Takanori Uchiyama

Static torque-angle relation of human elbow joint estimated with artificial neural network technique

Static relations between elbow joint angle and torque at constant muscle activity in normal volunteers were investigated with the aid of an artificial neural network technique. A subject sat on a chair and moved his upper- and forearm in a horizontal plane at the height of his shoulder. The subject was instructed to maintain the elbow joint at a pre-determined angle. The wrist was then pulled to extend the elbow joint by the gravitational force of a weight hanging from a pulley. Integrated electromyograms (IEMGs), elbow and shoulder joint angles and elbow joint torque were measured. Then the relation among IEMGs, joint angles and torque was modeled with the aid of the artificial neural network, where IEMGs and joint angles were the inputs and torque was the output. After back propagation learning, we presented various combinations of IEMGs, shoulder and elbow joint angles to the model and estimated the elbow joint torque to obtain the torque-angle relation for constant muscle activation. The elbow joint torque increased and then decreased with extension of the elbow joint. This suggests that if the forearm is displaced from an equilibrium point, the torque angle relation would not act like a simple spring. In a view of the musculoskeletal structure of the elbow joint, the relation between the elbow joint angle and the moment arm of the elbow flexor muscles seems to have a dominant effect on the torque-angle relation.

Comparison of linear prediction methods based on singular value decomposition

NMR spectra are usually obtained from free-induction-decay signals by means of the computationally efficient fast Fourier transform. However, in spectral analysis by FFT, there are several inherent limitations such as low resolution for the truncated signal, or side lobes due to windowing. Recently, several least-squares alternative methods to the FFT have been proposed to reduce these problems ( J-IO). Barkhuijsen et al. applied a linear prediction (LP) method based on singular value decomposition (SVD) to FIDs (LPSVD) (1). The LPSVD method has many advantages as described in the following paragraph, although the SVD computation is very time consuming. First, no preprocessing such as windowing or zero filling is needed. Second, the procedure based on singular value decomposition enhances the numerical stability and enables one to distinguish signals from the noise. Third, the LPSVD method is a parametrical approach as it directly estimates the entire set, including frequency, damping factor, amplitude, and the phase of each component in the FID. Therefore, this approach is fundamentally different from the nonparametrical approach used in the maximum entropy method, in which the spectrum is directly estimated. This point is very important in quantitative analysis. As parametrical approaches need no complicated procedures to determine line intensities or line positions of spectral peaks, one can perform quantitative analysis more easily. A more efficient method than the LPSVD has been applied by Tirendi et al., by using the total least-squares (TLS) procedure (2). The difference between the LPSVD and TLS methods is that the TLS takes the effects of the noise in both the LP data matrix and the observation vector into consideration while the LPSVD considers only the LP data matrix (2, 11). This assumption is more realistic. Moreover, Cadzow et al. proposed using the autocorrelation-like matrix in the TLS procedure instead of the LP data matrix ( 7, 12). They insisted that the improvement in the performance made possible by using this approach in the noisy case warranted this modification. We compared these parametrical LP methods based on singular value decomposition and investigated the features and the efficiency of each method. First, we assume that the FID signal comprises K exponentially damped complex sinusoids plus a complex Gaussian white noise sequence W(n), sampled at regular times n At. The data sequence x(n) is given by

System identification of evoked mechanomyogram from abductor pollicis brevis muscle in isometric contraction

The purpose of this study is to verify the applicability of a sixth-order model to the mechanomyogram (MMG) system of the parallel-fibered muscle, which was identified from the MMG of the pennation muscle. The median nerve was stimulated, and an MMG and torque of the abductor pollicis brevis muscle were measured. The MMGs were detected with either a capacitor microphone or an acceleration sensor. The transfer functions between stimulation and the MMG and between stimulation and torque were identified by the singular value decomposition method. The torque and the MMG, which were detected with a capacitor microphone, DMMG, were approximated with a second- and a third-order model, respectively. The natural frequency of the torque, reflecting longitudinal mechanical characteristics, did not show a significant difference from that of the DMMG. The MMG detected with an acceleration sensor was approximated with a fourth-order model. The natural frequencies of the AMMG reflecting the muscle and subcutaneous tissue in the transverse direction were obtained. Both DMMG and AMMG have to be measured to investigate the model of the MMG system for parallel-fibered muscle. The MMG system of parallel-fibered muscle was also modeled with a sixth-order model.

