The combined contraction of about thirty kinds of facial muscles induces numerous types of facial expression and wrinkling. It induces permanent wrinkling in the human face due to aging and induces serious problems to the human face from the viewpoint of beauty science. Lanir and Fung (1993, 1974) studied a constitutive model of rabbit skin by in vitro experiments using cut-out speciments, and the pipette suction method (Sato 1997) has been applied to obtain the uniaxial load-displacement curves for blood vessels of living humans. There has been no method for obtaining multi-axial load-displacement curves for skin in living humans, however. This paper presents the design and development of a convenient automatic loading and sensing system for obtaining in vivo multi-axial load-displacement curves for the facial skin of a living human. The basic concept of the system is that since there is no edge in the facial skin of living humans, an axial end of the follow cylinder, attached to the sensor subsystem, is pressed on the skin surface. Since it generates a disc-like subregion of skin, fixed at the circular boundary, it is used as the test region. The end of the loading bar is pasted on a point near the center of the disc skin surface, and force and displacement are applied on it by moving the bar in the vertical and rotational directions. Once the load-displacement curves are obtained through a series of experiments for various directions (x-x, y-y, x-y, ...), then by combining the experimental load-displacement curves with those of FEM computer modeling, one can determine the constitutive parameters in the multiaxial stress-strain relations of facial skin, using the least squares method, so that the gap between the experimental and the calculated curves can be minimized by varying the physical parameters in the constitutive model. The results for identification were compared with those obtained by experiment for various directions, and those obtained from FEM simulation coincided well with those obtained by experiment in every direction. After the constitutive parameters were identified, another type of experiment with a different loading condition was carried out and compared with results obtained by calculations using identified material parameters, in order to verify the applicability of the method developed. Satisfactory coincidence between them was obtained.

View full abstract

In their daily lives, people perform various movements, which are caused by muscle contraction. The contraction accompanies changes in the muscle's shape and biomechanical properties. Measurement of biomechanical properties of muscle in vivo is therefore an important subject. The authors have developed an impact force measurement system for biometric use. The purpose of this study was to measure biomechanical properties of muscle in vivo from the skin surface, using that system. Before we began measuring muscle, we used experimental models to examine how far from the surface mechanical properties were detectable. It was confirmed that the detectable area reached to a depth of around 20mm and that biomechanical properties of muscle should be measured from the skin surface. The loading period of an impact force, which had been found to be related to stiffness, was also defined beforehand. First, the relation between muscle contraction and its biomechanical properties was examined. The wave heights and loading periods were calculated from impact forces as parameters to express biomechanical properties. The contraction of an antebrachial flexor was induced by making subjects grasp a hand dynamometer. Impact forces and surface electromyogram (EMG) were measured at four graduations. Integrated EMG (IEMG) was calculated to confirm the activities of muscle fiber. We found that the wave heights increased and the loading periods decreased as IEMGs increased. These changes show the muscle gets harder, which agrees with tactile hardness. Both parameters had a tendency to be saturated according to IEMGs. It was consequently confirmed that there was a significant relation between muscle contraction and biomechanical properties measured from the skin surface. Next, the relation between muscle fatigue and its mechanical properties was also examined. Muscle fatigue was caused by applying a light load to an antebrachial flexor for a specified period. Impact forces and EMG were measured, and medium frequency (MF) of the EMG power spectra was calculated as a criterion of muscle fatigue. Together with fatigue, wave heights increased and loading periods decreased. These changes show the antebrachial flexor getting harder as muscle fatigue proceeds. The parameter shifts depending on contraction averaged +0.0768N and -6.04ms, respectively. On the other hand, those depending on muscle fatigue were +0.1247N and -2.75ms, respectively. Moreover, an adequate correlation between MFs and both parameters were observed, the correlation with loading periods being strong. This clearly demonstrated that muscle hardening caused by muscle fatigue is measurable from the skin surface.

View full abstract

Chronaxie is used to describe the relative excitability of a given tissue, in this case, muscle. The strength-duration curve is a plot of the lowest current required for stimulation versus pulse duration. Chronaxie is the duration at which the current required for stimulation is twice the long-duration asymptote, the rheobase a. The current required for stimulation is equal to a+b/t, where b/a is the pulse duration, the chronaxie, which is a constant that depends partly on the type of excitable tissue. This is the Weiss-Lapique expression for the strength-duration curve, which is hyperbolic-like, with the required current increasing with decreasing pulse duration. Chronaxie has importance when subcutaneous tissues are stimulated with skin-surface electrodes. The Chronax^<[○!a]> (OG Giken, CX-2) now on the market is used for clinical examination of muscle function. It takes a fairly long time to determine the strength-duration curve. Moreover, muscle vibration must be confirmed by visual inspection. It is not practically and clinically accepted. As it is impossible to apply it to muscle fatigue examination, it should be improved. In the present study, we developed a new computer-based chronaximeter. It consists of a notebook computer, an electrical interface circuit, and a stimulation electrode. The new meter is characterized by rapid measurement for the strength-duration curve and by portability. Muscle vibration is detected with a piezoresistive accelerometer mounted on the electrodes, instead of by visual inspection. Strength-duration curves were obtained for biceps brachii, flexor carpi radialis, tibialis anterior, and rectus femoris by using a stimulation electrode situated at the motor point of the given muscle. From these curves, the chronaxie values for biceps brachii were calculated to be 0.129 ms; for flexor carpi radialis, 0.168; for tibialis anterior, 0.265; and for rectus femoris, 0.188. The chronaxie values obtained were shown to be sufficiently different.

