Jing-Yi Guo

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

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.

Relationship between shear elastic modulus and passive muscle force: An ex-vivo study

As muscle is stretched, it reacts with increasing passive resistance. This passive force component is important for normal muscle function. Unfortunately, direct measurement of passive muscle force is still beyond the current state-of-the-art. This study aimed to investigate the feasibility of using Supersonic shear wave elastography (SSWE) to indirectly measure passive muscle force. Sixteen gastronomies pars externus and 16 tibialis anterior muscles were dissected from 10 fresh roaster chickens. For each muscle specimen, the proximal bone-tendon junction was kept intact with its tibia or femur clamped in a fixture. Calibration weights (0-400 g in 25 g per increment) were applied to the distal tendon via a pulley system and muscle elasticity was measured simultaneously using SSWE. The measurements were repeated for 3 cycles. The elasticity-load relationship of each tested muscle for each loading cycle was analyzed by fitting a least-squares regression line to the data. Test-retest reliability was evaluated using intraclass correlation coefficient (ICC). Results demonstrated that the relationships between SSWE elasticity and passive muscle force were highly linear for all the tested muscles with coefficients of determination ranging between 0.971 and 0.999. ICCs were 0.996 and 0.985, respectively, for the slope and y-intercept parameters of the regression lines, indicating excellent reliability. These findings indicate that SSWE, when carefully applied, can be a highly reliable technique for muscle elasticity measurements. The linear relationship between SSWE elasticity and passive muscle force identified in the present study demonstrated that SSWE may be used as an indirect measure of passive muscle force.

Fuzzy approximate entropy analysis of chaotic and natural complex systems : detecting muscle fatigue using electromyography signals

In the present contribution, a complexity measure is proposed to assess surface electromyography (EMG) in the study of muscle fatigue during sustained, isometric muscle contractions. Approximate entropy (ApEn) is believed to provide quantitative information about the complexity of experimental data that is often corrupted with noise, short data length, and in many cases, has inherent dynamics that exhibit both deterministic and stochastic behaviors. We developed an improved ApEn measure, i.e., fuzzy approximate entropy (fApEn), which utilizes the fuzzy membership function to define the vectors’ similarity. Tests were conducted on independent, identically distributed (i.i.d.) Gaussian and uniform noises, a chirp signal, MIX processes, Rossler equation, and Henon map. Compared with the standard ApEn, the fApEn showed better monotonicity, relative consistency, and more robustness to noise when characterizing signals with different complexities. Performance analysis on experimental EMG signals demonstrated that the fApEn significantly decreased during the development of muscle fatigue, which is a similar trend to that of the mean frequency (MNF) of the EMG signal, while the standard ApEn failed to detect this change. Moreover, fApEn of EMG demonstrated a better robustness to the length of the analysis window in comparison with the MNF of EMG. The results suggest that the fApEn of an EMG signal may potentially become a new reliable method for muscle fatigue assessment and be applicable to other short noisy physiological signal analysis.

Quantifying the passive stretching response of human tibialis anterior muscle using shear wave elastography

BACKGROUND Quantifying passive stretching responses of individual muscles helps the diagnosis of muscle disorders and aids the evaluation of surgical/rehabilitation treatments. Utilizing an animal model, we demonstrated that shear elastic modulus measured by supersonic shear wave elastography increases linearly with passive muscle force. This study aimed to use this state-of-the-art technology to study the relationship between shear elastic modulus and ankle dorsi-plantarflexion angle of resting tibialis anterior muscles and extract physiologically meaningful parameters from the elasticity-angle curve to better quantify passive stretching responses. METHODS Elasticity measurements were made at resting tibialis anterior of 20 healthy subjects with the ankle positioned from 50° plantarflexion to up to 15° dorsiflexion at every 5° for two cycles. Elasticity-angle data was curve-fitted by optimizing slack angle, slack elasticity, and rate of increase in elasticity within a piecewise exponential model. FINDINGS Elasticity-angle data of all subjects were well fitted by the piecewise exponential model with coefficients of determination ranging between 0.973 and 0.995. Mean (SD) of slack angle, slack elasticity, and rate of increase in elasticity were 10.9° (6.3°), 5.8 (1.9) kPa, and 0.0347 (0.0082) respectively. Intraclass correlation coefficients of each parameter were 0.852, 0.942, and 0.936 respectively, indicating excellent test-retest reliability. INTERPRETATION This study demonstrated the feasibility of using supersonic shear wave elastography to quantify passive stretching characteristics of individual muscle and provided preliminary normative values of slack angle, slack elasticity, and rate of increase in elasticity for human tibialis anterior muscles. Future studies will investigate diagnostic values of these parameters in clinical applications.

