Understanding what strains occur over the surface area of a ligament is important, in order for the results between specimens or studies to be more reliably compared. To overcome restrictions in existing techniques and enable strain analysis over the whole surface of a ligament, we applied a photoelastic-coating method to the measurement to obtain visualized strains. We tried several photoelastic materials on various ligaments, and tested the reinforcement effect caused by the adhesion, as well as the traceability to the ligament's strain. A specific kind of polyurethane film was found to have negligible effect on the Young's modulus of the ligament. That film showed optically high fringe sensitivity. The isochromatics appearing on it closely reflected the strain behavior of the ligament. Following the above experiments, a direct application of photoelastic coating method to the human ACL was performed. A cadaver knee was fixed to a specially designed device (simulator jig). A three-dimensional displacement and rotation of the knee associated with its flexion-extension were measured with the simulator jig. Next, the medial half of the femur was removed in order for the ACL to be exposed. Free flexion-extension of the knee was reproduced by manipulating the simulator jig even without tibio-femoral articulation. Then variations of strain distribution over the entire surface of the ACL were measured using photoelasticity. The fringes appearing on the ACL were measured to obtain the values and directions of principal strains, and the strain distributions along the principal strain trajectories. Measurement results revealed that the principal strain directions were close to the fiber directions. During knee motion, it was verified that the antero-medial and postero-lateral bundles functioned reciprocally, and a zero strain area was observed on the central/posterior side near the tibial insertion. In general, the strains were rather uniform along the principal strain lines and decreased rapidly near the insertions. Yet strain distribution varied from portion to portion in complex fashion according to knee angle.

View full abstract

Muscles have elastic- and viscous-like properties in the contractile mechanism itself. These elastic- and viscous-like properties are controlled by muscle activity. It is important to maintain muscle activity constant when these properties are investigated. The purpose of this study was to investigate the elbow joint torque-angle and torque-velocity relation at constant muscle activity in voluntary contraction. Our targets of experiments were static position control and isovelocity extending of the elbow joint on a horizontal plane. A subject was instructed to maintain the posture of the upper extremity at a pre-determined position (static condition) or to extend the elbow joint at constant velocity (isovelocity condition) and not to cause coactivation of antagonist muscles. First, we measured IEMGs at various elbow joint angles, extending velocities (0, 30, 60, 90[deg/s]), and with various external loads (0, 2.8, 5.6, 8.4[%] of the maximum voluntary contraction). Then we constructed a model describing the torque-angle-IEMG relation and torque-velocity-IEMG relation with the aid of an artificial neural network technique. Finally, we estimated the torque at various elbow joint angles or various extending velocities while the IEMGs were kept constant. The elbow joint torque increased with increasing elbow joint angle in both the static condition and the isovelocity condition. The increase of elbow joint torque means that the flexor muscles of the elbow joint have a spring-like property, because the extensor muscles are stretched with the increase of elbow joint angle. This property agrees with that of a monkey's triceps. The elbow joint torque decreased with increase in the extending velocity. This means that viscous-like property is not negligible. In the case of extracted muscle fibres, the torque-velocity relation represents a hyperbolic curve. However, in this study, the relation seems to be a linear function rather than a hyperbolic curve. The main reason was that the extending velocity was much slower than the maximum velocity of elbow joint extension (1400-2200[deg/s]). In this study, the elastic-like properties of some subjects were not enhanced while the muscle activity increased. However the properties of other subjects were enhanced. It is known that extracted muscle fibres have high elasticity when they are strongly stimulated. If we assume that the contractile force and the viscosity of muscles are proportional to both the muscle activity and the muscle length (elbow joint angle), the elbow joint angle is determined by the muscle activity.

View full abstract

Baseball is one of the most popular sports in Japan and in U.S.A. However, many players, especially pitchers, have shoulder injuries incurred during pitching motions because the shoulder suffers repeated stress in the late cocking phase. In that phase the shoulder is externally rotated far beyond the normal range of motion by the inertial force applied on the forearm due to the rapid forward movement of the shoulder. Many researchers have analyzed pitching motion. However no one could calculate the inter-joint force applied between the glenoid fossa and the humeral bone. They calculated the "joint force", which makes the center of gravity of the upper arm accelerate or decelerate. This force is different from inter-joint force. The purpose of this study was to calculate the joint moment, the muscle force, and the inter-joint force on the shoulder during pitching. The subjects were five elite baseball pitchers. Twenty-five reflective markers were attached to the subjects. Four 60-Hz CCD TV cameras (Expert Vision; Motion Analysis) were used to calculate 3D locations of these markers. Shoulder joint angles were calculated from the orientation of the markers. Eulerian angles were calculated to represent the spatial orientation of the upper arm and the fore arm. Angular velocity around the local axes fixed on the upper arm and the fore arm were calculated from Eulerian angles. Joint moments were calculated from the angular velocity and angular acceleration. As the first step of this study, joint moment and joint force, which were calculated by the ordinal method, were analyzed. The ball speeds of these trials were 71 to 87% of each subject's average speed during real baseball matches, so the trials were suitable for analyzed. The average time of the late cocking period and acceleration period were 73ms and 54ms respectively. These were comparable to the 60ms and 50ms of Pappas's results. Maximum external rotation of the shoulder was about 55 degrees, which was smaller than Feltner's data. The angular velocity, joint force, and joint moment were smaller than the Feltner's results. The reason why our results were smaller than others' was partly due to the difference in physical stature between Japanese and American players. Maximum external and horizontal adduction moments were found in the late cocking phase, not in the acceleration phase. All joint moments decreased when the ball was released. Approximately 54ms prior to the ball release, the shoulder was rapidly rotated in the external direction, and it was believed to cause the over-stress on the shoulder joint. At the same time, however, the shoulder was horizontally adducted to reduce the stress.

