Low-cost stimulation resistant electromyography.

Surface Electromyography (sEMG) is the non-invasive measurement of skeletal muscle contraction bio-potentials. Measuring sEMG of a stimulated muscle can prove particularly difficult due to large scale and long lasting stimulation-induced artefacts: if an sEMG device does not account for such artefacts, its measurements can be swamped and components damaged. sEMG has been used in a wide range of clinical and biomedical fields, providing measures such as muscular fatigue and subject intent. The recording of sEMG can prove difficult due to signal contamination such as movement artefact and mains interference. There are very few commercial sEMG devices that contain protection against large stimulation voltages or measures to reduce artefact transient times. Furthermore, most commercial or research level designs are not open source; these designs are effectively an inflexible black box to researchers and developers. This research presents the design, test and validation of an open source sEMG design, able to record muscle bio-potentials concurrently to electrical stimulation. The open source, low-cost nature of the design provides accessibility to researchers without the time and cost associated with design development. The design has been tested on the forearms of four able-bodied subjects during 25 Hz constant current stimulation, and has been shown to record subject volitional sEMG and M-wave without saturation.

Data captured using low-cost active electromyography.

Surface electromyography (sEMG) data was captured for three able-body subjects, from their right biceps brachii using the POLE sensor outlined in "Low-cost active electromyography" [1]. Data was captured for 45 seconds per subject, resulting in 12-21 contractions per subject. The raw data files, along with a sinusoidal waveform have been provided. This allows users of the POLE sensor to verify their low-cost sEMG device has been populated and configured correctly. This data also allows researchers/developers to compare their results against this low-cost, low noise sEMG device. The frequency content of the raw sEMG data is also of interest; this is calculated by applying a fast Fourier transform (FFT). The process applied to perform these algorithms is supplied in a MATLAB script.