Myoelectric signal processing

Parker, P., Englehart, K., Hudgins, B. (2006). Myoelectric signal processing for control of powered limb prostheses. Jnl. of Electromyography and Kinesiology, 16, 541-548. 

Evans, H., Pan, Z., Parker, P., and Scott, R. (1994). Signal processing for proportional myoelectric control, IEEE Transactions on Biomedical Engineering, 41 207-211. 

Myopulse modulation (Childress, 1973) offers a means of processing myoelectric signals that maximizes the speed of response of an externally powered component. This technique eliminates the delays that are introduced by analog or digital filtering methods, provides proportional control. 

Asghari Oskoei, M., Hu, H. (2007). Myoelectric control systems - a survey. Biomedical Signal Processing and Control, 4 (4), 275-294... 

FIGURE 6.18 A method for processing EMG data for a myoelectric control signal, (a) the raw EMG data and (b) the signal after being rectified and low pass filtered (10 Hz cutoff). 

Myoelectric control derives it name from the electromyogram (EMG), which it uses as a control input. When a muscle contracts, an electric potential (the EMG) is produced as a by-product of that contraction. If surface electrodes are placed on the skin near a muscle, they can detect this signal (Fig. 32.26). The signal can then be electronically amplified, processed, and used to control a prosthesis. While the intensity of the EMG increases as muscle tension increases, the relationship is a complex nonlinear process that depends on many variables, including the position and configuration of the electrodes (Heckathome and Childress, 1981). Although the EMG is nonlinear it is broadly monotonic, and the human operator perceives this response as more or less linear. 

The Russian hand was the first semipractical myoelectric hand to be used clinically. This hand also had the distinction of being the first to use transistors (germanium) to process the myoelectric control signal (Childress, 1985). In this country, following World War II, the Committee on Artificial Limbs contracted with IBM to develop several electrically powered limbs. These were impressive engineering feats in their day but never found use outside the laboratory (Klopsteg and Wilson, 1956). 

Figure 24.16 Schematic drawing of the concept of a myoelectrically controlled hand splint. EMC signals are recorded by means of surface electrodes and are processed by a myoelectric controller that drives the actuators of the splint.

---