View full abstract

The factors that determine human postural stability are not only the feedback postural control system but also the stretch reflex and joint viscoelasticity. It is expected that there is a theoretical functional sharing ratio among these components. To determine the functional sharing ratio, it is necessary to measure these components separately, but, few studies have been done. Intrinsic viscoelasticity was therefore measured using a newly developed system which allows rapid mechanical perturbation to be applied to the ankle joint of a standing subject. The viscoelasticity without stretch reflex was estimated by recursive calculation from the ankle joint angle and the torque during the first 50 ms of the measured step response. The steady-state elasticity was calculated from the torque difference at two different joint positions, as a reference value. The error between the estimated elasticity and the reference was 39 percent with plantarflexion disturbance and 32 percent with dorsalflexion disturbance. The experimental results in 10 normal subjects showed 30 nm/rad in sitting, 95 nm/rad in standing and 212 nm/rad in single-leg standing. The differences were statistically significant. Closing the eyes also showed significantly increased viscoelasticity during normal (95/121 nm/rad) and single-leg standing (212/241 nm/rad). The viscoelasticity sufficient to stabilize human upright posture without a feedback control system can be calculated as about 300 nm/rad using a pendulum model. It was experimentally confirmed that the human upright posture is stabilized by both viscoelasticity and the feedback postural control system, because the measured value was less than 300nm/rad. The increase in viscoelasticity with eye closing showed a complementary increase in viscoelasticity to compensate for the decrease in the feedback postural control gain. This suggests that the functional sharing ratio between viscoelasticity and the postural control system could be controlled, probably by a higher postural control system.

View full abstract

Joint resistance is the passive torque exerted by viscoelastic tissues such as ligaments, capsules, tendons, and muscles around the joint. The main function of joint resistance is restriction of the range of motion. However its positive role for human bipedal walking has not been clarified. In this study, we developed a three-dimensional passive-walking model that can walk on an inclined plane by utilizing gravitational force and joint resistance. By using the model, we can easily observe the influence of joint resistance on walking. The model consists of eleven rigid segments; head, chest, pelvis, upper arms, forearms, thighs, and shank-foot segments. The foot part is modeled with a semicircular plate and can roll over the slope. The joint resistance is approximated with a nonlinear viscoelastic torque element, which can prevent hyperextension and hypertwist of the joint. In order to prevent the knee joint from flexing at heel contact, minimum active torque exerted by the knee extensor during the first stance phase is measured from real walking and approximated as a nonlinear viscoelastic element. If the passive model is placed on a slope, the supporting leg naturally rotates down on the semicircular foot, and the other leg swings forward until it reaches to the slope surface. This motion is repeated to generate walking. Initial conditions of the segment angles, angular velocity, and walking velocity are determined by an optimization so as to minimize the difference in walking pattern between the first and second steps. For evaluation of the model, we calculated the passive walking with actual and artificially restricted knee properties, and these agreed well with actual walking patterns. We simulated passive walking patterns by measured changes in the knee and hip joint resistances and also the range of joint motion. Body proportions were also changed from those of a baby to those of an adult. These simulated results show the following roles and characteristics of the joint resistance by passive tissues: 1) knee joint resistance is important when active torque is applied and at the end of the stance phase, and hip joint resistance acts during only the last stance phase; 2) the supporting leg behaves like a stick during the first half stance phase; 3) elastic energy is charged up in the hip and knee joint by extension action around the ankle joint; 4) the charged energy is released at the swing phase, and the thigh swings forward and shank swings upward; 5) joint motions are not greatly affected by alternation of joint resistance; 6) the walking cycle lengthens if the resistance is weakened or the joint range becomes wider; 7) the strength of joint resistance relates to the body proportions, namely knee joint resistance relates to shank length, and hip joint resistance relates to the inertial moment of the leg; 8) the active torque around a knee joint has less influence in passive walking; 9) the patterns of joint resistance torque are similar to those of muscular torque in real walking; and 10) joint resistance saves walking energy. Consequently, we can understand that joint resistance is adapted to the body proportions and bipedal walking. This fact is useful in restoring fossil humans and their locomotion.

View full abstract

It has been suggested that the experience of falls by older adults is closely related to their decreased foot clearance during walking. The aging effect on foot clearance, however, is inconsistent in previous studies, probably because they used few subjects and a laboratory setting. In this study, we examined the relationship between foot clearance and aging for older adults participating in the TMIG-LISA (Tokyo Metropolitan Institute of Gerontology, Longitudinal Interdisciplinary Study on Aging). The participants were 628 older adults (65 to 91 years old) living in Nangai Village located in northern Japan. The participants were classified by sex and in four age groups (aged 65 to 69 years, 70 to 74 years, 75 to 79 years, and above 80 years). They walked on an 11-meter straight walkway at their own preferred speed. The coordinates of markers attached to participants' toes and heels were measured by the Vicon-370 system (Oxford Metrics, Oxford, England) during 5 meters toward the middle of the walkway. The sampling frequency was 60 Hz. Foot clearance was defined as the lowest height of the toe trajectory during the swing phase. The foot clearance was normalized by height. The mean value of foot clearance was 16.3mm (SD: 8.9mm) for males and 13.5mm (SD: 8.4mm) for females. The normalized foot clearance mean was 0.103 (SD: 0.06) for males and 0.093 (SD: 0.06) for females. A two-way analysis of variance (sex*age group) showed that the normalized foot clearance was not significantly different between the age groups and sexes. This result suggests that further investigations are needed to elucidate the functional relevance of foot clearance of the elderly.

