Myoelectric control of prostheses

The development of myoelectric control systems for powered limb prostheses has advanced rapidly in recent years. The main thrusts in this development have been in realizing self-contained prostheses and in realizing better prostheses control through improvements in the myoelectric signal processing techniques. This review considers the latter of these two areas. It first presents an historical look at myoelectric signal processing and identifies the problems. It then presents a general look at the myoelectric signal and those characteristics which give rise to these problems. A review of the literature related to various control strategies and signal processing techniques to overcome these problems is given. Finally, future trends to be expected in this area are discussed.     

Myoelectric control in neurorehabilitation

A myoelectric signal, or electromyogram (EMG), is the electrical manifestation of a muscle contraction. Through advanced signal processing techniques, information on the neural control of muscles can be extracted from the EMG, and the state of the neuromuscular system can be inferred. Because of its easy accessibility and relatively high signal-to-noise ratio, EMG has been applied as a control signal in several neurorehabilitation devices and applications, such as multi-function prostheses and orthoses, rehabilitation robots, and functional electrical stimulation/therapy. These EMG-based neurorehabilitation modules, which constitute muscle-machine interfaces, are applied for replacement, restoration, or modulation of lost or impaired function in research and clinical settings. The purpose of this review is to discuss the assumptions of EMG-based control and its applications in neurorehabilitation.     

Myoelectric control of an artificial hand for sequential movement in a tropical climate

Even a thin layer of sweat on the surface of the skin badly reduces the sensitivity of an e.m.g. channel. This has been found to be the greatest difficulty in operating any myoelectric device in hot climates. A new method for the discrimination of functions is described. This can operate an artificial hand successfully for sequential movements even if the sensitivities of the e.m.g. channels change at random within 30%. The method has been tested for five movements.     

Myoelectric control systems research at the Bio-Engineering Institute, University of New Brunswick

Myoelectric controls research began at the University of New Brunswick in 1960, and has continued without interruption since that time. The Bio-Engineering Institute has emphasized control systems which meet special clinical needs. These have included systems for amputees with only a single muscle remnant with which to control a prosthesis and, more recently, systems which provide the best possible appearance for children and for infants as young as 15 months of age. Throughout this work, however, the Institute has maintained a fundamental interest in the problem of control of complex multifunction prostheses. With recent developments in electromechanical systems this more advanced controls research will find clinical application shortly.     

Myoelectric prostheses: state of the art

The following is a brief introduction to powered prosthetics and myoelectric control. This paper reviews the present availability and clinical impact of myoelectric prostheses. A significant observation is that these systems have reached a sufficient degree of maturity that they are accepted by many health-care funding agencies as reasonable and cost-effective components of the rehabilitation process. The limitations of both the mechanical systems and the myoelectric controls are discussed in some detail, from the viewpoint of the potential user. Finally, an overview is given of current research in this field with comments on probable directions of development.     

Myoelectric properties of the cat ileocecal sphincter

Myoelectric activity and intraluminal pressures were recorded simultaneously from the ileum, ileocecal sphincter (ICS), and proximal colon in chloralose-anesthetized cats. Slow-wave activity, seen at all areas, showed coupling of frequency in the distal ileum and ICS. ICS spike activity was both isolated and associated with ileal or colonic spike activity and correlated with phasic contractions (r = 0.86; P less than 0.01). Ileal distensions caused ICS relaxation and decreased spike activity 33.8% of the time. Colonic distensions caused contraction and increased spike activity 46.9% of the time. Migrating action-potential complexes (MAPC) induced by castor oil, ricinoleic acid, or cholecystokinin propagated to the ICS and through to the colon significantly more frequently than ileal non-MAPC (P less than 0.05). Both spike potential-dependent and spike potential-independent mechanisms were involved in ICS contraction. Bethanechol increased spike activity and phasic and tonic contractions. Phenylephrine, despite loss of spike activity in all leads, caused tonic contraction of the ICS. Isoproterenol caused loss of spike activity and decreased ICS pressure. Thus, ICS myoelectric activity appears to be important in determining sphincter function during neurohumoral and mechanical stimulation, with ICS contractions occurring through both a phasic spike-related mechanism and a tonic mechanism without spike activity.     

