Title :
A matrix-pencil approach to blind separation of multi-channel bio-signals
Author :
Xu, Weichao ; Chun, Francis H Y ; Chan, F.H.Y.
Author_Institution :
Dept. of Electr. & Electron. Eng., Hong Kong Univ., Hong Kong
Abstract :
Many biological and physiological processes can be modeled as a linear Multi-Input Multi-Output (MIMO) system. Blind deconvolution is ideal for retrieving the input signals and/or system parameters, given only the outputs and some input statistical information, but not the inputs themselves. In general biomedical signals sources have distinct, non-stationary, colored power spectral density, so, we use our algorithm developed before which works well for such kinds of signals. Simulation results are presented
Keywords :
Gaussian noise; deconvolution; eigenvalues and eigenfunctions; matrix algebra; medical signal processing; physiological models; spectral analysis; white noise; algorithm; biological processes; biomedical signals sources; blind deconvolution; blind separation; distinct nonstationary colored power spectral density; input signals; input statistical information; linear Multi-Input Multi-Output system; matrix-pencil approach; multi-channel bio-signals; physiological processes; simulation results; system parameters; Additive noise; Additive white noise; Blind source separation; Deconvolution; Eigenvalues and eigenfunctions; Frequency; Gaussian processes; Information retrieval; MIMO; Zinc;
Conference_Titel :
[Engineering in Medicine and Biology, 1999. 21st Annual Conference and the 1999 Annual Fall Meetring of the Biomedical Engineering Society] BMES/EMBS Conference, 1999. Proceedings of the First Joint
Conference_Location :
Atlanta, GA
Print_ISBN :
0-7803-5674-8
DOI :
10.1109/IEMBS.1999.804161