Continuous monitoring of electromyography (EMG), mechanomyography (MMG), sonomyography (SMG) and torque output during ramp and step isometric contractions

In this study we simultaneously collected ultrasound images, EMG, MMG from the rectus femoris (RF) muscle and torque signal from the leg extensor muscle group of nine male subjects (mean±SD, age=30.7±.4.9 years; body weight=67.0±8.4kg; height=170.4±6.9cm) during step, ramp increasing, and decreasing at three different rates (50%, 25% and 17% MVC/s). The muscle architectural parameters extracted from ultrasound imaging, which reflect muscle contractions, were defined as sonomyography (SMG) in this study. The cross-sectional area (CSA) and aspect ratio between muscle width and thickness (width/thickness) were extracted from ultrasound images. The results showed that the CSA of RF muscles decreased by 7.25±4.07% when muscle torque output changed from 0% to 90% MVC, and the aspect ratio decreased by 41.66±7.96%. The muscle contraction level and SMG data were strongly correlated (R(2)=0.961, P=0.003, for CSA and R(2)=0.999, P<0.001, for width/thickness ratio). The data indicated a significant difference (P<0.05) in percentage changes for CSA and aspect ratio among step, ramp increasing, and decreasing contractions. The normalized EMG RMS in ramp increasing was 8.25±4.00% higher than step (P=0.002). The normalized MMG RMS of step contraction was significantly lower than ramp increasing and decreasing, with averaged differences of 12.22±3.37% (P=0.001) and 12.06±3.37% (P=0.001), respectively. The results of this study demonstrated that the CSA and aspect ratio, i.e., SMG signals, can provide useful information about muscle contractions. They may therefore complement EMG and MMG for studying muscle activation strategies under different conditions.

Sonomyographic responses during voluntary isometric ramp contraction of the human rectus femoris muscle

This paper aims to investigate the relationship between torque and muscle morphological change, which is derived from ultrasound image sequence and termed as sonomyography (SMG), during isometric ramp contraction of the rectus femoris (RF) muscle, and to further compare SMG with the electromyography (EMG) and mechanomyography (MMG), which represent the electrical and mechanical activities of the muscle. Nine subjects performed isometric ramp contraction of knee up to 90% of the maximal voluntary contraction (MVC) at speeds of 45, 22.5 and 15% MVC/s, and EMG, MMG and ultrasonography were simultaneously recorded from the RF muscle. Cross-sectional area, which was referred to as SMG, was automatically extracted from continuously captured ultrasound images using a newly developed image tracking algorithm. Polynomial regression analyses were applied to fit the EMG/MMG/SMG-to-torque relationships, and the regression coefficients of EMG, MMG, and SMG were compared. Moreover, the effect of contraction speed on SMG/EMG/MMG-to-torque relationships was tested by pair-wise comparisons of the mean relationship curves at different speeds for EMG, MMG and SMG. The results show that continuous SMG could provide important morphological parameters of continuous muscle contraction. Compared with EMG and MMG, SMG exhibits different changing patterns with the increase of torque during voluntary isometric ramp contraction, and it is less influenced by the contraction speed.

Sonomyography (SMG) Control for Powered Prosthetic Hand: A Study with Normal Subjects

Our previous studies have demonstrated that the muscle thickness change detected by ultrasonography during contraction, namely sonomyography (SMG), can be used for functional assessment of skeletal muscles and has the potential for prosthetic control. In this study, we further investigated the feasibility of using one-dimensional SMG (1-D SMG) signal for controlling a powered prosthesis with one degree of freedom. The performance of SMG control in visual pursuit tracking of opening-closure patterns of the prosthesis was evaluated. Nine normal subjects including seven males and two females participated in the experiment. SMG signals were collected from the extensor carpi radialis muscle to control the opening position of the prosthetic hand. The subjects were instructed to perform the wrist extension movement to match the prosthesis response to the target sinusoid and square tracks under different movement rates as accurately as possible. The normalized root mean square (RMS) tracking error between the target track and the degree of the prosthetic hands opening position, which was measured by an electronic goniometer, was calculated to evaluate the control performance. It was found that the mean RMS tracking errors of SMG control under different movement rates were 12.8 +/- 3.2% (mean +/- SD) and 14.8 +/- 4.6% for sinusoid and square tracks, respectively. Two-way analysis of variance (ANOVA) revealed significant differences in RMS tracking errors among the three movement rates (p = 2.0 x 10(-6)) and between the two target tracks (p = 0.007). The results suggest that SMG signal, based on further improvement, has potential to be an alternative method for prosthetic control.