View full abstract

The purpose of this present study is to develop a new ergometer controlling an exercise workload by the combination of heart rate and respiratory frequency in order to drive an ergometer with safe and moderate workload for exercisers. To achieve this purpose, the main system, an ergometer, is used with two subsystems for estimating and analyzing the Anaerobic Thresh-old (AT). The main system is the ergometer controlling an exercise workload by the combination of heart rate (HR) and respiratory frequency (RF). The core of the ergometer is made by Takei Electric Corporation, and is an off-the-shelf-unit. An HR detection unit is connected to an RF detection unit, and the date detected through these units are transformed by an A/D converter to be analyzed by a personal computer. These data are then transformed into information to control an exercise workload by a specific program. Feedback to the ergometer is managed through a pulse feedback control unit. The procedure to control an exercise workload is as follows; select the maximum value (Max) and the minimum value (Min) of HR and RF, and input them into a personal computer. In driving this ergometer, an exercise workload increases until either the HR or RF value reaches its Max. Soon after that the exercise workload declines to one of the Min. The exercise workload changes in triangular waves between the Max and Min for HR and RF. The subsystem has two functions, detecting and analyzing the breathing curved line (BCL). By means of a thermistor, the BCL, which is brought about due to a change of breathing gas temperatures, is detected. Through a circuit, the BCL is transformed by an A/D converter and is transformed into a power spectrum by means of FFT functions of a hard disk attached to the computer. In addition, the BCL data are forwarded into another computer, which analyzes peak-to-peak interval (PPI) variability in BCL in relation to heart rate variability.

View full abstract

The incidence of circulatory diseases caused by arteriosclerosis is increasing in Japan. Cerebrovascular and heart diseases account for about one-third of all fatalities in Japan, killing more people than cancer does. Nonetheless, there are currently no established methods to assess arteriosclerosis. Due to the recent development of high-resolution ultrasound echography, it has become possible to non-invasively diagnose arteriosclerosis. Even newly developed diagnostic methods, however, only examine the morphology of arteries, and since they cannot quantitatively assess arteriosclerosis, this disease can only be diagnosed when the morphology of arterial walls changes. By the time arterial walls have thickened or calcified, arteriosclerosis has already been established. Therefore, to prevent the onset of diseases caused by arteriosclerosis, we have to diagnose arteriosclerosis before the appearance of morphological changes and take appropriate measures. We thought that for a diagnostic method for arteriosclerosis to be truly useful in preventing diseases, the functional tissue characterization of vascular walls must be quantitatively assessed. Furthermore, from the viewpoint of preventive medicine, this diagnostic method should be noninvasive, suitable for integration into mass screening programs, and capable of assessing arteriosclerosis that is undetectable by exising ultrasound echography. We have studied the functional tissue characterization of the common carotid artery using a time series of ultrasound echography images. Shifts in the wall of the common carotid artery and their velocities were calculated from ultrasound echography images. Based on the blood pressure of the same artery, the motion of the wall of the common carotid artery was also analyzed. The blood pressure of the common carotid artery was non-invasively measured by tonometry using a continuous sphygmomanometer, and the pulse wave of the common carotid artery was determined by a carotid artery pulse wave sensor, taking into account the time gap between the common carotid artery and the radial artery. We have studied the functional tissue characterization of vascular walls in subjects that were 13 to 89 years old (mean: 57.4 years) with various health conditions (including 37 healthy individuals, 68 obese individuals, 76 hypertensive patients, 35 diabetic patients, 30 patients with hyperlipidemia, 14 patients with chronic renal failure, 12 patients with cerebral infarction, and 6 patients with old myocardial infarction). The results of clinical laboratory tests on these subjects were also recorded, and their relationships with the functional characteristics of the wall of the common carotid artery were statistically analyzed. Patients with cardiac arrhythmia were excluded from the present study. We found that the cross-sectional area of the common carotid artery and its expansion velocity are important parameters in a dynamic model of the common carotid artery. Next, we clarified that a nonlinear viscoelasticity was necessary in this type of analysis, and that the nonlinear viscoelastic parameters changed with age. By analyzing these parameters, we were able to detect functional arteriosclerosis. So far, we have found that elderly and hypertensive patients can be differentiated from the healthy individuals by analyzing the viscoelasticity of the arteries.

