evidence of non-linear coupling of
distant cortical sites (Felix Darvas)
Large scale synchronization is widely believed to reflect coordination
of neuronal assemblies during complex tasks and as such is fundamental
in the study of human cognition. Presently, the majority of methods
have been focused on synchronization within a single frequency or
frequency band or on harmonic couplings of those frequencies. However,
cross-frequency coupling of cortical areas has been the subject of
recent studies, where a focus has been on phase-amplitude coupling of
cortical signals. It has been shown however that pure phase–phase
relationships, i.e. phase synchronizations, between distant cortical
sites are important and can occur without local amplitude changes.
We present a new method, the bi-phase locking value (bPLV), to study
such phase relationships across frequencies. We estimate time varying
quadratic phase coupling of electrophysiological signals. We use this
method to show in a simple cued motor paradigm, that robust task
related non-linear phase-coupling between the mu rhythm and high gamma
frequencies exist. Since little a priori knowledge is available on
such interactions, we implemented data driven methods to extract these
interactions from intra cranial recordings, i.e. ECoG, from epileptic
patients, which were implanted prior to surgery. ECoG data are ideal
for such hypothesis testing, as they exhibit a much better
signal-to-noise ratio than non-invasive data, e.g. EEG or MEG. Also
recordings from ECoG electrodes do not suffer from linear cross-talk
between putative cortical interaction sites, which affects
synchronization detection methods.
The existence of such non-linear interactions demonstrates that the
human cortex is using a much wider range of communication between
distant neuronal populations than can be revealed by single-band
interactions. Consequently we speculate that multiplicative
interaction and, more generally, phase–phase interactions may be a
fundamental mode of communication between distant cortical areas.
Cross-frequency phase and amplitude interactions among human cortical
oscillations. (Matias Palva)
Oscillatory interactions and synchronization may be a key mechanism
for coordinating scattered neuronal activity into transient neuronal
assemblies to serve coherent action and perception at the systems
level. How is the co-operation of oscillatory assemblies in distinct
frequency bands regulated? Two distinct forms of cross-frequency (CF)
phase interactions, phase-phase and phase-amplitude coupling, are in a
position to mechanistically underlie such cross-spectral coordination,
but their prevalence and phenomenology in cognitive tasks has remained
unclear.
Neuronal synchronization in the human brain can be investigated
non-invasively with millisecond-range temporal resolution by using
magneto- (MEG) and electroencephalography (EEG) but it is essential to
incorporate source reconstruction methods into M/EEG data analyses to
recover and disentangle information about the underlying cortical
sources. In this presentation, I discuss framework for integrated
mapping of within- and cross-frequency phase interactions in M/EEG
data. I first review a family of interaction metrics and discuss the
favourable and less favourable properties of three established
circular statistics; phase-locking value, pairwise-phase consistency,
and mutual information, and also present a novel method for estimating
causal directionality in phase interactions. I then demonstrate how
the predominant confounding factors inherent to M/EEG, e.g., signal
mixing, signal-to-noise ratio dynamics, and evoked activity, bias the
interaction statistics and how these biases could be compensated for.
Finally, I will describe the cross-frequency phase interaction
networks for the coupling among theta-, alpha-, and beta-frequency
band oscillations during visual working memory. In line with prior
observations, these data suggest that cross-frequency phase
interactions may indeed support cross-spectral functional integration.
DICS variations for non-invasive cross-frequency coupling detection.
The method and preliminary results. (Alex Ossadtchi)
Synchronization of activity between distinct cortical regions
underlies the mechanism of functional integration that forms a
foundation of all our actions. Recently, the role of non-linear
interactions manifested in cross-frequency(across scale)
synchronization has been emphasized as facilitating the exchange of
information between cell assemblies. Such synchronization has been
found in many experimental paradigms and is currently under active
study. Unfortunately spatially and temporally precise analysis
available only in a limited number of cases corresponding to
neurological patients with implanted cortical grids. In order to
provide the flexibility in experimental designs and allow for more
specific studies tools for analysis of such non-linear
synchronizations are to be developed. Instrumentally, MEG is a unique
technology that allows for mapping of cortical activations and
provides high temporal resolution. The use of beamformers supported
by sufficiently accurate forward models allows for reasonable (0.5 cm)
spatial resolution. The time frequency representation of MEG signals
is natural and captures the nature of MEG observed cortical activity
as consisting of short time narrow-band bursts.
In this work our goal was to combine the above and develop a signal
processing method for identification of the cortical spatial structure
of cross-frequency coupling between the oscillations in the two
non-overlapping time-frequency windows. Our method is a statistical
test contrasting the results of adaptive beamformer based inverse
mapping obtained using the original and cross-term deprived
time-frequency domain data covariance matrices by calculating the
ratio of the two inverse values. We use multiple comparison corrected
randomization statistical tests for identification of significant
source space coupling.
