DocumentCode
2992724
Title
Time-series based features for EMG pattern recognition: Preliminary results
Author
Knox, Robert R. ; Brooks, Dana H. ; Manolakos, Elias ; Markogiannakis, Stelios
Author_Institution
CDSP Center, Northeastern Univ., Boston, MA, USA
fYear
1993
fDate
18-19 Mar 1993
Firstpage
1
Lastpage
2
Abstract
A summary of results for features obtained from upper limb electromyographic (EMG) signals is given. The features are based on the autoregressive (AR) model and include model coefficients, reflection coefficients, and cepstral coefficients. Some of these coefficients demonstrate potential for pattern recognition of upper limb movements for a EMG controlled prosthesis
Keywords
electromyography; medical signal processing; pattern recognition; time series; EMG controlled prosthesis; EMG pattern recognition; autoregressive model; cepstral coefficients; model coefficients; reflection coefficients; upper limb electromyographic signals; upper limb movements; Acoustic reflection; Cepstral analysis; Data mining; Electrodes; Electromyography; Euclidean distance; Pattern recognition; Power system modeling; Prosthetics; Speech processing;
fLanguage
English
Publisher
ieee
Conference_Titel
Bioengineering Conference, 1993., Proceedings of the 1993 IEEE Nineteenth Annual Northeast
Conference_Location
Newark, NJ
Print_ISBN
0-7803-0925-1
Type
conf
DOI
10.1109/NEBC.1993.404442
Filename
404442
Link To Document