Myoelectric signal compression using zero-trees of wavelet coefficients

Recent progress in the diagnostic use of the myoelectric signal for neuromuscular diseases, coupled with increasing interests in telemedicine applications, mandate the need for an effective compression technique. The efficacy of the embedded zero-tree wavelet compression algorithm is examined with respect to some important analysis parameters (the length of the analysis segment and wavelet type) and measurement conditions (muscle type and contraction type). It is shown that compression performance improves with segment length, and that good choices of wavelet type include the Meyer wavelet and the fifth order biorthogonal wavelet. The effects of different muscle sites and contraction types on compression performance are less conclusive.A comparison of a number of lossy compression techniques has revealed that the EZW algorithm exhibits superior performance to a hard thresholding wavelet approach, but falls short of adaptive differential pulse code modulation. The bit prioritization capability of the EZW algorithm allows one to specify the compression factor online, making it an appealing technique for streaming data applications, as often encountered in telemedicine.     

Abstract and Proportional Myoelectric Control for Multi-Fingered Hand Prostheses

Powered hand prostheses with many degrees of freedom are moving from research into the market for prosthetics. In order to make use of the prostheses’ full functionality, it is essential to study efficient ways of high dimensional myoelectric control. Human subjects can rapidly learn to employ electromyographic (EMG) activity of several hand and arm muscles to control the position of a cursor on a computer screen, even if the muscle-cursor map contradicts directions in which the muscles would act naturally. But can a similar control scheme be translated into real-time operation of a dexterous robotic hand? We found that despite different degrees of freedom in the effector output, the learning process for controlling a robotic hand was surprisingly similar to that for a virtual two-dimensional cursor. Control signals were derived from the EMG in two different ways, with a linear and a Bayesian filter, to test how stable user intentions could be conveyed through them. Our analysis indicates that without visual feedback, control accuracy benefits from filters that reject high EMG amplitudes. In summary, we conclude that findings on myoelectric control principles, studied in abstract, virtual tasks can be transferred to real-life prosthetic applications.     

Myoelectric changes in the triceps surae muscles under sustained contractions

Summary Synergistic behaviour of triceps surae muscles (medial gastrocnemius-MG, lateral gastrocnemius-LG, soleus-SOL) during sustained submaximal plantarflexions was investigated in this study. Six male subjects were asked to sustain an isometric plantar flexor effort to exhaustion at two different knee angles. Exhaustion was defined as the point when they could no longer maintain the required tension. The loads sustained at 0 and 120 degrees of knee flexion represented 50% and 36% of their maximum voluntary contraction (MVC) respectively. MVC was measured at 0 degree knee flexion. During the contractions, electromyograms (EMG) from the surface of the triceps surae muscles were recorded. Changes in the synergistic behaviour of the triceps surae were assessed via partial correlations of the average EMG (AEMG) between three muscle combinations; MG/LG, MG/SOL, LG/SOL, and correlation between SOL/MG+LG and MG/SOL+LG. The latter combinations were based on either common fibre type or innervation properties. Two types of synergisms were identified: trade-off and coactivation. Trade-off and coactivation synergies were defined by significant (p<0.05) positive and negative correlations respectively. Coactivation synergism was found to occur predominantly under conditions of high load or reduced length of the triceps surae, and increased with the duration of the contraction. Trade-off synergism was evident when the muscles were at their optimum length and the loads sustained were submaximum. Complete shutdown of one muscle activity was ruled out. It is postulated that, in the absence of voluntary strategies on the part of the subjects, changes in the syznergistic behaviour of the triceps surae muscles, manifested through trade-off and coactivation, is dependent on the load placed on the muscle and the muscle effectiveness as characterized by the force/length curve.     