A model of the feline medial gastrocnemius motoneuron-muscle system subjected to recurrent inhibition

Abstract.Recurrent inhibition in the mammalian spinal cord is complex, and its functions are not yet well understood. Skeletomotoneurons (α-MNs) excite, via recurrent axon collaterals, inhibitory Renshaw cells (RCs), which in turn inhibit α-MNs and other neurons. The anatomical and functional structure of the recurrent inhibitory network is nonhomogeneous, and the gain and filtering characteristics of RCs are modulated by inputs circumventing α-MNs. This complex organization is likely to play important roles for the discharge and recruitment properties of α-MNs. Modeling this system is a way of investigating hypothesized roles for normal functioning including muscle fatigue and different forms of physiological pathological tremor. In this paper, a detailed model including α-MNs, RCs, and the muscle fibers innervated by the α-MNs is presented. Outlines of the experimental data underlying the model and the modeling philosophy and procedure are presented. Then the behavior of a RC model is compared with experimental data reported in the literature. Model and experimental data agree well for burst responses elicited by synchronous single-pulse activation of different numbers of motor axons. In addition, the static relation between motor-axon activation rate and RC firing rate agree fairly well in model and experiment, and the same applies to the dynamic responses to step changes in motor-axon rate. The ultimate objective is to use this model in probing the role of recurrent inhibition in the control and stability of (isometric) muscular force under normal and altered conditions occurring during fatigue and muscle pain.

Static and dynamic input-output relations of the feline medial gastrocnemius motoneuron-muscle system subjected to recurrent inhibition: a model study

Abstract.The physiological function of spinal recurrent inhibition is still a matter of debate because of the experimental difficulty or impossibility of observing recurrent inhibition at work in normally behaving animals. The purpose of this study was to investigate, by computer simulation, the role of recurrent inhibition in shaping the input-output (I/O) relationships between descending command signals (DCS) as inputs and motoneuron (MN) and Renshaw cell (RC) firing rates and muscle force as outputs. Changing the spatial (topographical) distribution of recurrent inhibition from nonhomogeneous (as in the standard model) to homogeneous did not alter the I/O relationships significantly, while changing the functional distribution related to MN types did. Altering the global gain of recurrent inhibition, as happens naturally in various motor acts, changes the slopes and positions (at high inputs) of the I/O relationships, making recurrent inhibition a suitable means of gain control. Coupling a decrease in recurrent inhibitory gain with an increase in DCS input, as could occur during slow dynamic contractions, would increase the MN and force gains during the act. Short dynamic ramp-and-hold DCS inputs generate MN firing patterns, to which recurrent inhibition contributes interspike-interval variability and damped oscillations, which are related to issues of tremor and its control.

Superresolution of FT-NMR spectra by the maximum entropy method and AR model fitting with singular value decomposition

The complex maximum entropy method and complex autoregressive model fitting with the singular value decomposition method (SVD) were applied to the free induction decay signal data obtained with a Fourier transform nuclear magnetic resonance spectrometer to estimate superresolved NMR spectra. The practical estimation of superresolved NMR spectra are shown on the data of phosphorus-31 nuclear magnetic resonance spectra. These methods provide sharp peaks and high signal-to-noise ratio compared with conventional fast Fourier transform. The SVD method was more suitable for estimating superresolved NMR spectra than the MEM because the SVD method allowed high-order estimation without spurious peaks, and it was easy to determine the order and the rank.