View full abstract

To gain insight into the mechanism of MCL injury, we performed both experimental measurements and theoretical model analyses. In the initial experiment, the strain distributions over the entire MCL in a rabbit's femur-MCL-tibia complex were measured using a photoelastic coating method. The results of tensile testing revealed that the isochromatic line was concentrically higher near the tibial insertion of the MCL than in other areas, and rupture began with the avulsion of the periosteum at this site in all cases. With application of a valgus moment, the isochromatic line concentrated on the medial femoral condyle, and changed over the medial epicondyle. Next, histological studies were performed on the rupture sites. Histological observation noted that, for the specimen after a tensile test, rupture was observed in the nonmineralized fibrocartilage zone mainly at the MCL tibial insertion site, while for the specimen after valgus bending, rupture was present over the medial epicondyle from the region of the medial femoral condyle. Thus histological findings demonstrated that rupture sites and increased strain concentration sites correlated closely. Histological inspection further revealed that collagen fibers were directly inserted and anchored into osseous tissue at the femoral insertion, and the development of mineralized fibrocartilage was greater at the femoral insertion than at the tibial insertion. Investigations of actual MCL injuries showed that the ruptures commonly occurred on the femoral side irrespective of its mechanical strength. This suggested that the mechanism of MCL injury could be induced by particularly large tensile force on the femoral insertion side when a valgus moment was applied to the knee joint. We therefore attempted to introduce stress distributions on the MCL with mathematical models when a tensile force and a valgus moment were applied. In the first simulation, a constitutive equation for the MCL composite was formulated on the assumption that the MCL can be idealized as being composed of a homogeneous matrix in which densely distributed extensible fibers are embedded. Using the finite element method, we performed a simulation whose results showed that stress concentration was located in the tibial or in the femoral insertion site respectively, when a simple tensile force or a valgus moment was applied. Next, using the free body force and mechanical funicular diagram, we calculated tension on the MCL during application of valgus moment. The results demonstrated that when a valgus moment was applied, an imbalance tensile force was generated at the femoral insertion. An impingement phenomenon on the medial femoral condyle was seen as well. This phenomenon explained the higher incidence of MCL injuries on the femoral side seen in the clinical setting.

View full abstract

In recent years SQUID (Superconducting Quantum Interference Device) technology has developed rapidly in both sensitivity and number of recording channels. Biomagnetic measurements based on SQUID technology are considered to have great potential in the analysis of brain and heart functions. They are also applicable to skeletal muscles and may provide a new method for diagnosing neuromuscular functions. To clarify the capability for biomagnetic measurements, the magnetic recording technique was applied to the vastus lateralis and the vastus medialis of three healthy male adults. Magnetic fields were measured with a 64-channel SQUID system. Discharges of single motor units were simultaneously detected by surface electromyography under a weak voluntary contraction. The magnetic signals were averaged for 64 to 158 times at the zero-crossings in the surface electromyogram. Six motor units were detected in the three subjects. The isofield maps of magnetic fields showed current sources arising from the motor endplate regions and spreading in opposite directions to the tendons. A current octupole moving along muscle fibers explains these magnetic fields. Because the magnitude of the magnetic fields is directly proportional to the intensity of the currents in the muscle fibers and is independent of the conductivity of the surrounding medium under certain conditions, it is possible to calculate the intensity of the currents in the muscle fibers. To improve the accuracy of such calculations, a model of the muscle fiber action currents was developed, taking into consideration the intensity and duration of the current source. A magnetic field was calculated from an octupole current model. The measured magnetomyographic signal waveform was deconvoluted with the calculated magnetic field signal produced by a single muscle fiber. The area of the deconvoluted waveform represents the number of active muscle fibers, which was estimated at 708 to 1,791 (average 1,088±480) for the six motor units detected. These numbers were 6.5 times larger than those estimated from the intensity of the current source alone without considering its duration, and were close to the invasively obtained values. The number of muscle fibers contained in a muscle or a motor unit has until now been estimated only by an anatomical method. Noninvasive magnetic measurement should therefore contribute to the diagnosis of neuromuscular diseases that cause the decrement or shrinkage of muscle fibers.

View full abstract

Masticatory or bite forces applied to the teeth are known to generate compressive forces in the temporomandibular joint (TMJ), which are referred to as "TMJ loading." Normal TMJs are both tolerant and adaptable to a certain amount of loading, but excessive loading is believed to be one of the causes of TMJ dysfunctions. Thus TMJ loading can be considered to be controlled to a certain extent by the stomatognathic system, so as not to exceed a certain limit. In an attempt to clarify that control mechanism during biting, we previously analyzed TMJ loading using a two-dimensional static jaw model and then proved, by computer simulation, that the TMJ loading is minimized when the load vector points in a certain direction. This minimization can be achieved merely by coordinating the activities of the masseter and the temporalis. We then analyzed the modification of such coordinated activities of masticatory muscles by changing the bite point and validated the jaw model by comparing our simulation results with somatometric data. Our jaw model comprises two rigid bodies, the upper and lower jaws, and a spring-element model of the articular disk, including masticatory muscles as follows: 1) the masseter, including the internal pterygoid; 2) the temporalis; and 3) the lateral pterygoid, all of which function dominantly during biting. Bite force is assumed to be applied to a single point on the occlusal plane. Muscle forces applied to our model were determined by referring to morphological and electromyographic data reported previously. First, we carried out experiments in order to obtain masseter and temporalis forces minimizing TMJ loading under various locations of bite point on the occlusal plane. Simulation results indicated that 1) the location of the bite point solely affects the activity of the masseter under the condition that the magnitude and direction of bite force are both constant and TMJ loading is minimized; and 2) the activity balance between the masseter and the temporalis minimizing TMJ loading is sensitive to the direction of bite force. Comparison of our simulation results with EMG data for the masseter and temporalis and biteforce data demonstrated that our results almost coincided with somatometric data for biting at the canine or the premolar, and for biting weakly at the first molar. This strongly suggests that minimization of TMJ loading is actually realized by the stomatognathic system under certain bite conditions.