View full abstract

The vibration called tremor is widely seen in human beings. In the medical field, the many diseases display tremor which is typically represented by Parkinson's disease and essential tremor. The number of patients with Parkinson's disease is about 100,000 in Japan and will increase in the future. Parkinson's disease has three typical signs, namely tremor, rigidity, and hypokinesia. Tremor is the most dominant symptom in the triad. It is difficult to discriminate Parkinson's disease from several other diseases simulating Parkinson's symptoms. Medical treatment is often effective in Parkinson's disease, but in general these are transitory. In this study, we tried to discriminate these diseases by measuring tremor using accelerometers. We developed a measuring and analyzing system for tremor in the upper limbs using accelerometers, which were put on both hands. The posture of the subjects during measuring their tremor was as follows: Subjects sat on a chair and put their elbows on a desk. We placed their forearms at an angle of 45°from the horizontal plane. Tremor signals were analyzed and the dominant frequency was calculated as "main frequency". The severity of the tremor can be quantified by the spectra. We adopted the main frequency and its tremor strength as two tremor indicators. The measuring system was applied to 42 patients with Parkinson's disease (mean age, 73.8 years), 5 patients with drug-induced Parkinsonism (mean age, 72.6 years) and 36 patients with essential tremor (mean age, 67.8 years). The system could significantly discriminate the drug-induced Parkinsonian tremor from the other two types of tremor diseases (p<0.01). The degree of tremor could be quantified using an engineering index, namely, the power of the main frequency. We studied the relationship between age and the symptoms of the Parkinson's disease and essential tremor. The results were as follows: In Parkinson's disease the frequency decreased with increasing length of tremor history. In essential tremor, the main frequency decreased and the power of the frequency increased with the age of the patients. We tried to control tremor using the biofeedback technique in three healthy subjects. An effect on tremor frequency was observed during the biofeedback training. The power of high-frequencies tremors (which we defined as tremors at frequency between 5 and 15 [Hz]) increased in all patients. We assume it possible to control the patients' tremors using the biofeedback technique. Biofeedback is considered to be a useful therapy for all patients with tremor, because it is not invasive and it doesn't require any medicines such as L-Dopa.

View full abstract

When electrostimulation is applied to a living body, the subject cannot feel the stimulation if its amplitude is small or if the muscle is fatigued. The purpose of this study was to estimate muscle fatigue utilizing electrostimulation. A biphasic stimulation pulse, with a pulse width of 1ms and an interpulse interval of approximately 90 ms, is applied to the subject's lower back. Biphasic pulse pairs produce less long-term skin reddening and a more comfortable sensation. The positive pulse induces a conducted action potential, and the negative pulse reverses the electrochemical process that has resulted from the positive pulse. Therefore, balanced-charge biphasic pulses do not cause charge accumulation because the charge in the positive and negative pulses is the same. The stimulus electrode used is a large conductive rubber electrode, which can be used for low-frequency electrical stimulation. The amplitude of the stimulation pulse can be adjusted by changing the power supply voltage. As the pulse amplitude is increased gradually, the subject first feels the stimulation, and the critical perceptible stimulus intensity occurs at a time which is referred to as the electrostimulus threshold. In the study, muscle fatigue is caused by sitting on a standard or lower-quality vehicle seat during a long period of driving. The "standard seat" is the normal seat used in the test vehicle, and the "worse seat" is remodeled to induce muscle fatigue in the subject's lower back by removing some sponge rubber and two "S" springs for lumbar support. After the subject is seated, an electrostimulus is applied every 30 minutes for 3 hours and the subject then rates the subjective pain in the lower back using a six-grade classification from 0 (right after sitting) to 5 (a pain that the subject cannot endure). This suggests an adequate correlation between the subjective fatigue feeling and the stimulus threshold. The time-dependent variations, however, using the standard and worse seat show various remarkable patterns. It is apparent that the feeling of subjective fatigue can be quantified by the electrostimulus threshold.

View full abstract

The aim of this research is to examine the kinetic and kinematic changes of infant gait by longitudinal observation of infants at the onset of bipedal walking. We are especially interested in the trajectory of the foot force as a kinetic parameter which is important in determining the developmental stage of an infant. Six infants, ranging from 10 months to 2 years and 1 month of age, were examined longitudinally by means of a force plate system and a three-dimensional motion analysis system. Two characteristics of the trajectory of foot force distinguish an infant from an adult during a single stance phase. First, the sagittal track length of the trajectory of the infant, standardized by foot length, is shorter than that of the adult. The length is correlated with the duration of single stance phase. Second, the transverse sway of the trajectory is characteristic of the infant gait. The above results are related to displacement of the joints, that is, transverse displacement of the hip joint, vertical displacement of knee joint, and the pattern of transverse movements of shoulder, hip, and knee joints. The transverse displacement of the hip joint during stance phase is larger in the infant gait than in the adult gait. It decreases with development. Rapid decrease is observed between 10 months and 15 months. The vertical drop of the knee joint is conspicuous at the earlier month of independent walking. This action of the knee joint causes an impact to the force plate. The pattern of transverse movements of shoulder and hip joints differs between infants and adults. The shoulder and hip joints of infants sway together during stance phase. The shoulder and hip joints of adults first begin to move laterally following heel contact. When the load acts on one foot, the adult shoulder joint moves in the opposite direction. The knee joints of infants and adults both move in the opposite direction to the shoulder joints at that time. The sway of each joint causes instability of the center of gravity and changes the trajectory of foot force. The trajectory of foot force and the displacement of joints accordingly express the stability of the center of gravity. They can be added to the series of parameters useful in judging the developmental stage of the infant gait.

View full abstract

The purpose of this study was to clarify the changes in Galvanic Skin Reflex (GSR) and Heart Rate (HR) while appreciating a dance piece under various conditions and while imagining it, and to survey the time errors while imagining it. Conditions for appreciating the dance piece were as follows: I: audio and visual II: only accompaniment music (audio) III: only picture (visual) The length of the dance piece used in this study was about 10 minutes. Subjects were 12 female university students (4 performers of this piece, 4 students of the department of dance, and 4 general students), aged from 20 to 23 years. Results of the experiment were as follows: 1) In the group of performers, the rate of increase of HR was rather high under all conditions (from 16% to 34%). The group for which this rate was particularly high during imagining included not only performers but other students of the department of dance. 2) In the group of performers, there were similarities between the GSR patterns while appreciating the piece. In the group of students of the department of dance, the GSR pattern under condition I was similar to that under condition II. In the group of general students, there weren't similarities between the GSR patterns under any conditions. 3) In the case of a subject whose level of dance was advanced, the GSR pattern under condition I was similar to that during imagining. The results can be summarized by saying that the emotional autonomic response while appreciating a dance piece has something to do with experience of dancing. Similarities between the GSR patterns while appreciating and while imagining a dance piece, show that skill in dancing is significant factor.

