The measurement of stiffness on the body surface is of great importance for palpation, percussion and pressation such is used in a detection of an edema in clinics. From the other standpoint of an industrial measurement, it is also significant for getting hold of a soft object such as the hardness sensor of a manipulator. In order to estimate biomechanical properties of the body surface by means of biomechanical impedance, we have developed a measurement system for biomechanical impedance. The system consists of a measuring device, a measuring probe and a personal computer. A vibrator, an impedance head and a load-cell are included in the measuring probe. A random vibration is applied to the body surface through the probe, and force and acceleration at the driving point are detected by the impedance head. The measuring probe is of useful size, and almost all parts of the body surface are measured by the probe. The random vibration method makes a short measurement time possible. The preload (contact force) of probe to the body surface during a measurement is kept constant by means of computer programming. This allowed satisfactory accuracy and measurement repeatability. The mechanical impedance of the body surface is significantly influenced by the measuring conditions (diameter of the vibrating tip of the probe and the preload). It also depends on stratified body structures, such as a bone under the skin surface. The applied vibration is reflected at the layer according to its form and thickness. Consequently there are three typical patterns (soft, intermediate and hard) in the biomechanical impedance spectra. We have proposed a new index SI (Stiffness Index), which is obtained from the mechanical impedance spectrum. The SI has [N/m] dimensionally and indicates a kind of extended spring constant into a visco-elastic medium. The stiffness distribution on the back of the hand is illustrated as a SI mapping. From the mechanical impedance spectrum with a soft pattern, we are able to obtain a visco-elastic constant of the body surface. When the obtained visco-elasticity includes the influences of the stratified structure of the body, it should be termed 'apparent' visco-elasticity. In order to obtain the visco-elasticity of the skin and subcutaneous tissue independent of body structure, we have proposed a correction method for the measured results. This method was applied to the measured result for the chest and the corrected visco-elasticity of the tissue in situ was obtained.

抄録全体を表示

In dentistry, the opposing relations between the upper and lower teeth during contact, which is referred to as occlusion, are evaluated chiefly by the location and/or total area of their contact portions. Considering the amount of force applied to the teeth during chewing, however, the proximity between the opposing surfaces as a whole should also be evaluated. In order to evaluate the relations of opposing teeth, including both contact and clearance, the distance transformation was applied, in which any surface point was mapped into the distance between that point and its opposing surface. That distance and its distribution on the surface are referred to as the interocclusal distance and the distance map, respectively. This map provides a closed surface-region where the interocclusal distance is less than or equal to xmm, denoted as R(x). This region can be interpreted as a generalization of contact region R(0) by the interocclusal distance. The proximity can be evaluated by the area of R(x), denoted as s(x)mm^2, which decreases monotonously as the two surfaces approach each other. To verify the validity of s(x) for the occlusal analysis, it was applied to that during lateral excursions. The occlusal surfaces of the upper and lower first molars were digitized by a laser scanner, and the jaw movements were measured in 6 degrees of freedom. The combination of these measurements provides any positional relation between the opposing surfaces during lateral excursions. Proximity s(x) of the upper first molar was computed at various occlusal positions, where the incisal point was less than or equal to 2.0mm from its location in the intercuspal position (ICP). Proximity s(x) relative to that in the ICP, referred to as the relative proximity, was compared on the working and non-working sides. For s(0.3), the difference between the two sides was not significant statistically on the relative proximity in all the lateral positions, suggesting that the portions relatively close to the opposing surface remained even on the non-working side. For s(2.0), however, the relative proximity on the working side was on average larger than that on the non-working side, as was expected. These results revealed that proximity {s(x)|x≧0} considerably reflected the opposing relation of the two surfaces as a whole.