View full abstract

Effects of erect bipedal standing exercise on the hind-leg bone density and cross-sectional geometric properties were simultaneously investigated in seventeen growing male rats, divided into a control group and an exercise group. Using a bipedal training box, in which rats achieved a fully upright stance through positively reinforced operant conditioning, the exercise group was burdened with bipedal standing exercise from 64 days to 140 days of age, totalling 136 to 138 sessions. At the age of 140 days, the left femur was dissected out, and ten serial cross sections of the femoral diaphysis were cut from proximal to distal. In the present experiment, bone density and cross-sectional geometric properties were measured at a 46% level in a sectioned block of 45 to 50% level, using pQCT (peripheral quantitative computed tomography). The bipedal standing exercise had the following effects on the femoral diaphysis: (1) Bone density per unit volume tended to increase, though without statistical significance; (2) total bone area of the cross section of the femur shaft, cortical area of the cross section, cortical thickness (mean value), and minimum principal moment of inertia increased with statistical significance. These results suggest that adaptation of mechanical strength of a rat femur to bipedal standing exercise depends mainly on the changes in the cross-sectional geometric properties of the bone, rather than on an increase in bone density.

View full abstract

During human walking, the body accelerates during the second half of the phase in which the foot touches the ground, before it decelerates. Recently developed real-time ultrasound imaging makes it possible to observe the behavior of the muscle-tendon complex during walking. Muscle fiber (MF) of the m. gastrocnemius medialis (MG), which acts as one of the main agonist muscles, contracted isometrically during the body acceleration phase, whereas the MG tendinous tissue (TT) was lengthened. Before toe clearance, both the MF and TT shortened drastically without MG activation. No mechanical power exerted by the MF contraction was observed while the foot was on the ground, but the TT provided mechanical power in lengthening (negative) following the shortening (positive) during the body acceleration phase. It was observed that during walking the MG fiber contracts isometrically whereas the tendinous tissue is stretched, causing the storage of elastic energy in tendinous tissue. Before toe clearance, the abrupt shortening of tendinous tissue and total muscle-tendon complex is caused by release of the elastic energy stored in tendinous tissue.

View full abstract

This article summarizes newly published findings obtained in our recent studies of eye-head-trunk coordination strategies during walking. I attempt to place them within the historical framework of previous studies, and discuss underlying mechanisms that control head and eye movements during walking. Study of vertical head and trunk movements during walking throughout the range of possible speeds enabled us to obtain quantitative information about translation and rotation of the head and trunk. We also obtained a clearer picture of the motor mechanisms responsible for head movements and their relationship to trunk motion during locomotion. Our results suggest that two mechanisms are used to maintain a stable head fixation distance over the optimal range of walking velocities. The relative contribution of each mechanism to head orientation depends on the frequency of head movement and consequently on walking velocity. Considering the frequency characteristics of the compensatory head pitch, we inferred that the angular vestibulocollic reflex achieves head stability at low walking speeds, whereas the linear vestibulocollic reflex is predominately responsible for producing compensatory head pitch movement at higher speeds. We also found that the naso-occipital axis of the head aims approximately at a single point, the head fixation point (HFP), 1 meter in front of the subject. The HFP position was stable despite changes in walking velocity. This led us to design a second set of experiments in which we tested vertical eye and head coordination as a function of viewing distance during locomotion. The major finding from the second study is that during natural walking, the phase of eye velocity relative to head pitch velocity is dependent on the viewing distance, whereas head translation and rotation are relatively unaffected and tend to maintain the HFP despite changes in viewing distance. During viewing of a distant (2-meter) target, the vertical eye velocity is 180°out of phase with the head pitch velocity, indicating that the angular vestibuloocular reflex (aVOR) is generating the eye movement response. With a close (0.25-meter) target, eye velocity is in phase with head pitch and compensates for vertical head translation, suggesting that activation of the linear vestibuloocular reflex (lVOR) contributes to the eye movement response. The results of supplementary experiments using fixed-body active head pitch rotation while viewing a head-fixed target indicate that visual suppression modifies both the gain and phase characteristics of the aVOR at frequencies encountered during locomotion. We propose that visual suppression may shift the phase of the aVOR to augment the lVOR when viewing close targets during locomotion. In this context, we have taken the view that correct transduction and integration of signals from otoliths and canals is essential to maintaining stable vision and head orientation control during natural linear walking.

