The electromyogram (EMG) as a control signal for functional neuromuscular stimulation. I. Autoregressive modeling as a means of EMG signature discrimination

Toward subtle intimate interfaces for mobile devices using an EMG controller

“…The simplicity of the gesture recognition algorithm illustrates that it is possible to obtain interesting results even without calibration. Further development should include the use of more advanced analysis techniques, such as autoregressive modelling, which has been reported to be successful in some EMG literature [14].…”

“…The first examples of EMG based real-time control systems can be found in the field of prosthesis control and functional neuromuscular stimulation. Hefftner et al [14], for example, report successful results for a system that can recognise two gestures generated from the shoulder and upper arm. The system must be specifically calibrated for each subject and uses EMG signals from two channels.…”

Functional neuromuscular stimulation controlled by surface electromyographic signals produced by volitional activation of the same muscle: adaptive removal of the muscle response from the recorded EMG-signal

“…For example stimulating with 20 Hz and using a adaptive filter will give an averaging time of 350 ms. In order to use the adaptive filters in a practical system, the time delay should not exceed 500 ms [9].…”

“…Assuming that the filter removes all the muscle responses and that the volitional EMG is band-limited Gaussian noise, the power of the output signal is given by out (9) where is the power of the volitional EMG in the th frame. It is assumed that the power of the volitional EMG is constant in the frames.…”

Myoelectric signal recognition using artificial neural networks in real time

“…9 Autoregressive model: Heffiner et al [2] have studied the above-lesion EMG signals for the control of FNS by using pure autoregressive parametric models. However, the use of single channel autoregressive model to extract the features of EMG linear envelope (LE) has shown rather poor classification results [4].…”