Application of the method to an event-related MEG dataset from a
single subject (imagined hand rotation) yielded plausible results with
interacting pairs falling into physiologicsally plausible cortical
sites. We observed beta-gamma coupling between frontal and
parietal-occipital regions, consistent with published signal space
analysis. We also observed beta-gamma prefrontal/frontal and
alpha-gamma temporal/frontal couplings.
Studying neuronal n:m phase synchronization with the frequency-shift
approach (Vadim Nikulin)
Neuronal synchronization has been hypothesized to be the mechanism
through which efficient communication between the neurons can be
achieved. In addition to conventional interactions at the same
frequency range, phase synchronization between different frequency
ranges has been demonstrated recently. Such cross-frequency phase
synchronization is usually studied in a sensor space or with
computationally demanding beamformer techniques, leading to a very
large number of statistical comparisons. Here we present a novel
method for the extraction of neuronal components showing
cross-frequency phase synchronization. The method allows a compact
representation of the sets of interacting components (along with their
spatial patterns) without the need to perform inverse modeling. In
general it works for the detection of phase interactions between
components with frequencies n and kn, where n and k are integers. This
class of interactions includes 1:2 and 1:3 synchronization frequently
observed in EEG and MEG recordings. We refer to the method as
Cross-Frequency Decomposition (CFD), which consists of the following
steps: a) extraction of n-oscillations with spatio-spectral
decomposition algorithm (SSD); b) frequency-shift transformation of
the oscillations obtained with SSD, and c) finding kn–oscillations
synchronous with n-oscillations using least-squares estimation. Our
simulations showed that CFD was capable of recovering interacting
components even when the signal-to-noise ratio was as low as 0.1. An
application of CFD to the real EEG data demonstrated that
cross-frequency phase synchronization between alpha and beta
oscillations can originate from the same or remote neuronal groups.
While interactions occurring at the same spatial location can
potentially indicate quasi-sinusoidal waveform of neuronal
oscillations, the synchronization between spatially remote populations
is likely to be a marker of genuine neurophysiological coupling
between different oscillations.
Oscillatory cortico-subthalamic connectivity in Parkinson's patients/
Brain Connectivity and Model Comparison (Vladimir Litvak/Will Penny)
Insights into how brain structures interact are crucial for
understanding the principles of distributed neuronal function and may
finesse diagnosis and therapy. To study the interactions between
cortex and basal ganglia we acquired simultaneously
magnetoencephalography (MEG) and direct recordings from the
subthalamic nucleus (STN) in 17 Parkinson's disease patients. The
patients performed self-paced button presses with three fingers either
simultaneously or sequentially. The experiment was repeated with and
without the dopamine prodrug levodopa.We examined the effect of
movement complexity and drug on induced power in the contralateral
primary motor cortex (M1) and STN. Changes in mean power appeared
similar in the two structures, but only M1 exhibited prolonged high
gamma (50-90 Hz) activity for sequential movement. Levodopa caused a
wide-band increase in power around movement onset in M1 and an
increase limited to the gamma band in the STN. We used a novel
analysis method - dynamic causal modelling for induced responses to
assess the coupling between M1 and STN and its modification by
dopamine replacement therapy and by the movement complexity.
Relevant papers
1. Palva JM, Monto S, Kulashekhar S, Palva S (2010) Neuronal synchrony
reveals working memory networks and predicts individual memory
capacity. Proc Natl Acad Sci U.S.A. 107: 7580-5.
2. Palva S, Monto J and Palva JM (2010). Topological properties of
synchronized cortical networks during working memory maintenance.
Neuroimage 49: 3257-68.
3. Dennis J.L.G. Schutter , Claudia Leitner , J. Leon Kenemans, Jack
van Honk, Electrophysiological correlates of cortico-subcortical
interaction: A cross-frequency spectral EEG analysis, Clinical
Neurophysiology 117 (2006) 381–387
4. F. Darvas, J.G. Ojemann, L.B. Sorensen, Bi-phase locking — a tool
for probing non-linear interaction in the human brain, NeuroImage 46
(2009) 123–132
5. Vadim V. Nikulin, Guido Nolte, Gabriel Curio, A novel method for
reliable and fast extraction of neuronal EEG/MEG oscillations on the
basis of spatio-spectral decomposition,NeuroImage 55 (2011) 1528–1535
6. B. Kralemann, A. Pikovsky, and M. Rosenblum,Reconstructing phase
dynamics of oscillator networks,Chaos, xx, p. xxxx, 2011, submitted.
7. A. Ossadtchi, R.E. Greenblatt, V.L. Towle, M.H. Kohrman, K. Kamada,
Inferring Spatiotemporal Network Patterns from Intracranial EEG Data,
Clin, Neurophysiology, June 2010 ; 121(6):823-35
Workshop format
The workshop consists of 7 presentations of 30 minutes each plus 10
minutes for questions/discussion. A final session of 45 minutes is
planned for the final discussion.
More information about the Comp-neuro
mailing list