View full abstract

In adult walking, ground reaction force parameters are related to time-distance measurements. There is little agreement as to this relationship in the early period of independent walking. The aim of the present research was to examine the relationship between ground reaction force parameters and cadence in independent walking of one-year-old infants. Twenty-four infants, ranging from 10 to 23 months of age, were examined by means of a force plate system and a three-dimensional motion analysis system. Parents of all subjects were told the purpose and contents of this research before measurement. As control subjects, 11 adult males were examined. Maxima of the brake component and the propulsive component in infant walking are smaller than those in adult walking, when differences between subjects in body mass are cancelled by dividing by body weight. The first and second maxima of the vertical component in infants are smaller than those in adults. The minimum of the vertical component and the impulse of the transverse component in infants are larger than in adults. The cadence in infant walking is higher than in adult walking, the cadence decreasing as age increases. Though the result is the same as in previous reports, the present research revealed the new fact that the cadence of infants increases until 17 months of age, peaks at 17 to 20 months, and decreases after 20 months. The minimum of the vertical component and the impulse of the transverse component decrease, and the maxima of the brake and propulsive components increase when the cadence increases. Force parameters of infants differ from those of adults in the same range of cadence. For example, the maximum of the propulsive component of infants is remarkably smaller than that of adults. No differences were observed between infants and adults in the minimum of the vertical component, the maximum of the brake component, or the impulse of the transverse component in the same range of relative cadence (cadence×√<stature>). The maximum of the propulsive component of infants is smaller than that of adults at the same relative cadence. Comparison between infants and adults revealed that the angular displacement of the hip joint is related to the maximum of the propulsive component.

View full abstract

Introduction: The gait of older adults is characterized by such gait parameters as slow walking velocity and small range of change in joint angle. The effect of aging on these parameters has been analyzed independently, but the strength of the aging effect on each parameter has not been determined in detail. In the present study, using principal component analysis and analysis of variance, we clarified the strength of the aging effect on gait parameters. Method: The participants were 616 older adults (65 to 91 years old) living in Nangai Village located in northern Japan and 45 young adults (20 to 39 years old). The participants were classified by gender and into five age groups (aged 20 to 39 years, 65 to 69 years, 70 to 74 years, 75 to 79 years, and 80 years or above). They walked on an 11-meter straight walkway at their preferred speeds. The coordinates of markers attached to participants' iliac spines, knees, ankles, toes, and heels were measured using a Vicon-370 system (Oxford Metrics, Oxford, England) during 5 meters near the middle of the walkway. The sampling frequency was 60 Hz. From this three-dimensional data, we calculated 34 representative gait parameters. For these parameters, through principal component analysis and two-way analysis of variance by gender and age groups, we determined aging-related principal components and parameters belonging to each component. Results and Discussion: Through principal component analysis, we detected ten components with over one point of eigen value. After two-way analysis of variance for component score, we classified components according to whether they were age-related or not. The age-related components were the first principal component, with 26.0% of contribution rate. In the first component, stride length, toe height on heel contact, peak extension angle around the hip joint, and magnitude of vertical sway had a high component score; we labeled this component the "stride length" component. The parameters of cadence, single stance time, and magnitude of lateral sway had a high score in the second component (11.0% of contribution rate), which we labeled the "stride duration" component. In the second component, we detected significant gender difference. Components with less than 5% explanation of variance were mainly related to parameters on relative time in one walking cycle, such as timing of maximum knee extension. These results suggest that parameters related to stride length may strongly characterize the gait of older adults. In other words, the gait of older adults may be recognized mainly from their short stride length. The low explanation of variance for timing parameters may suggest that the pattern of kinematic change in one gait cycle is determined strictly, whereas kinematic magnitude is strongly affected by aging.

View full abstract

We proposed a neural network model for force control of skeletal muscles. The model consisted of a single motor cortex output cell, α motoneurons (MNs), Renshaw cells, and muscle units. The model was based on the size principle and had recurrent inhibition by Renshaw cells. The first purpose of this study was to construct a neural network model whose relation between the firing rate of αMNs and force is similar to that observed in human muscles. The second purpose was to investigate the mechanism for force control. The number of motor units used for models of human brachialis muscle, extensor digitorum muscle, and first dorsal interosseous muscle were 101, 99, and 101, respectively. We made the following simplification and assumptions: (1) all the motor cortex output cells (CMN) innervating one muscle were simply modeled as a single CMN; (2) a Renshaw cell, RC_M (M=1,..., n) receives excitatory impulses from all αMN and sends inhibitory impulses back to all or some of them; (3) an RC_M is located adjacent to, or is strongly connected to, an MN_M; (4) afferent impulses from various kinds of receptors are ignored while αMNs receive them. Many neural network models were constructed with various numbers of recurrent inhibitory connections from Renshaw cells to αMNs and various ranges of RIPSP, which were determined empirically considering the physiological values. For all muscles, only small αMNs fired when the force was small. Large αMNs began to fire as the force increased. This was consistent with the size principle. When the appropriate number of connections between Renshaw cells and αMNs and magnitude of RIPSP were utilized, the neural network models provided the following relationships. For brachialis muscle, the firing rate of the small and medium-sized αMNs increased rapidly after the αMNs began to fire. Then they increased gradually. Subsequently, the firing rate of the medium-sized αMNs increased rapidly once again. The neural network model of the first dorsal interosseous muscle provided an almost linear relation between the firing rate of αMNs and force. For extensor digitorum muscle, the firing rate of αMNs increased, first rapidly and then gradually, as muscle force increased. These findings agree with those for human muscles. In conclusion, the size distribution of motor units has a dominant effect on the relation between firing rate of αMNs and force. The details of the relationship were inconsistent with those observed in human muscles. In constructing the model, we need to take into account the afferent input from muscle spindle and tendon or the nonlinear inhibition by Renshaw cells. Further study is needed.