Myoelectric signal as a quantitative measure of muscle mechanical output

The estimation of muscle tension and velocity of shortening from the myoelectric signal have been considered in numerous papers. These papers consider the estimates of each variable separately, with the other appearing in the estimation as a constant parameter. The work described in this paper develops a model for the relationship between a muscle’s mechanical outputs and the myoelectric signal. The model suggests that the myoelectric signal is related directly to the muscle mechanical power via a nonlinear differential equation in velocity of shortening. The model is general in that it includes as special cases the isometric and anisometric constant-velocity work of other authors and agrees with their experimental results. In this work anisometric experiments are performed on the biceps brachii muscle to verify the model. Estimates of muscle velocity of shortening and mechanical power are obtained from the myoelectric signal during anisometric contractions and the results agree well with the actual velocity and power. The model points out that the myoelectric signal is a direct measure of muscle tension only under isometric conditions.     

利用表面肌电信号评价肌肉疲劳的方法

表面肌电信号分析是评价局部肌肉疲劳有效的工具。过去由于受信号处理技术的限制,对肌肉疲劳的评价仅局限于等长、恒力收缩。由于适合于非平稳信号分析的频谱估计技术的发展,使得动态收缩条件下的肌肉疲劳评价得以实现,在工效学、康复医学和运动医学等方面开启了新的应用领域。本文论述了利用表面肌电信号评价肌肉疲劳,特别是在肌肉动态收缩期间评价肌肉疲劳的方法,为进一步探讨肌肉疲劳的发生机制提供理论依据。     

Myoelectric reactions to ultra-low frequency and low-frequency whole body vibration

Summary 5 healthy males were exposed to vertical sinusoidal whole body vibration (WBV) at 5 frequencies (F1=0.315 Hz, F2=0.63 Hz, F3=1.25 Hz, F4=2.5 Hz, F5=5.0 Hz) and 2 intensities (11=1.2 ms^–2 rms, F1-F5; I2=2.0 ms^–2 rms, F2–F5). Erector spinae EMGs were derived at the levels of the first thoracic (T1) and third lumbar (L3) spinous processes, rectified and synchronously averaged, as were the accelerations of the seat and the head. WBV induced vibration-synchronous EMG activity (T1 and L3) which exceeded the activity without WBV during enhanced gravitation and decreased during lowered gravitation from F1 to F3. At F4 and F5, these phase relations changed drastically, thus suggesting a different trigger mechanism. The extreme average EMG-amplitudes remained nearly constant at F1 to F3 and increased at higher frequencies. Maximum EMG activity was higher at I2 than at I1. WBV from F1 to F3 is supposed to cause tonic muscular activity triggered by the otoliths; at higher frequencies, stretch reflexes probably gain additional importance. The results hint at an increasing sensory conflict with decreasing frequency of WBV and are interpreted within the theoretical framework of different modes of motor control. Relations between transmissibility and muscle activity suggest the usefulness of including time-variant spring-characteristics into biomechanical models.     

Myoelectric power spectrum dependence on muscular contraction level of elbow flexors

Summary The influence of the strength of contraction on surface recorded myoelectric power spectra was studied for three elbow flexors. Four subjects performed brief (3–5 s) isometric contraction levels (5–80% MVC). The experiment was repeated 23–26 times on different days. The surface myoelectric signal was recorded from the biceps brachii, the brachialis and the brachioradialis. By fast Fourier transform the myoelectric power spectrum was computed. The mean power frequency (MPF) was calculated and used as a single estimate of the myoelectric power spectrum. The MPF was found to increase with contraction strength with low level contractions. At levels in excess of 25–30% of MVC, the MPF became independent of contraction level. This dependence of the MPF on low level contractions is explained by tissue filtering effects and the recruitment order and distribution of motor units.This investigation was supported by grants from the Swedish Work Environment Fund and the Medical Faculty of Umeå University     

Stability of Myoelectric Slow Waves and Contractions Recorded from the Distal Colon

