Title :
Decomposition of MEG signals with sparse representations
Author :
Özkurt, Tolga E. ; Sun, Mingui ; Sclabassi, Robert J.
Author_Institution :
Univ. of Pittsburgh, Pittsburgh
Abstract :
We suggest an iterative method for the decomposition of MEG signals into some user-specified parts. It is based on a technique called morphological component analysis (MCA), which seeks sparse representations. A numerical simulation is carried out to reveal the performance characteristics of this method.
Keywords :
iterative methods; magnetoencephalography; medical signal processing; signal representation; MEG; iterative method; morphological component analysis; signal decomposition; sparse representations; Bayesian methods; Electroencephalography; Gaussian distribution; Independent component analysis; Iterative methods; Magnetic separation; Maximum likelihood estimation; Numerical simulation; Source separation; Surges;
Conference_Titel :
Bioengineering Conference, 2007. NEBC '07. IEEE 33rd Annual Northeast
Conference_Location :
Long Island, NY
Print_ISBN :
978-1-4244-1033-0
Electronic_ISBN :
978-1-4244-1033-0
DOI :
10.1109/NEBC.2007.4413304
