Title :
Resolving Superimposed Action Potentials for EMG Signal Decomposition
Author :
Ren, Xiaomei ; Mo, Site ; Huang Hua ; Lihua Deng
Author_Institution :
Coll. of Electr. Eng. & Inf. Technol., Sichuan Univ., Chengdu, China
Abstract :
A practical method for resolving superimposed action potentials detected in the real electromyography (EMG) signal is describes in this paper. This algorithm is based on the peel off approach, which uses waveform peak alignment when the template waveforms have been subtracted from the superimposed action potential. Moreover, the algorithm also uses the firing pattern information to select the template waveform. Both the alignment procedure and the firing pattern information can be used to reduce the search space. The algorithm is tested using real EMG signals. After peeling off the templates from the superimposed action potentials the mean amplitude square summation of the remaining waveform was reduced to 4.96 from 45.32, which is the mean amplitude square summation of the original superimposed action potentials.
Keywords :
bioelectric potentials; electromyography; feature extraction; independent component analysis; medical signal processing; neurophysiology; EMG signal decomposition; ICA; electromyography; firing pattern information; independent component analysis; peel off approach; search space; superimposed action potentials; template waveforms; waveform peak alignment; Background noise; Electrodes; Electromyography; Filtering; Independent component analysis; Instruments; Muscles; Needles; Signal resolution; Wavelet packets;
Conference_Titel :
Bioinformatics and Biomedical Engineering (iCBBE), 2010 4th International Conference on
Conference_Location :
Chengdu
Print_ISBN :
978-1-4244-4712-1
Electronic_ISBN :
2151-7614
DOI :
10.1109/ICBBE.2010.5514674