Cross-fuzzy entropy: A new method to test pattern synchrony of bivariate time series

A new method, namely cross-fuzzy entropy (C-FuzzyEn) analysis, that can enable the measurement of the synchrony or similarity of patterns between two distinct signals, is presented in this study. With the inclusion of fuzzy sets, the similarity of vectors is fuzzily defined in C-FuzzyEn based on the exponential function and their shapes, rather than on the Heaviside function used in the conventional cross sample entropy (C-SampEn). Tests on simulated data sets and real EEG signals showed that C-FuzzyEn was superior to C-SampEn in several aspects, including giving the entropy definition in the case of small parameters, better relative consistency, and less dependence on record length. The proposed C-FuzzyEn was then applied for the analysis of simultaneously recorded electromyography (EMG) and mechanomyography (MMG) signals during sustained isometric contraction for monitoring local muscle fatigue. The results showed that the C-FuzzyEn of EMG-MMG signals decreased significantly during the development of muscle fatigue. The C-FuzzyEn showed a similar trend with the mean frequency (MNF) of EMG, the commonly used muscle fatigue indicator. However, C-FuzzyEn of EMG-MMG demonstrated a better robustness to the length of the analysis window in comparison with the MNF of EMG. The results suggested that the proposed C-FuzzyEn of EMG-MMG may potentially become a new reliable method for muscle fatigue assessment. It can also be applied to other bivariate signals extracted from complex systems with short data lengths in noisy backgrounds.

Dynamic monitoring of forearm muscles using one-dimensional sonomyography system

We introduce a method, known as one-dimensional sonomyography (1-D SMG), that uses A-mode ultrasound signals to detect dynamic thickness changes in skeletal muscle during contraction. We custom-designed a 1-D SMG system to collect synchronized A-mode ultrasound, joint angle, and surface electromyography (EMG) signals of forearm muscles during wrist extension. We extracted the 1-D SMG signal from the ultrasound signal by automatically tracking the corresponding echoes, which we then used to calculate muscle thickness changes. We tested the right forearm muscles of nine nondisabled young subjects while they performed wrist extensions at 15.0, 22.5, and 30.0 cycles/min and their largest wrist extension angle ranged from 80 degrees to 90 degrees . We found that the muscle deformation and EMG root mean square signals correlated linearly with wrist extension angle. The ratio of deformation to wrist angle was significantly different among the subjects (p < 0.001) but not among the trials of different extension rates for each subject (p = 0.9). The results demonstrate that 1-D SMG can be reliably performed and that it has the potential for skeletal muscle assessment and prosthesis control.

Test-Retest Reliability, Repeatability, and Sensitivity of an Automated Deformation-Controlled Indentation on Pressure Pain Threshold Measurement

OBJECTIVE The purpose of this study was to construct a computerized deformation-controlled indentation system and compare its test-retest reliability, repeatability, and sensitivity with a manual algometer for pressure pain threshold (PPT) measurements. METHODS Pressure pain threshold measurements were made on 16 healthy subjects for 2 sessions on bilateral erector spinae muscles at L1, L3, and L5 spinal levels, consisting of 5 repeated trials each using computerized algometry on one side and manual algometry on the other side. Mean, SD, coefficient of variation, standard error of measurement, minimal detectable change, and intraclass correlation coefficient were calculated for both manual and computerized PPT measurements. Effects of session, level, method, and side on PPT measurements were evaluated using analysis of variance. RESULTS Manual PPT measurements were significantly larger than computerized PPT measurements (P = .017), and session 2 was significantly larger than session 1 (P = .021). Coefficient of variation, intraclass correlation coefficient, standard error of measurement, and minimal detectable change of the manual and computerized PPT measurements were 10.3%, 0.91, 0.19 kg/cm(2), and 0.54 kg/cm(2) and 15.6%, 0.87, 0.26 kg/cm(2), and 0.73 kg/cm(2), respectively. CONCLUSIONS Although computerized algometry offers the benefits of eliminating the effects of operator reaction time, operator anticipation, alignment error, and variation in indentation rate on PPT measurements, these results indicate that manual algometry using load-controlled strategy may be better than computerized deformation-controlled algometry in terms of test-retest reliability, repeatability, and sensitivity. Constant load-controlled indentation protocol may be more favorable for PPT measurements. Future computerized instrumentation for PPT measurements should adopt a load-controlled mechanism.