抄録全体を表示

Soft tissues of the temporomandibular joint (TMJ) influence the position and the movement of the condyles. The examination of the laxity of the TMJ provides valuable diagnostic findings concerning the function and pathological changes in the soft tissues of the TMJ. There have, however, been few studies of biomechanical properties of the soft tissues of the TMJ. The purpose of this study was to non-invasively estimate the mechanical properties of the TMJ regions. The viscoelastic properties of the TMJ regions were analyzed by an automatic system which was composed of a measuring device for the mechanical mobility of the TMJ regions and a data analysis unit with a micro-computer. First, the soft tissues of the TMJ at rest position were randomly vibrated through the lower central incisor by a portable vibrator with a small impedance head. During this measurement, a standardized preload 0.49N (50gf) was applied to the tooth. The impedance head generated the output signals which were proportional to input acceleration and force response. They were amplified by charge amplifiers and sampled by an A/D converter inserted in the micro-computer. The mechanical mobility spectrum which gave information about viscoelasticity of the lower tooth and the TMJ regions was obtained. Next, the subjects were instructed to bite resin blocks, and the mobility spectrum which gave the viscoelasticity of the lower tooth was obtained. Then we abtained the mechanical mobility of the TMJ regions by subtracting the second data from the first data. We proposed a mechanical model of the TMJ regions. Furthermore, biomechanical parameters (damper: c, spring: k and mass: m) were analyzed from curve fitting using a personal computer. The viscoelastic parameters for the TMJ regions at rest position were taken of 28 normal subjects (14 men and 14 women). Statistical testing of the data employed a test for the difference of means and correlation coefficient. In normal subjects, the mechanical properties in men were significantly greater than those in women (p<0.05). We applied this system to temporomandibular disorder (TMD) patients (2 men and 2 women). Difference was found for three parameters between normal subjects and TMD patients. The mechanical properties of the TMD patients were out of the range of those of normal subjects. The coefficient of variation in the mechanical properties of TMD patients varied more than those in normal subjects both over one day and over one week. In the next stage of our resarch, we improved this system in order to be able to measure parameters by using chin vibration. During measurement, the preload on the chin was kept at 1.96N (200gf). Parameters of the TMJ regions were taken for 24 normal subjects (12 men and 12 women). The mechanical properties in men were significantly greater than those in women (p<0.01). Our method enabled us to estimate the viscoelastic parameters of the TMJ regions noninvasively.

抄録全体を表示

The skeleton of an animal must have enough robusticity to support its body. In this study scaling of limb bones in mammals was examined from the viewpoint of the strength of materials. Mammals who locomote and support their bodies with their four limbs were compared. Allometric equations were calculated from the osteological measurements and the cross-sectional geometry on the long limb bones of quadrupedal mammals. Osteometrical data were consisted of 79 species, 36 families, and 13 orders ranging from the shrew (5.6g of body mass) to the elephant (5000kg). The cross-sectional geometry was measured at the mid-length of the femur and the humerus on 18 species, 14 families, and 7 orders. Allometric equations were obtained from the data by least-squares regression, linear dimensions against the body mass, and by standard major axis analysis between diameters and bone length after transformations to logarithms. Thirty-eight correlations for Tables 1, 2, and 3 were all statistically significant. Most exponents of allometric equations for outside linear dimensions of bones against the body mass were significantly near the 1/3 required for geometric similarity. Exponents for bone lengths did not in the least agree with the hypothesis of elastic similarity by McMahon (1973). The robusticity of mammal long bones is not known only from the outside, because the marrow cavity diameter is not geometrically similar to the outside diameter. The inside wall of a hollow bone grows and decreases relatively independently of the outside wall. The cross-sectional geometry is a good indicator for the robusticity of bone. Exponents for cross-sectional geometry statistically corresponded to the robusticity against the bending or torsional moments, which may be caused mainly by muscle forces. The axial force produced by only the body weight did not agree with the exponent data on the geometry of shaft cross sections or on the sizes of epiphyses. The shapes and sizes of mammal long bones were considered to be adapted mainly to the moments caused by muscle forces during daily life activity.

