T h e electromyogram (EMG) from the gastrocnemius muscle was measured for thirty five healthy subjects during normal level walking. The signal bandpass was 20-300 Hz. The DC offset in these raw surface EMGs were eliminated. An autoregressive (AR) model of the signal was determined. The order was found to be between 8 and 12. The order chosen was 9 for our analysis. The power spectral density (PSD) of the signal was then calculated from the predicted values of the signal. The dominant frequencies of the predicted signal were found to be 60 Hz and 130 Hz. The results indicated that the AR coefficients and frequencies are repeatable measures for the specific muscle. However, further studies must be conducted before using these for defining a normal index of the EMG act ivi ty during gait. These parameters may be used for the classification of "normals" and "clinical" subjec(s in the study of muscle contribution during gait.
[1]
Gideon F. Inbar,et al.
On Surface EMG Spectral Characterization and Its Application to Diagnostic Classification
,
1984,
IEEE Transactions on Biomedical Engineering.
[2]
A Fuglsang-Frederiksen.
Quantitative electromyography. I. Comparison of different methods.
,
1987,
Electromyography and clinical neurophysiology.
[3]
J.-J.J. Chen,et al.
Temporal feature extraction and clustering analysis of electromyographic linear envelopes in gait studies
,
1990,
IEEE Transactions on Biomedical Engineering.
[4]
Steven Kay,et al.
Modern Spectral Estimation: Theory and Application
,
1988
.
[5]
Anders Fuglsang-Frederiksen,et al.
Power spectrum analysis of the EMG pattern in normal and diseased muscles
,
1989,
Journal of the Neurological Sciences.