Muscle stiffness estimation using a system identification technique applied to evoked mechanomyogram during cycling exercise

The aims of this study were to develop a method to extract the evoked mechanomyogram (MMG) during cycling exercise and to clarify muscle stiffness at various cadences, workloads, and power. Ten young healthy male participants were instructed to pedal a cycle ergometer at cadences of 40 and 60 rpm. The loads were 4.9, 9.8, 14.7, and 19.6 N, respectively. One electrical stimulus per two pedal rotations was applied to the vastus lateralis muscle at a knee angle of 80° in the down phase. MMGs were measured using a capacitor microphone, and the MMGs were divided into stimulated and non-stimulated sequences. Each sequence was synchronously averaged. The synchronously averaged non-stimulated MMG was subtracted from the synchronously averaged stimulated MMG to extract an evoked MMG. The evoked MMG system was identified and the poles of the transfer function were calculated. The poles and mass of the vastus lateralis muscle were used to estimate muscle stiffness. Results showed that muscle stiffness was 186-626 N /m and proportional to the workloads and power. In conclusion, our method can be used to assess muscle stiffness proportional to the workload and power.

Bioelectrical Impedance Measurement of Subcutaneous Fat Thickness Using Apparent Resistivity

Measurements of subcutaneous fat thickness provide valuable information regarding human health. In previous studies, subcutaneous fat thickness was estimated by bioelectrical impedance; however, this method required sophisticated equipment and analysis. The objective of this study was to develop a simple method to determine subcutaneous fat thickness using apparent resistivity. A single-frequency 50-kHz bipolar pulse was applied to a tetrapolar electrode, while steady-state pulses were used to determine the apparent resistivity. Subcutaneous fat thickness was determined using ultrasound tomography. We obtained a linear correlation of R = 0.916 between subcutaneous fat thickness and apparent resistivity from measurements at 20 sites on a human anterior thigh. The obtained regression equation suggests that subcutaneous fat thickness can be estimated using the apparent resistivity.

Relationship between muscle hardness estimated by the indentation method and muscle contractile level

The purpose of this study was to investigate the relationship between muscle contractile level and hardness by the indentation method. Eleven healthy male subjects were involved in this study. The subjects put their arms horizontally on a table. The subjects were instructed to keep the isometric contractile force constant at three elbow angles. The indentation depth and reaction force of the biceps brachii muscle were measured, and electromyogram (EMG) were done. First, EMGs before and during the indentation were compared. The EMGs showed no significant increase during the indentation. The indentation did not affect muscle activities. Next, the force indentation depth curves were approximated with the non-linear Voigt model by the least square method, and the indices of the elasticity and the viscosity were estimated. Then the relationship between the indices and the contractile levels were investigated. The contractile level was the force normalized by that in the maximum voluntary contraction. The elastic indices increased almost linearly as the contractile levels increased. The relationships at different elbow angles did not show significant differences. However, the characteristics of the viscous indices varied depending on the subjects.

Elastic- and viscous-like properties of the upper arm estimated by the indentation method

The purpose of this study is to estimate the elastic-and viscous-like properties of the human upper arm using the indentation method. Healthy male subjects put their upper arms horizontally on a table. An indenter was placed 1-2 mm over the upper arm. When the indenter was moved vertically with a robot arm, the displacement of the indenter and the force were measured. First, the relationships between the displacement and the force were modeled using several variations on the non-linear Voigt model. The model based on Herzs theory did not provide a good approximation of the indentation responses, but the proposed model did. Next, the indentation point dependency of the upper arm was investigated. The muscle hardness (elasticity and viscosity) was invariant in a relatively large area of the upper arm. Finally, the elastic and viscous coefficients of the upper arm were estimated when the subject was holding a weight. The heavier the weight was, the larger the elastic and viscous coefficients were. The results suggest that the muscle hardness can be evaluated quantitatively with the elastic and viscous coefficients estimated by the proposed model.