Electromyographic signal capture

Abstract

Electromyography (EMG) signals are widely used both in the clinical and research fields to analyse neuromuscular activation and neuromuscular diseases. The amplifiers that are usually used to read the signals are considerable in size, consume a substantial amount of power and cost hundreds if not thousands of euros. Furthermore, fluctuating skin-toelectrode impedance which introduces powerline interference (PLI) increases the complications. Consequently, few EMG capturing devices are available on the consumer market and those present have considerable limitations. This project aims to investigate the feasibility of having a cheap micro-module multi-channel acquisition system with on-board electrodes, design novel front-end and conditioning analog systems which limit the performance dependency on skin-to-electrode impedance and PLI, manufacture these systems, and finally digital filtering on the EMG signals themselves. In the feasibility analysis it was confirmed that with a limited cash and power budget a 16-electrode pair acquisition system with ADCs, digital filtering and bluetooth transmission is possible to construct. A new type of front-end circuitry the transimpedance amplifier, was attempted and testing provided encouraging usability results. Four analog conditioning systems, each an improvement over the previous one were designed. Through results it could be concluded that, if the skin-to-electrode impedance increases and the leakage on the skin increases, a drive electrode very close to signal electrodes becomes ineffective as the averaging effect on the skin makes the input of the drive negligible. However, the DC voltage of the body can still be controlled. When comparing a buffered input against the transimpedance input, the latter produced slightly better results. With skin-to-electrode impedances up to 1MΩ, the corresponding magnitude plot of the system remained flat over the frequency of interest, while dipping slightly with 10 MΩ and deviating up-to -3 dB. Furthermore, PLI injection was also less at higher impedances but the system generated more pink noise. On-skin testing indicated that further research is required to refine the systems for better acquisition and reliability. Appropriate stainless-steel electrodes need to be implemented to reduce impedance and improve bandwidth. Nevertheless, through this study it was confirmed that high density acquisition systems with acceptable reliability and performance for commercial use are possible.