Title :
Decomposition of electromyographic signals by iterative message passing in a graphical model
Author :
Koch, V.M. ; Loeliger, H.-A.
Author_Institution :
Dept. of Information Technol. & Electr. Eng., ETH, Zurich, Switzerland
Abstract :
The problem of decomposing electromyographic (EMG) signals is considered. Starting from a graphical model (factor graph) of the problem, a new iterative decomposition algorithm is proposed. It is based on the sum-product (or belief/probability propagation) algorithm and is able to decompose heavily superimposed single- and multi-channel signals.
Keywords :
electromyography; iterative methods; medical signal processing; message passing; physiological models; belief/probability propagation; electromyographic signal decomposition; factor graph; graphical model; heavily superimposed multi-channel signals; heavily superimposed single-channel signals; iterative message passing; sum-product; AWGN; Electrodes; Electromyography; Finite impulse response filter; Graphical models; Message passing; Muscles; Signal analysis; Signal processing algorithms; Signal resolution; EMG signal; factor graph; signal decomposition; superposition;
Conference_Titel :
Engineering in Medicine and Biology Society, 2004. IEMBS '04. 26th Annual International Conference of the IEEE
Conference_Location :
San Francisco, CA
Print_ISBN :
0-7803-8439-3
DOI :
10.1109/IEMBS.2004.1403091