View full abstract

Recently, human skills have received a great deal of attention from researchers. Human skills have two aspects: one is motion ability and the other is spiritual ability. Violin playing, which involves both skills, was selected as an example. In this paper, we focus on Kansei as the spiritual ability in human skills. Kansei is a Japanese word that is similar in meaning to sensibility. Kansei affects human motion. The effects of Kansei on human motion have not been clarified, however. In this paper, timbre words, which are adjectives such as "bright", are considered as Kansei information, and playing parameters are considered as motion parameters. In a previous paper, the relationship between timbre words and playing parameters was analyzed in terms of a professional violinist's performance. As the next step in this study, we will analyze a listener's Kansei. The goal of this paper, therefore, is to clarify the relationship between timbre words, bowing parameters, and harmonic tone patterns from the results of sensory tests and sound analyses using fast fourier transforms (FFT). First, the process of converting a musical score into human motion that can play musical instruments is discussed and the five items of information are analyzed. These items are musical score, image, sound information, playing parameters, and playing motion. We assume that Kansei affects the process of converting a musical score into playing parameters. In order to analyze the relationship between timbre words and the three bowing parameters, we performed sensory tests on a single violin sound. Violin sounds were produced by a bowing machine that we constructed. This machine can control three bowing parameters. Each bowing parameter was set at three scales, and the sounds were evaluated by eleven subjects using twelve timbre words. The sounds were then analyzed using FFT. From the analysis of the data, we clarified the relationship between bowing parameters and timbre words, and derived the following results; (1) Eight timbre words of the twelve contributed to sound evaluation. The eight timbre words were divided into two groups: "bright" and "lean." (2) The harmonic tone patterns were divided into five groups, and two of them were evaluated by eight timbre words. (3) Bow force and sounding point (the place where the bow touches the string) affected the image of violin sounds.

View full abstract

Some kinds of myoelectric prosthetic hand, which use myoelectric signals (EMGs) of muscles as control signals for opening and closing, are used to replace the functions of a natural hand lost by amputation. Chances for an amputee to watch, touch, and use a myoelectric hand are very rare in Japan, because few hospitals give information on them and are provided with actual myoelectric hands. A simulator of a myoelectric hand can enable an amputee to use myoelectric hands virtually. The purpose of this study is to develop a 3 D-CG (three-dimensional computer graphics) simulator of an on-off control myoelectric hand and a biomimetic myoelectric hand. The former is produced by Otto Bock, and the latter was developed by our group. Both prosthetic hands are one-degree-of-freedom units with simultaneous opening and closing of thumb, index, and middle fingers. In the on-off control hand, either opening or closing is determined by the EMG amplitudes of flexor and extensor muscles. In the biomimetic hand, both the finger angle and the compliance are controlled by the EMGs. First, in this simulator, the EMGs are detected with special surface electrodes used for each hand. Second, the finger angle of the prosthetic hand is calculated from the EMGs with the model. Third, positions of the subject's upper limb are measured. Finally, the hand and the upper limb are displayed with 3 D-CG. It was found, in the on-off control hand, that the rate of change of finger angle was dependent on the finger angle. We constructed a model of the on-off control hand; the rate of angle change was determined from the output of the EMG electrode and the present finger angle. Using this model, the finger angle was calculated from the EMG electrode output and the initial finger angle. In the model of the biomimetic hand, an ideal relation between EMG input and finger angle output was utilized, although the real biomimetic myoelectric hand had complex nonlinear properties, such as friction and saturation. Using the model, the finger angle was calculated from the EMG output of the surface electrode. The positions of shoulder, elbow, and wrist were observed with a three-dimensional position sensor, and the posture of the upper limb was calculated. In this simulator, the posture of the upper limb was drawn with a stick picture and the orientation of the prosthetic hand was displayed with 3 D-CG. The experiment was conducted on normal subjects. Watching the 3D-CG of the simulator, they could easily control the finger angle and orientation of the hand.

View full abstract

Locomotion can be spontaneously generated in the relatively lower nervous system by autonomously coordinating the rhythmic activity of a Central Pattern Generator with afferent sensory information from the proprioceptors; it is not precisely controlled by the higher center. Inspired by these physiological findings, some computer simulation models of human locomotion have been proposed that mimic human locomotory neuronal mechanisms. These models, however, do not correspond neuro-physiologically to the actual human nervous system, and simulating successful locomotion does not necessarily imply elucidation of the neuronal mechanism of human locomotion. In this study, in contrast to these synthetic approaches, we attempted to estimate the human locomotory neural network, based on an actually measured kinematic and kinetic data set for human bipedal locomotion. The kinematic and kinetic data may describe structure and behavior of the neural network, since human motion is an output of the nervous system, as well as an input to it, in as much as the resultant motion is perceived by proprioceptors. Using the measured data for natural human walking and a two-dimensional musculo-skeletal model of a human lower extremity, muscular forces while walking were calculated by solving the inverse dynamic problem. The motor commands sent to muscles by alpha motoneurons can be estimated based on the calculated muscle forces. Two kinds of proprioceptors (the Golgi tendon organ and the muscle spindle), and tactile receptors on the plantar surface of the foot were considered. Activities of these sensors can also be estimated from changes in the physical quantities (such as muscle strain) that are actually perceived by each of these sensors during locomotion. The signal generated by the Central Pattern Generator is modeled as a sine curve, having a period equal to the walking period, since it basically generates the simple flexion-extension motion of a limb while walking. The estimated motor commands of the alpha motoneurons were then reconstructed by summation of the estimated signals of the proprioceptors, the foot tactile receptor, and the rhythm generators, according to the basic structure of the peripheral nervous system, such as reciprocal innervations. The calculated result suggests that the activity of alpha motoneurons can be represented by the weighted linear summation of the sensory and rhythmic inputs, indicating that locomotion can be generated autonomously by integration of the sensory inputs according to the structure of the peripheral nervous system. The estimated weights of connections were then compared to select important sensory information for generating locomotion. As a result, it was found that locomotion could be generated by a relatively simple neural network, and the sensory inputs, especially the GII signals of the muscle spindles, are more dominant than the rhythm pattern generator for generation of steady state locomotion.