View full abstract

In order to clarify the control mechanism of jaw movements, we developed an autonomous jawmovement simulator, JSN/1 D, which incorporates cable-tendon DC-servo actuators, simulating the masseter, temporalis, lateral pterygoid, and digastric muscles. The actuators were controlled adaptively under an impedance-control mechanism, utilizing data for biteforce, tooth contact, cabletension, and cablelength. In order to achieve more lifelike open-close movement on the simulator, we introduced a control hypothesis that during closing, the horizontal mandibular position is determined by antagonistic activities of the posterior portion of the temporalis muscles and the lateral pterygoid muscles. This hypothesis was materialized by an impedance control for the posterior temporalis actuators and an antagonizing tension-control of the lateral pterygoid muscles. The actuators were activated in the following manner, so as not to contradict anatomical and physiological knowledge of the jaws. The masseter actuators were activated only during biting, in order to exert bite force. The anterior temporalis actuators under impedance control were solely responsible for determining the vertical position of the mandible during closing and supported the masseter actuator during biting. During opening, the digastric and lateral pterygoid actuators were co-activated to lower the mandible. Both actuators were driven under tension control, due to the lack of the muscle spindles in corresponding muscles. At rest, all the actuators functioned as a weak elastic body like actual muscles. Experimental results demonstrated that the aforementioned control scheme can produce a reproducible lifelike open-close movement on the simulator , while referential experiments with less activation of the α-γ linkage of the posterior temporalis control showed kinetic instability, particularly in the closing phase. All these suggest a possibility of stabilizing the mandible during closing through position control of the posterior temporalis muscles and antagonistic tension-control of the lateral pterygoid muscles. A single alternative to this control scheme is to employ an excessive co-activation of the two muscles to make the position of the mandibular head less compliant. This is hardly acceptable, however, because no such strong activities are observed in both muscles during closing. Physiological validation of the proposed control scheme is a task that should be dealt with in a subsequent study.

View full abstract

I regarded an oval as a cross section of a torus (defined as closed curve E) cut by a plane II, parallel to the axis of rotation. Closed curve E is expressed as follows: [numerical formula] where a is the distance from the axis of rotation to plane II, and r is the distance from the axis of rotation to the center of the generating circle. I showed four shapes for closed curve E, which are similar to those of "real" eggs, by varying a values under constant r values. Then I demonstrated that shapes constructed by a computer are fairly similar to the "real" egg's shapes. I assumed that the most important function performed by the oval shapes of eggs is that eggs can stay near the original place when they are forced to roll. I showed that the greater the a value is, the shorter the radius of rotation of rolling eggs is.

View full abstract

The Desmostylus, known to have lateral-type limbs like a reptile, was a large mammal that lived about 15 million years ago. Most of its bones have already been discovered; due to the peculiar shape of its skeleton, however, adequacy of its posture restoration had not been verified, and a paleoecological restoration closely related to the posture had never been attempted. In this study, we constructed a three-dimensional musculo-skeletal model of the Desmostylus based on Inuzuka's restored skeleton, which laterally extending the limbs, and verified the adequacy of the peculiar posture by estimating muscular loads to sustain its posture. The musculo-skeletal model was constructed by referring to data obtained from three-dimensional measurement of the restored skeleton and a whole body model made by Inuzuka. Its body was represented by total of 16 rigid links: 3 links in each leg and 4 in the torso. The rigid link properties, such as mass and center of gravity of each segment, were calculated by representing the body shape as elliptical plates. A total of 49 muscles were modeled, including torso muscles. In order to sustain its body weight efficiently, a nonlinear elastic element, like a ligament, was attached around each joint. Body length of the model was 2.6 meters and its body weight was 1.2 tons. We calculated muscular loads to support the body at posture by varying four parameters defining its posture (height of hip joint; inclination angle of a motion plane composed by scapula, upper and fore limbs; directions of toe tips; and positions of the toe tips). In order to maintain the posture with limbs extending laterally, the elastic element at each joint became very important to reduce the muscular loads, instead of the structural support of the posture by bone. Therefore, we determined the posture so as to minimize the change of supporting moment and to maximize the step length. Constant moment can be easily exerted by passive elements. The results were as follows: 1) The calculated posture of Desmostylus is 0.45 meter hip height, 1.06 meters lateral distance between right and left forelimbs, and 0.7meter for that of hindlimbs; 2) muscle cross-sectional area to support the body weight exceeded its permissible leg thickness, therefore, abdomen should touched the ground during daily locomotion; 3) it is more feasible to extend its forelimbs forward, and hindlimbs backward, and the step length is 0.65 meter; 4) the hindlimbs were able to land at a different position from that of the forelimbs; 5) hands and feet were adapted to scratch backward if they were positioned perpendicular to the ground; 6) by inclining the neck downward about 20 degrees from the estimated posture, it could place its mandible parallel to the ground; 7) when in such a low posture, the total muscular load may be comparable with the load of a rat-type posture. We concluded that the lateral posture of Desmostylus touching its abdomen to the ground was adequate and adapted the mammal to live on tidelands.