抄録全体を表示

The hardness of a mattress is one of the main influential factors in the comfort of bedding, because it determines vertebral loads and pressure distribution on the tissues. In this paper, we analyzed the comfort condition of a mattress using a developed experimental bed with mechanisms for adjusting the hardness and a link model calculating internal forces such as vertebral loads and compressions. The mattress is divided in the longitudinal direction into 16 parts: the 8 parts for the torso are 100mm wide; the others, 150mm. The hardness of each section can be set arbitrarily from 0.35N/mm to 46N/mm. Reaction forces relating to surface pressure and the downward motions affecting the vertebral loads are measured. Twenty-five adult male subjects tried to adjust their mattress to be comfortable. Fat subjects and subjects sensitive to lumbar pressure preferred a uniformly hard mattress or partially hard mattress at the hip part in order to maintain a flat posture. Thin subjects selected soft mattress to decrease overall reaction forces. These results show that the internal loads at the lumbar parts and the pressure on the back tissues are important factors in achieving a comfortable fit. To verify the adequacy of the assumed determinative factors, a two-dimensional rigid link model of the human body was developed to calculate internal loads in the recumbent position. The model contains 21 segments: a head, 17 divided torso sections, thigh, shank, and foot. Elastic elements were attached as a vertebral disc or a ligament between spinal joints and as the tissues at the back body surface. The sinking posture depending on the input hardness distribution was calculated by the Newton-Raphson method. The calculated posture almost agreed with the measured one. Using this simulation, we confirmed that the internal load become minimal at the selected distribution of hardness. Inversely, the distribution achieving minimum internal loads was also calculated for several body conditions. Consequently, the concentrated distribution of hardness on the hip part satisfied the biomechanical comfort requirements.

抄録全体を表示

This study was designed to analyze the jumping forehand smash of elite players based on three-dimensional kinematic data, and to gain insight into the basic badminton smash technique. Jumping smashes of four male elite players were filmed with two high-speed cameras operating at 250Hz with exposure times of 1/1500 and 1/1250 s. Nine jumping smashes were selected for the analysis, and were digitized from the take-off of the jump to the end of the swing in the air. Thirty three-dimensional coordinates for the segment endpoints and racket were computed by a Direct Linear Transformation Method. Small reference poles were fixed to the forearms of the swing arms of the subjects to detect the movements of the radio-ulnar joint and wrist joint. The following six joint angle changes were obtained throughout the smash motion. (a) abduction/adduction angle at the shoulder joint; (b) internal rotation/external rotation angle at the shoulder joint; (c) flexion/extension angle at the elbow joint; (d) pronation/supination angle at the radio-ulnar joint; (e) radial flexion/ulnar flexion angle at the wrist joint; (f) palmar flexion/dorsiflexion angle at the wrist joint. The results showed that internal rotation of the shoulder joint, extension of the elbow joint, and pronation of the radio-ulnar joint seemed to contribute to produce great velocities of the racket head, because the three rotations occurred over the greatest range in the shortest time in the six rotations immediately before contact with the shuttle. Preliminary to the three motions were motions in the opposite direction: external rotation of the shoulder, flexion of the elbow, and supination of the forearm were detected. These motions in the opposite direction would be useful to extend the range of the motion in each joint angle. The results also appeared to be related to intrinsic muscle properties, that greater power can be exerted by the stretching-shortening cycle of the muscles. The time of the last joint rotation starting immediately before contact was in the order of decreasing inertia. Times required for the rotation until contact became shorter in order of occurrence. The faster the rotation was, the later it occurred, and the shorter was the rotation time. This kind of chain and continuous movement of different joints, and different freedom with the same joint, may accelerate the racket head efficiently. The averages of elbow angle and racket angle (the angle between the forearm and racket shaft) were 160.0°and 147.0°at contact, respectively the values may be the suitable choices between the contact height and contact speed in practical play.