View full abstract

Many computer simulation studies of human bipedal walking have been conducted in the field of biomechanics. The musculo-skeletal systems in previous models, however, have been simplified two-dimensionally, and theoretical methods in robotics have been applied for the motor control mechanisms. The purpose of this study was to develop a more precise simulation model of human walking in order to improve the practicability of the simulation method. As a result, a model was created in which the musculo-skeletal system of the entire body is modeled three-dimensionally, and a mechanism for motor control was constructed by a neuronal system model having a hierarchical structure. The inertial properties of the entire human body were represented by a three-dimensional , 14-rigid-link system. These links include the feet, calves, thighs, pelvis, lower lumbar region, upper lumbar region, thorax, upper arms, and forearms. The body's dynamic model is driven by 42 muscles for the entire body. The arrangement of each muscle was represented as a series of line segments, the direction of which changes according to joint angle. Energy consumption, including heat production, in the muscle was calculated from the generating tension. The hierarchical neuronal system includes three levels. First, at the highest level, there is a neuronal system corresponding to the higher center level. The function of adjusting changes in walking pattern is assumed to exist at this system level. The model was expressed by a computational multi-layered neural network. Second, at the middle level , there is a neuronal system corresponding to the spinal cord level. This neuronal system, representing a rhythm-generation mechanism, was modeled as a network system consisting of neural oscillators. They generate the neuronal stimulus combined for each degree of freedom by receiving nonspecific stimulus from the higher center and feedback signals from the somatic senses. Each neural oscillator is mathematically expressed by two differential equations. Third, at the lowest level, there is a neuronal system corresponding to the peripheral level. The neuronal system divides the combined neuronal stimulus from the neural oscillator into the neuronal stimulus to each muscle. The model is mathematically represented as an optimization problem. The simulated walking pattern was continuous and stable. The walking pattern closely agrees with actual human walking in terms not only of joint movement but also of muscle activities and energy consumption. In order to investigate the effects of higher center system functioning in adjusting walking patterns, we compared a walking pattern generated by a model incorporating a higher center system with the patterns obtained from a model without the higher center system, in terms of robustness of mechanical perturbation. Although the model without the higher center system could not stabilize its walking pattern and finally fell down, the model with the higher center system could perform continuous walking without falling down.

View full abstract

The Desmostylus was an ungulate mammal as large as a present hippopotamus, that lived approximately 15 million years ago. Although fossil bones of the whole body have been discovered, various postures have been restored because of the creature's unusual skeletal structure. In this study, we attempted to judge the reliability of each proposed restoration and also to investigate the creature's most probable walking style, using three different approaches: biomechanical analysis by computer simulation, comparison with living animals, and synthesis of natural motion by a quadrupedal walking machine. A three-dimensional musculoskeletal model of Desmostylus with 98 muscles and 16 bone segments was developed from the measured data for the whole body's skeletal structure and marks of muscle insertion. Fundamental standing posture was determined so as to minimize change in muscle length while walking and to maximize stride length. The sloth-type posture proposed by Domning and the crocodile-type posture proposed by Inuzuka were simulated using the musculoskeletal model. These were chosen since they are the only two restoration models that avoid dislocation of leg joints. The calculated results for muscle loads and stride length show that the crocodile-type posture is superior to the sloth-type. The results also showed that both postures must be supported by the abdomen because of the lack of muscle force. We concluded that the crocodile posture was the basic posture of the Desmostylus. The toe locus during walking was determined while maintaining an efficient range of muscular length so as to minimize body swaying in the stance phase and to maximize the toe lift in the swing phase. The cycle of leg motion was estimated from pendulum motion of the leg in the swing phase. Restored walking velocity was 36 m/min at a trot gait supported by the abdomen. The posture and walking were visualized using computer graphics techniques in order to confirm the validity of the restoration. In order to verify the restoration method, the posture and walking of an alligator (Caiman) were calculated using the same method, and the results were compared with the actual posture and walking pattern. The results obtained by computer simulation showed good agreement with those of an actual alligator, indicating that this restoration method is valid. The posture and walking of Desmostylus were also evaluated using a musculoskeletal robot. The developed 1/5-scale robot (580mm, 1.8 kg) can generate natural leg motion according to the innate musculoskeletal structures of Desmostylus. Referring to the muscle functions of an alligator, we classified the muscles of Desmostylus into the following three types: driving muscles, such as proximal big muscles; two-joint muscles, which are almost isometric during walking; and distal spring-like muscles with a long tendon. The isometric muscle and the bones parallel to the muscle act as a four-link mechanism and automatically generate coordinated leg motion. The four-link mechanism is extended by proximal driving muscles and flexed by distal spring-like muscles. The leg can therefore be driven only by two actuators for propulsion and by lifting of the toe. The restored posture and walking using the musculoskeletal robot agreed with the simulated results. Consequently, the restored Desmostylus has crocodile-type posture and trot walks supported by its abdomen. The proposed biomechanical restoration method can be easily applied to other extinct animals because it requires only the animal's skeleton and common locomotion principle. In addition, the mechanically imitated musculoskeletal robot is a useful tool for understanding the importance of body construction.