View full abstract

In recent years, the mechanism of organic evolution has been investigated by computer simulation in the field called artificial life. The creatures simulated in these studies are mainly lower and abstract ones, and their body dynamics are not considered. However, biomechanical factors such as muscle tension certainly play important roles in the evolution of animals. That is, motions are reasonably adapted to the mechanical constraints of the body system, whereas a repetitive motion modifies body construction adapting to its motion. Clarifying the mechanism of interactive adaptation between body shape and repetitive motions is helpful for studying evolutional processes in animals. This study focuses on the evolutional adaptation observed in animals and the acquisition process of animal hopping motions, which are considered to be strongly influenced by biomechanical factors, was reproduced using the techniques of computer simulation and a neuro-musculo-skeletal model. A body model was constructed of seven rigid links in the sagittal plane, in which the right and left legs were made up into one leg. These links were driven by twelve muscles. The nervous system was modeled into a rhythm pattern generator consisting of twelve neural oscillators. Hopping motions were generated through mutual entrainment between the neural oscillation and the body dynamics of the musculo-skeletal system. The evolutional algorithms that change body motions and body shape consist of acquired processes corresponding to neural learning and inherent processes using genetic algorithms. Firstly, a group of individuals was generated according to genes. In the acquired processes, neural connectivity was modified by evaluating the adaptability of the synthesized hopping motions for each individual. In the inherent processes, individuals whose value of fitness was small and those who fell down were weeded out. In addition, genetic operations such as mutation produced various body shapes. In the initial stage of evolution, continuous and high-speed hopping motions were not synthesized. First, the locomotive-speed gap between objective and actuality, specific power, and moment of inertia of each leg were defined as the criteria for motions and fitness for genetic algorithms. As selection and multiplication were repeated, high-speed quadruped hopping motions were obtained. Second, minimizing the ground reaction force of the foreleg was added to the criteria and fitness mentioned above. In this case, bipedal hopping motions were obtained, and the body shape became kangaroo-like. From these results, it is thought that adaptation to high-speed locomotion and facility in motion control are important factors in the evolution of hopping motions and body shape. Virtual evolution, such as this study, not only provides a novel methodology for research on animal evolution, but also can be applied to various fields such as mechanical engineering.

View full abstract

We have previously developed an automatic diagnostic system for tooth mobility, for measuring the biomechanical properties of human periodontium. This paper presents a new application of the system in evaluation of oral implant function and the biological rhythm in tooth mobility, and it introduces the more favorable instrument, T-M tester and its further improvements. The IMZ implant system contains an intramobile mechanism, which was inferred to possess a stress-breaking function. To ascertain whether the IME is able to absorb or disperse oral stress effectively, we first fabricated the superstructure of the IMZ with gold alloy, then inserted a polyoxymethylene (POM) element and titanium element into the implant body alternately, and finally measured the horizontal mobility of the superstructure of IMZ with the automatic diagnostic system for tooth mobility. We found that the values of mechanical parameters (c_1, c_2, and k) in the titanium element were higher than those in natural teeth, while the POM's resemble those of natural teeth. The results indicated that the IME possessed a mobility functions which was similar to the natural teeth. For investigating the daily variation of tooth mobility, the horizontal mobility of maxillary central incisors in young women was measured by our system. The basal body temperature was also recorded simultaneously. Our results showed that daily variations in tooth mobility exist in women, and that the mobility during the lutein phase is greater than that during the follicular phase. Since the impedance head of the automatic diagnostic system is too large to measure mobility in an implant abutment and a molar, a simplified testing instrument called T-M tester was developed. In this tester, a sinusoidal vibration is applied to the tooth, the acceleration response is detected, and an MI (mobility index) corresponding to the tooth movement is obtained. The tester is characterized by its portable size and rapid measurement. It was used to measure quantitatively not only the mobility of maxillary and mandibular teeth (implants) but also diurnal variations in physiological mobility. Recently, a tooth-movement transducer using bender-type piezoelectric ceramics has been developed for the probe of the tester, making the probe even smaller and lighter than before. The new type Tooth Mobility tester is very convenient for examination of tooth mobility in any teeth, in both vertical and horizontal directions. In the near future, we will use it routinely in clinics to measure both tooth and implant mobility.

