Classification of the mechanomyogram signal using a wavelet packet transform and singular value decomposition for multifunction prosthesis control

Abstract

Previous works have resulted in some practical achievements for mechanomyogram (MMG) to control powered prostheses. This work presents the investigation of classifying the hand motion using MMG signals for multifunctional prosthetic control. MMG is thought to reflect the intrinsic mechanical activity of muscle from the lateral oscillations of fibers during contraction. However, external mechanical noise sources such as a movement artifact are known to cause considerable interference to MMG, compromising the classification accuracy. To solve this noise problem, we proposed a new scheme to extract robust MMG features by the integration of the wavelet packet transform (WPT), singular value decomposition (SVD) and a feature selection technique based on distance evaluation criteria for the classification of hand motions. The WPT was first adopted to provide an effective time-frequency representation of non-stationary MMG signals. Then, the SVD and the distance evaluation technique were utilized to extract and select the optimal feature representing the hand motion patterns from the MMG time-frequency representation matrix. Experimental results of 12 subjects showed that four different motions of the forearm and hand could be reliably differentiated using the proposed method when two channels of MMG signals were used. Compared with three previously reported time-frequency decomposition methods, i.e. short-time Fourier transform, stationary wavelet transform and S-transform, the proposed classification system gave the highest average classification accuracy up to 89.7%. The results indicated that MMG could potentially serve as an alternative source of electromyogram for multifunctional prosthetic control using the proposed classification method.

Estimation of elbow flexion force during isometric muscle contraction from mechanomyography and electromyography

“…Although MMG is influenced by many factors of muscle morphology [22] and the physical milieu, such as intra-muscular pressure, muscle stiffness, and osmotic pressure, and although the sensor placement [26] and movement artifacts also affect the amplitude and spectral feature of the MMG signal, as the control signals for HMIs, MMG can provide some notable advantages over EMG [36]. First, due to its propagating property through the muscle tissue, the placement of the MMG sensor does not need to be precise or specific [2,31].…”

“…First, due to its propagating property through the muscle tissue, the placement of the MMG sensor does not need to be precise or specific [2,31]. Second, MMG is a mechanical signal; thus, it is not influenced by the change in the skin impedance due to sweating [36].…”

Novel insights into skeletal muscle function by mechanomyography: from the laboratory to the field

“…As far as the first application is concerned, MMG signal permits to discriminate differences in muscle activation between patients and controls. Although the sensitivity of MMG is not significantly higher than EMG, in the light of the lack of influence of the change in the skin impedance due to sweating and its better Sport Sci Health portability [29,118,130], some authors consider the MMG as a reliable alternative method in assessing neuromuscular activation and MTU behaviour in those conditions where the EMG cannot be easily detected. In consideration of the second point, the use of MMG as a trigger to activate or deactivate some devices represents to date a promising area of investigation.…”

“…The application fields of MMG are numerous and increasing: they span from the assessment of muscle function during isometric or dynamic contractions under different physiological conditions, such as fatigue [16][17][18], muscle temperature manipulation [19,20], stretching [21][22][23], ageing [24], training [25,26], to the analysis of the effects of rehabilitation programs [27,28], the development of prosthesis [29], and/or the use of the MMG as a triggering signal [30].…”

Symplectic geometry spectrum analysis of nonlinear time series

“…The MMG data are at a sample rate of 1000 Hz. A detailed description of the experimental protocol and data recording procedures can be found in Xie et al [48]. The reason for setting this step length is that the minimal length of a moving window satisfying real-time control in MMG or EMGbased HMI is 32 in most cases [48].…”