View full abstract

Exercises for the inner muscles of the shoulder have been developed using elastic bands for postinjury rehabilitation and injury prevention. It is important to maintain a careful balance between the inner muscles (rotator cuff muscles) and the outer muscles when performing various movements of the shoulder joint. If this balance is not maintained, that is if the function of the inner muscles is at a lower level than that of the outer muscles, pain or damage can arise in the shoulder. Although low loads are generally used to improve such imbalances in inner muscle exercises, little is known about appropriate low-level exercise loads for selective conditioning of the inner muscles. The purpose of our study was to determine effective ranges of loads and motion angles during low-level shoulder exercises by analyzing electromyographic (EMG) data from outer muscles. Twelve normal male subjects without any history of shoulder injuries participated in the experiments (mean age 23.3). Using a Cybex 770 dynamometer to control the exercise load and the angle of motion, the subjects performed external rotations and abduction movements for the shoulder. The range of loads was 0 to 20 N・m, the range of motion was 0 to 120 degrees, and the angular velocity for the movements was fixed at 15 deg/sec. Dynamic EMGs were recorded with surface electrodes over the ten shoulder muscles. The raw EMG signals for each load, which were separated into eight equal parts (each covering 15 degrees), were rectified and integrated for each part. The integrated EMG data, which were normalized with the EMG data for a maximum voluntary contraction, were compared for each load and each motion angle condition. For external rotation of the shoulder, loads of less than 10 N・m for motion angles below 60 degrees and of less than 8 N・m for angles below 90 degrees were found not to produce strong electromyographic activity in the outer muscles. The ranges which did not produce significant activity in the outer muscles, as compared with no-load condition, were loads of less than 6 N・m for motion angles below 30 degrees and of less than 4 N・m for angles below 45 degrees. For shoulder abduction, loads of less than 10 N・m for motion angles below 30 degrees, of less than 8 N・m for angles below 60 degrees, and of less than 4 N・m for angles below 75 degrees were found not to produce strong activity in the outer muscles. The range which did not produce significant activity in the outer muscles, compared with no-load condition, was that of loads less than 4 N・m for motion angles below 30 degrees. The results of our experiments clearly show that there are upper limits for loads and motion angles in performing external rotation and abduction movements of the shoulder without involving the outer muscles. The exercise condition under this upper limit could be recommended for the selective training of the inner muscles. We suggest, in addition, that there is a shift in the range of loads and motion angles from levels that mainly activate the inner muscles to levels that mainly activate the outer muscles.

View full abstract

Myoelectric (ME) signals are useful for estimating muscle fatigue, but the estimation of fatigue in the field is difficult. Fatigue is related not only to the neuromuscular system but also to the cardiorespiratory system. We studied a method of evaluating muscle fatigue using a superimposed M wave (SM wave) and background activity at several contraction levels. The SM wave is the electrically elicited M wave superimposed on a voluntary contraction. We estimated the mean power frequency (MPF) of background activity and the instantaneous frequency (IF) at the first peak of the SM wave. The MPF and IF were uncorrelated, or sometimes showed negative correlation, during low-level contractions. At high-level contractions, however, MPF and IF were closely correlated at the beginning and then became uncorrelated as muscle fatigue progressed. The samples {mpf, ifs} were classified into two groups, G 1 and G 2, depending on the features between two contraction phases: muscle force sustaining phase, and degeneration phase. Hence the purpose of this study was to expand our method of fatigue estimation in the field, based on results at different contraction levels. The use of low-level contractions seems suitable in field assessment. Two healthy male subjects (22 years old) ran on a treadmill for three consecutive days. On the first day, the running speed was 7.5 km/h; on the second day, an incline of 4.5 degrees was added; and on the third day, an incline of 9 degrees was used. We measured 5 minutes of biosignals at 30% of maximum voluntary contraction (MVC) during the endurance period before and after exercise. Each subject was seated in a chair, with a force transducer attached to the instep of the foot. We fixed a pair of stimulation pads on the motor point area and adjusted the stimulation levels to obtain the highest superimposed M wave. We measured heart rate, force output, and ME signals from the tibialis anterior muscle. The MPF-IF pattern of the first day was the same as those at low-level contractions. On the third day, however, they showed features of muscle fatigue like those at high-level contractions. This change in the MPF-IF patterns was probably correlated with the accumulation of muscle fatigue and agreed with subjective reports. Furthermore, the spectrum analysis of heart rate variability showed augmented autonomic nervous activity on the third day: increase in frequencies and decrease in amplitudes of the low-and high-frequency components. Our method is, therefore, effective for assessing muscle fatigue quantitatively and has potential for evaluating fatigue from local muscular fatigue.