The stability of physiological activity in the distal colon was investigated by recording 5-6 hours in each of 6 healthy adults. Contractions and myoelectric slow waves were recorded from the sigmoid colon (25-30 cm from the skin surface) and rectum (10-15 cm), and pressure waves were recorded from the proximal small intestine. The activity index (sum of areas of all waves divided by recording time) varied by 200% to 800% across 4-min samples for all motility and myoelectric slow wave recordings. Spectral analysis indicated that contractile activity waxed and waned in a cycle with a period of 40-55 min in the colon and 64-80 min in the small intestine. Myoelectric slow wave activity in the colon cycled with a period of 30-40 min. Contractile activity in the sigmoid colon was correlated with similar activity in the rectum, but myoelectric slow wave activity in the colon was not correlated with myoelectric slow waves in the rectum. The frequency composition of contractions and slow waves was unstable over time.     

握速可调式肌电假手的系统研究

针对假手无法根据使用者的意愿来控制其握合速度的问题,设计了一种握速可调的智能肌电假手。在该肌电假手系统中,电极引入的表面肌电信号,放大150倍后,在频域进行信号处理。经过阈值和等值计算,最终输出控制信号到驱动电路对假手的直流微电机进行控制。10名受试者参加了抓握试验,每人重复5次,记录输入肌电信号强度为500、1 000、1 500 mV时的输出电流、输出电压和握力等值,并拟合假手输出电压、输出电流、握力与输入肌电信号大小的曲线。拟合结果表明,假手速度与输出电压成正相关,在带有负载且不过载时,握力随着输出电流增大而增大。本假手系统具有稳定性,同时和假手握速可通过检测到的肌电信号强弱进行调节,提高了假手的灵活性。     

Myoelectric control of a computer animated hand: a new concept based on the combined use of a tree-structured artificial neural network and a data glove

This paper proposes a new learning set-up in the field of control systems for multifunctional hand prostheses. Two male subjects with a traumatic one-hand amputation performed simultaneous symmetric movements with the healthy and the phantom hand. A data glove on the healthy hand was used as a reference to train the system to perform natural movements. Instead of a physical prosthesis with limited degrees of freedom, a virtual (computer-animated) hand was used as the target tool. Both subjects successfully performed seven different motoric actions with the fingers and wrist. To reduce the training time for the system, a tree-structured, self-organizing, artificial neural network was designed. The training time never exceeded 30 seconds for any of the configurations used, which is three to four times faster than most currently used artificial neural network (ANN) architectures.     

Myoelectric response of the human triceps brachii to displacement-controlled oscillations of the forearm

Summary The dynamic relations between the surface myoelectric activity in tonically contracting triceps brachii and the forearm rotation (proportional to triceps stretch) were measured by imposing small, sinusoidal, displacement-controlled perturbations on the forearm position. Three normal, adult, male subjects participated in these experiments. The amplitude of the forearm rotation, the driving frequency, and the tonic contraction level were all carefully regulated. The mean rectified triceps EMG (the output) showed a strong harmonic at the driving frequency, and the frequency-response characteristics were computed directly by comparing the amplitude and phase of this harmonic to that of the forearm flexion angle (the input). The (electrical) reflex gain is defined as the amplitude ratio of output to input. The system response was measured from 2 to 18 Hz, at two tonic contraction levels and two forearm rotation amplitudes, about a mean position of 90° forearm flexion. The results show clearly that the system response is nonlinear: the reflex gain decreases with forearm rotation amplitude. (This gain also increases with tonic contraction level for sufficiently low values of the latter variable.) The measured frequency-response characteristics of the system can be modeled approximately as a second order linear lead filter with a single time delay, followed by a saturating nonlinearity. Both model independent estimates and least-squares model fitting, yielded values of the time delay of the order of 25 ms, suggesting that a segmental mechanism mediates reflex activity. Simplified calculations and limited measurements are presented to show that a nonlinear system of the type we have identified with constant displacement driving may appear linear under constant torque driving. Our directly-measured frequency-response characteristics differ from those reported by investigators employing random, rather than periodic, driving; possible reasons for these apparent discrepancies are discussed.