Three papers on matching pursuit analysis of EEG and software
used in these studies, described below, are available from
http://brain.fuw.edu.pl/~durka
title: Stochastic Time-Frequency Dictionaries for Matching Pursuit
journal: IEEE Tran. SP, March 2001
url: http://brain.fuw.edu.pl/~durka/papers/stochmp
abstract:
Analyzing large amounts of sleep electroencephalogram (EEG) data by
means of the matching pursuit (MP) algorithm, we encountered a
statistical bias of the decomposition, resulting from the structure of
the applied dictionary. As a solution we propose stochastic
dictionaries, where the parameters of the dictionary's waveforms are
randomized before each decomposition. The MP algorithm was modified
for this purpose and tuned for maximum time-frequency resolution.
Examples of applications of the new method include parametrization of
EEG structures and time-frequency representation of signals with
changing frequency.
title: A unified parametrization of EEG
journal: submitted to IEEE
url: http://brain.fuw.edu.pl/~durka/papers/unification
abstract:
70 years since the first recording of human electroencephalogram
(EEG), visual analysis of raw EEG traces is still the major clinical
tool and point of reference for other methods, in spite of its
inherent limitations: low repeatability and high cost.
7 years since the introduction of the matching pursuit, we collected
evidence suggesting that adaptive time-frequency approximation is a
good candidate for a universal high-resolution parametrization of EEG,
compatible with the visual and spectral analysis, and applicable to a
large class of problems. In the following we briefly discuss the need
for a generally applicable method for a mathematical description
(parametrization) of the signal, which would be directly related to
the heritage of the traditional EEG analysis. The main section
discusses in this context application of the matching pursuit
algorithm. We present recent advances in analysis of sleep EEG and
discuss earlier works on event-related potentials and epileptic
recordings.
title: Time-frequency microstructure of event-related EEG
desynchronization and synchronization
journal: Med and Biol Eng & Computing, May 2001
url: http://brain.fuw.edu.pl/~durka/papers/microstr
abstract:
We present a new method for analysis of event-related EEG phenomena,
in particular ERD/ERS related to a voluntary movement, which offers:
-high time-frequency resolution, and hence also increased ERD/ERS
sensitivity (especially in the gamma band, where improvement may
exceed order of magnitude),
-possibility to analyze the whole picture of energy changes at once,
without setting a priori the analyzed frequency bands,
-parametric description of signal's structures.
The main idea is based upon averaging energy distributions of single
EEG trials in the time-frequency plane. As the estimator for signal's
energy density we choose matching pursuit with stochastic Gabor
dictionaries. Other possible estimates are presented on simulated
signal and discussed briefly. Consistence of results with previous
findings is evaluated on the data from a classical voluntary finger
movement experiment.
Software:
A program for matching pursuit decomposition of signals in stochastic
time-frequency dictionaries, with complete source code and binaries
for GNU/Linux and Windows, plus Java tool for t-f visualization of
results, is available at http://brain.fuw.edu.pl/mp. This
implementation is described in the paper "Stochastic Time-Frequency
Dictionaries for Matching Pursuit", IEEE Tran. SP, March 2001,
available at
http://brain.fuw.edu.pl/~durka/papers/stochmp
Sincerely,
Piotr Durka
-- Piotr J. Durka, Ph.D., Asst. Prof. Warsaw University, Laboratory of Medical Physics, ul. Hoza 69, 00-681 Warszawa, Poland phone (48 22) 6254535, fax (48 22) 6226154 http://brain.fuw.edu.pl/~durka
|
Home Login Meetings Courses Belgium Maillist Credits Servers Links Archive
Please send comments and suggestions to
Mike Wijnants. |
|
Page last updated on Thursday, 30-Aug-2001 17:29:15 CEST © BBF 1998 all rights reserved |