Use of muscle thickness change to control powered prosthesis: a pilot study

Nowadays, most of the commercial upper-limb externally powered prosthetic devices are controlled by electromyography (EMG) signal. It is detected from the remaining muscles of amputated arms to control the prostheses. However, there are some inherent limitations of EMG control, such as muscle cross talk. On the other hand, it has been demonstrated that the muscle thickness change collected by ultrasound during contraction, namely sonomyography (SMG), could be used for muscle assessment and had the potential for prosthetic control. In this study, we investigated the feasibility of controlling a powered prosthesis by one-dimensional SMG (1-D SMG) signal and compared the performances of SMG and EMG control in tracking the guided patterns of wrist extension. SMG and EMG signals, collected from the extensor carpi radialis, were used to control the open-close motion of a powered prosthesis respectively. It was found that the mean RMS tracking errors of SMG control under different movement rates were 12.8+/-3.2% (mean+/-SD) and 14.8+/-4.6% for sinusoidal and square guiding waveforms, respectively, while the corresponding values of EMG control were 24.1+/- 5.0% and 22.9+/-5.5%, respectively. Paired t-test showed that the RMS errors of SMG control were significantly smaller than those of EMG control. The results suggested that SMG signal, based on further improvement, may have great feasibility to be an alternative method to control prostheses.