View full abstract

Experimental results are described to clarify the relation between the stance phase control function of an external prosthetic knee joint and mechanical characteristics exerted on the prosthesis in level walking, and to establish a clear standard for selecting an external prosthetic knee joint for trans-femoral prostheses. The stance phase control functions of prosthetic knee joints are (1) preventing sudden knee flexion at heel strike, (2) decreasing movement of center of gravity by small knee flexion (double knee action) in middle stance phase, and (3) shifting smoothly to swing phase at the end of stance phase. In this paper, functions (1) and (2) are examined to clarify the effect of improving the prosthetic gait. Gaits of three normal subjects using a knee-disarticulation prosthesis for experience and three trans-femoral amputees were measured by a pylon load cell installed in the shank of the prosthesis to get six force quantities, and flexible electro-goniometers made of electroconductive rubber placed on the side of ankle, knee, and hip joint. The angular displacement and joint moments of the hip, knee, and ankle were measured. Normal subjects are used to examine the experimental process for preliminary evaluation of the function of various knee joints. Prosthetic knee joints used in the experiment were 3 R 20 (solid knee joint), 3 R 15 (weight-stabilizing knee joint), Total Knee (six-bar linkage knee joint), 3 R 60 (five-bar linkage knee joint), and 3 R 49 (non-friction knee joint). The prosthetic foot used was a single axis foot, in which ankle joint stiffness was changed from hard to soft. Before the experiment, each subject was instructed how to effectively use each knee joint and experienced walking with the knee joint for a long period. Knee joints were selected from the series of 3 R 20, 3 R 15, Total Knee, 3 R 60, and 3 R 49, in which the amount of voluntary control changes from small to large. From the data for prosthetic gait in level walking, the following results are obtained. 1) The extension moment of the hip joint of trans-femoral amputees became large when the non-friction knee joint was used. The non-friction knee joint has no function to prevent knee flexion in stance phase. 2) The knee joint of trans-femoral amputees flexed about ten degrees in the middle of stance phase when the Total Knee and 3 R 60 were used. The angular displacement pattern was very close to that of normal gait, and decreased movement of the center of gravity. This knee flexion relates to the angular displacement and the moments of the hip and knee joints in the stance phase. 3) The stance phase control function of the prosthetics knee joint was significantly influenced by the stiffness of plantar flexion of the prosthetic foot. As the stiffness of plantar flexion of the prosthetic ankle joint decreased, the amount of voluntary control decreased in the case of the free knee (3 R 49), and function was impaired in the case of the other knee joints except the solid knee (3 R 20).

View full abstract

Since the introduction of the Intelligent prosthesis in the prosthetic field, many trans-femoral amputees have acquired high functional ability. This is caused by the use of a high-performance prosthetic knee and the introduction of new training programs. It became clear that the physical capacities of T/F amputees were reduced because the poor function of the prosthesis prevented them from doing physical activities to keep their bodies as active as those of sound persons. In using the IP, initial gait training through sports came to be considered as a training series divided into three steps. Conventional gait training is the first step. The second step is to change the walking speed and to walk as normally as possible. The third step is the sports training including games, running, and sprinting. When those prosthesis users who have finished at least the second step are walking, their energy consumption is about 15% higher than normal persons, whereas for those who are not using the IP, energy consumption is about 50% higher than for normal persons, especially at a high walking speed. To achieve better training results, several new training means have been introduced. These include using sensors and multimedia devices. By using sensors, training results are quantitatively evaluated, and a realtime feedback works better to make the trainee understand the instructions of the physiotherapists. A video image of the lateral view is also useful for them to see how their gait appears from the side. The concept of the IP is considered applicable to other joints, such as the hip joint and anklefoot. A hip joint is being developed with the aim of future mechatronic control. Sensor signals will be needed to control these devices. As an experiment, several sensor outputs were accumulated to test what kind of information could be extracted. As a result, cycle time, knee flexion angle, heel contact, toe off, mid stance, stance time, swing time, shank bending moment, thigh flexion angular velocity, and thigh angle in space are considered to be identifiable from those data. Prosthesis constructs a human-machine system with the user. Hardware cannot overcome a functional disability by itself alone. When a new functional device is developed, a new training means should be developed so that the human-machine system can work at its best with the user.

View full abstract

Desirable characteristics of ankle-foot orthoses (AFOs) for hemiplegic patients were clarified by the authors' previous studies. The most important mechanical characteristics of AFOs are the magnitude of the dorsiflexion assist moment and the initial ankle angle. It is necessary to adjust these characteristics to the individual hemiplegic patient. It was also found that the plantar flexion assist moment generated by conventional AFOs is not necessary, but rather is injurious to the hemiplegic gait. An AFO called DACS AFO, which means Dorsiflexion Assist Controlled by Spring AFO, was developed to incorporate these characteristics. The DACS AFO consists of two plastic parts, the foot and the shank, connected at the ankle joint. An assist device that generates the dorsiflexion assist moment is set at the rear of the shank. When the ankle joint rotates into plantar flexion, a piston compresses the spring and the assist device generates an assist moment proportional to the plantar flexion angle. The dorsiflexion assist moment can be changed easily by using four different kinds of springs. When the ankle joint rotates into dorsiflexion, it rotates freely because a slider in the piston can move without friction. The initial ankle angle can be changed by altering the length of the assist device. The weight of the assist device is 70 grams. The gait of hemiplegic patients with DACS AFOs, with conventional posterior AFOs, and without any AFOs were measured by a three-dimensional motion analysis system (Oxford Metrics VICON 370 and force plates). Before the gait analysis, the characteristics of the DACS AFO were adjusted to the condition of each patient through observation of the gait. The gait cycle decreased and walking velocity increased when patients walked with DACS AFOs. The maximum magnitude of the plantar flexion ankle joint moment increased when the patients walked with DACS AFOs although the DACS AFOs did not generate any plantar flexion assist moment. This result means that the ankle joint moment generated by the plantar flexors increases because of the smooth propulsion of the center of gravity of the body during the mid to late stance. The performance of the DACS AFO was assessed through daily use by approximately 70 hemiplegic patients. The acceptance by patients was good but some defects were clarified, such as weight, appearance, and noise during gait.We intend to improve them and to put the DACS AFO on the market.