View full abstract

Purpose Exercise of patients was carried out to maintain and improve their body functions. The load on the hip joint during exercise was analyzed using the integrated EMG and force measurement of agonist and antagonist muscles. Method Experiment 1: The relationship between the integrated EMG and the muscle force of hip abduction was examined to identify their linearity up to 100% MVC using a Cybex 6000. Twelve lower extremities of six healthy males were used. Experiment 2: The load on the hip joint was estimated through this experiment, using twenty lower extremities of ten healthy males. Exercises such as straight leg raising (SLR), hip abduction, and knee extension were performed. The integrated EMG at 100% MVC and the muscle force of agonist and antagonist were measured. Then the integrated EMGs of agonist and antagonist were measured to determine the muscle force in proportion to the force at 100% MVC. Mathematical models were used to analyze the load on the hip joint in each exercise. Results Experiment 1: The integrated EMG and the muscle force of the hip abduction showed a strong linearity up to 100% MVC. Experiment 2: In SLR, the resultant force on the hip joint was 908 N and 1.4 times body weight at 10 degrees hip flexion. It was 765 N and 1.2 times body weight at 20 degrees, and 657 N and equal to body weight at 30 degrees. In hip abduction in the lateral position, it was 1.8 times body weight at 10 degrees hip abduction, and it decreased with increasing hip abduction. In knee extension with sitting, it was 127 N and 0.2 times body weight at a 60-degree knee flexion angle, and it increased gradually with knee extension. Conclusion The analyzed values showed good agreement with those from sensorized prostheses. The proposed method in this study was considered appropriate for evaluating the load on the hip joint during exercise. In SLR, the load was 1.4 times body weight, which was unexpected. Our approach will be applicable to other exercises in a rehabilitation protocol.

View full abstract

This paper describes a biomechanical evaluation of tasks using wheeled lifting devices with a universal sling. The tasks were partitioned into eight phases: sling application, strap attachment, patient elevation, patient transport, patient lowering, strap detachment, sling removal from the legs, and sling removal from the back. Each task was measured using a Vicon system with seven cameras, a force plate, and an EMG measurement device. Three-dimensional moment at the lower back was calculated from the movement and the ground reaction force through a body segment model. The results of the EMG showed that sling application, patient transport, and sling removal from the legs required large muscle activities. Sling application and sling removal from the legs required a large peak moment and large average moment. Sling application also indicated long duration. Moments caused by an upper body weight were then calculated. As a result, these two phases were shown to generate a large moment caused by the stooped posture of the caregiver, who was bending forward during these phases. Patient transport required large twisting moments for turning of the lift to the right and left. These moments were related to a moment about a center of pressure of ground reaction force. Angles of caregiver's trunk inclination were measured during sling application and sling removal from the legs. A belt sling and a universal sling were selected and bed heights of 0.55, 0.65, and 0.73 meter, and wheelchair heights of 0.38 and 0.43 meter were set as measurement conditions. The moment index calculated from the angles showed no significant difference between the two slings in the sling application task. A lower back load index, which was the moment index multiplied by duration, however, indicated that the belt sling required significantly smaller lower back loads than did the universal sling. On the other hand, in sling removal from the legs, the moment index and lower back loads index showed that the belt sling required significantly smaller lower back loads than did the universal sling. In addition, it was indicated that a higher bed required smaller lower back loads on caregivers. There was no relationship between wheelchair height and lower back loads. Subsequently, ground reaction moments about the center of pressure of ground reaction force were measured during turning of the lifting device toward the right. There were two methods: turning it centered around the caregiver (method A), and turning the lifting device centered around itself (method B). Method B required smaller peak lower back loads than method A. Large peak moments were indicated in method B, however, when the subjects did not step in the direction of rotation. These results showed the significance of evaluating assistive devices using biomechanical methods. They will likely improve the education of caregivers and the development of new devices.

View full abstract

Methodology for design of the femoral component in total hip arthroplasty is closely related to evaluation of shape of the proximal femoral canal. Shape of a canal in a tubular bone, such as femur, is defined by the shapes of its axis and cross section. In this study we proposed a novel methodology for determination of three-dimensional shape and arrangement of the axis and cross section, using geodesic curves on a surface of a cone as its axis, the cross section of a torus as its cross section, and some measurements obtained from X-ray photographs. Geodesic curves on a surface of a cone are expressed by the following formula: [mumerical formula] Cross section of a torus is expressed as: [mumerical formula] We named a stem designed according to this method a spiral stem. The spiral stem has a smooth surface and has both spiral and taper properties. The spiral stem, which therefore has excellent fit and fill, torsional load, and dispersion of force, is expected to be designed using this novel method.

View full abstract

The durable years of artificial hip joints have progressively improved. Due to certain factors, such as aliment, burdens during the daily activity, and deterioration of the material with aging, however, the implant tends to need replacement because of partial but fatal wear. Understanding the relation of the wear and executing a proper mounting angle have therefore become important. In this study, we estimated the load charged on the socket head from an extirpated artificial hip joint. The surface properties of two samples were measured and inverse analysis was performed using the finite-element method (FEM) to determine the factor. The joint heads were measured with a heliumneon laser and shearing interferometer. The results showed that the deformations of the samples were between 1.5μm and 1.7μm. The abrasions found were similar in both implants. This is explained by the similarity of two subjects' physical situation; the implants were both placed in the right hip, and both subjects were trained to walk on a crutch. Results of the FEM inverse analysis show that the burden is loaded mainly from two directions. The load toward the longitudinal direction is extrapolated as body weight, since it fits the data previously covered in a medical report. The load toward the lateral direction matches the direction of the force from the muscles that stabilize the joint and the force that is applied from the gluteus maximus in gait. The calculated results, the computed force in the gravity direction divided by the subject's body weight and implant's period of usage, using an analysis on contact problem of Hertz, were in the range of 2.69 to 4.43 in subject A and 2.33 to 4.00 for subject B. As seen, the values for subject A were larger. The conjectured cause for this difference is that the joint in subject A had a duration of usage seven years longer than that of subject B. The analysis method in this study proved its effectiveness by presenting a match between our results and the medical report data. We intend to add more samples to the database using our measuring and analysis method. Our goal is to improve the ideal form and surface properties for designing an artificial hip joint socket head that is suitable for Japanese subjects.

View full abstract