View full abstract

The aim of this project is to supply a new, cost-effective stair lift for wheelchair users. For this purpose, a new mechanism, known as "crawltype," has been developed. This system is composed of three parts: traction guides, drive guides, and a hoist carrier. The former two guides are attached to a wall. The traction guide is constructed by a pair of parallel pipes, and the drive guide is set between the traction guides. The hoist carrier is attached to the traction guides by two rollers and can slide on the traction guides. The hoist carrier has one rotating disk driven by an electric motor. On the latter's rim, some free-rotating rollers are mounted, separated by equal degrees along the normal direction of the disk. The disk is set on the hoist carrier at 15 degrees along the line perpendicular to the slope of staircase, parallel to the wall surface of the staircase. The drive guide has been designed for the uniform rotating velocity of the roller attached to the disk to be converted to uniform straight velocity along the traction guide. The resulting drive guide line has a complicated three-dimensional shape. The roller follows on the drive guide by disk rotation. In the case of either straight stairs or spiral stairs, the number of rollers on the disk is four. However, in the case of half-turn stairs, the number of rollers on the disk is five, thereby reducing the number of varieties of the drive guides. This mechanism has many advantages over conventional ones: (1) The traction guide and the drive guide are not mechanical parts, therefore they do not need lubricating oil. This does not soil the clothes of the user on the staircase; (2) drive guides are set closer to the wall, and are set on the inner side from the surface formed by a pair of traction guides. They are composed of cylindrical pipes without sharp edges. This means that they do not hurt children by their indiscreet behavior; (3) the shape of the drive guide is complicated however each drive guide is the same shape. This means that manufacturing cost is inexpensive; and (4) the mechanism is safe. The disk is driven by an electric motor via a pair of worm gears. When the electrical power is cut accidentally, the disk holds its own position. Furthermore, the disk is set at 15 degrees toward the wall, therefore a held disk interferes with the drive guide. By virtue of this structure, there is little danger of the hoist carrier sliding down along the traction guides.

View full abstract

Currently, most of the manipulators used in industry are fixed in a single position, and the work range is limited. From now on, to expand the ability of a manipulator, it is necessary that a manipulator itself will have a mobile ability like that of human beings. A human being performs a variety of work by coordinating hands, arms, and legs. This study aims to clarify basic rules of human motion, from the standpoint of biomechanism, by analyzing the results of an experiment in handling a mainpulator based on a guide shaft with body movements, and to apply them to task planning for a mobile anthropomorphic manipulator. This is a different approach from the conventional method, which uses an evaluation function with weighting coefficient. This paper presents the analysis of human motions between the upper body and legs, and application of the rule to a two-wheeled robot with a 7-degrees-of-freedom manipulator and 1 degree of freedom for rotating the manipulator itself. The experiment was conducted under conditions which require 1500-mm hand movement from three different start positions and angles. From the analysis of the results, three basic rules are derived as follows: (1) Human beings change the direction of the body as the first step in the motion to obtain mobility for walking. (2) Human beings do not change posture in the body coordinate, and mainly conduct the work by the legs. They keep an upper body posture for which the manipulability measure is high, and it is approximately 60% to 80% in the maximum work range of the arm. (3) In a straight line trajectory, human beings keep the same velocity of movement and hand, and keep a track parallel to the given hand trajectory. These basic rules are applied to the task planning algorithm for a mobile anthropomorphic manipulator which needs to determine the path of the two-wheeled robot and the redundancy of the manipulator. From the simulation, the hand is always determined in the range of 60% to 80% in the maximum work range of the arm in which the manipulability measure is high, and the minimum radius path is considered for the wheeled robot. The simulation result shows that it can make a plan which considers the manipulability of the manipulator and the motion of the wheeled robot.

View full abstract

Many groups are researching a biped walking robot, although they have different objectives in mind such as applications of modern control theory, the study of mechanisms, or practical use to medical fields. The authors and others are engaged in studies of biped walking robots, with "human form" as a key word, from two points of view: one is human engineering, and the other is toward the development of anthropomorphic robots. The authors and others have obtained the following results to date. In 1984, the authors and others succeeded in achieving a dynamic biped walking of 1.3 [s/step] by the use of a hydraulic biped walking robot, WL-10 RD (Waseda Leg-No.10 Refined Dynamic). From 1986 to 1994, the authors developed hydraulic biped walking robots of the WL-12 series that compensated for lower limbs moment using an upper body and realized not only fast dynamic biped walking (0.54 [s/step] with a step length 0.3 [m]) but also walking on an unknown surface. In 1995, the authors developed an electrical powered biped walking robot WL-13, in which each leg joint is driven antagonistically via a rotary-type, nonlinear spring mechanism, and realized quasi-dynamic walking (7.68 [s/step] with a 0.1 [m] step length). In the current research concerning a biped walking robot, however, there is no developed example of a life-size biped walking robot which can perform manipulation and locomotion by dynamically coordinating arms and legs. Therefore, the authors proposed the construction of a biped humanoid robot that has a hand-arm system, a head system with visual sensors, and antagonistic driven joints using a rotary-type non-linear spring mechanism, on the basis of WL-13. We designed and built it. In addition, as the first step to realize the dynamically coordinated motion of limbs and trunk, the authors developed a control algorithm and a simulation program that generates the trunk trajectory for a stable biped walking pattern even if the trajectories of upper and lower limbs are arbitrarily set for locomotion and manipulation respectively. Using this preset walking pattern with variable muscle tension references corresponding to swing phase and stance phase, the authors performed walking experiments of dynamic walking forward and backward, dynamic dance and carrying, on a flat level surface (1.28 [s/step] with a 0.15 [m] step length). As a result, the efficiency of our walking control algorithm and robot system was proven. In this paper, the mechanism of WABIAN and its control method are introduced.

View full abstract