抄録全体を表示

In an aging society, it is important to study the states and mechanisms of decline in body functions with aging and of changes in motion with aging. It is especially valuable to investigate the functions of the lower extremities, since they are main factors determining safety as well as competitiveness in almost all sports activities. The purpose of this study is to cross-sectionally investigate the changes in joint moment and power of lower extremities for the elderly during a standing long jump using a force platform and a high-speed VTR camera. The subjects of this study were thirty-three healthy elderly males (62yr. to 86yr.). They were divided into three age groups: the 60s(n=12), 70s(n=14), and 80s(n=7). They were videotaped at 200 frames per second with high-speed VTR camera while they were performing the standing long jump from a force platform with maximal effort. Using ground reaction force data from the force platform and coordinate data of seven body landmarks (toe, ball of the foot, heel, ankle, knee, great trochanter, and suprasternale) from VTR images, equations of motion for thigh, shank, and foot were solved, and then joint moments, powers, and mechanical works were calculated. Joint moment curves and power curves were evaluated after being normalized by time from the start of the jump (ST) to the takeoff (OFF) and averaged for each group. For the hip and knee joint extensor's moment acted predominantly from ST to the point of about 90% completion of the jump, and then flexor's moment acted predominantly up to the OFF. For the ankle joint, extensor's moment acted predominantly through-out the jump. The power curve for the hip and knee has three peaks. The first one was power by extensor's eccentric contraction (EEh, EEk), the second was power by extensor's concentric contraction (ECh, ECk), and the third was power by flexor's eccentric contraction (FEh, FEk). The power curve for the ankle has only two peaks (EEa, ECa). From the investigation of these curves and these maximum values, the following results were obtained. 1) Although the magnitude of downward kicking force in a standing long jump hardly changed with aging, magnitude of backward kicking force decreased with aging, and jumping distance for the elderly also decreased proportionately. 2) Maximum extensor's moment and maximum power of joints in lower extremities decreased with aging. The magnitude of decrease was the largest in the hip joint, the second largest in the ankle joint, and smallest in the knee joint. 3) The power by concentric contraction of the extensors in the standing long jump was influenced by aging and decreased. The power by eccentric contraction of the extensors, however, was hardly influenced by aging. 4) In a standing long jump, which requires that the body be largely inclined at the takeoff, elderly people tended to jump with smaller forward incline of their bodies to avoid being in an unstable situation.

抄録全体を表示

The purposes of this study were to identify the changes in moments, powers, mechanical works, and contributions of the lower limb joints in vertical jump takeoff and landing from vertical fall, and to propose an interpretation of recruitment of the body segments in sports skills. Fifteen male track and field athletes as subjects performed vertical jumps with no arm swing in six different levels of effort: 100% (maximum effort), 90, 80, 60, 45, and 30%. Eleven subjects landed from the five different heights of 0.25, 0.4, 0.6, 0.8, and 1.0m so as to minimize the landing impact force as much as possible. Five reflective landmarkers on the lower extremity were automatically detected at 60Hz with Quick-Mag System I, and interpolated using cubic spline function to convert data equivalent to sampling rate of 100Hz. Ground reaction force data were sampled at 1000Hz with a Kistler force platform. Two-dimensional inverse dynamics were performed on a four-link segment model to obtain the moments and powers of the hip, knee, and ankle. Mechanical works and contributions of the lower limb joints were calculated from these data. The results were summarized as follows. 1. With changes in relative jumping height the maximum joint moment and power on the hip increased remarkably, while those of the knee and ankle did not show such a sharp increase. Relative contributions of the ankle, knee, and hip to the total positive mechanical work changed from 37.6, 40.4, and 22.0% for 40% jumping height to 29.4, 29.0, and 41.6% for maximum jump, respectively. 2. With increases in heights of fall, the maximum joint moments and negative powers increased in all three joints. The hip showed the maximum muscle moments, followed by the knee and ankle. The knee joint exerted the greatest negative power at every height to dissipate the energy of the fall. Relative contributions of the ankle, knee, and hip were 19.6, 51.8, and 28.6% for height of 0.25m and 14.8, 48.8, and 36.4% for height of 1.0m. Considering these results together with the inertia properties and mechanical work capacities of the lower limb muscles, it would be speculated that the smaller segments must first be recruited so that energy and load on the body can be minimized, but in the case of the maximum power output, the greater segments with large inertia and mechanical work capacities may join the preceding segments, where achieving the objectives of the skills may take priority over the principles of minimizing energy and muscle stress.

抄録全体を表示

There has been a great deal of research on human walking on level ground because level walking is the most fundamental transportation method for us. However, we frequently encounter stairs and slopes during the course of daily activities. Stair-climbing and slope-walking require much more energy than level walking, so biomechanical analysis might be needed especially for elderly people and the physical-handicapped. But few reports on the stair-climbing and slope-walking have been published. The purpose of the present study was to analyze the energy aspect of stair-climbing and slope-walking. The energy consumption during stair-climbing and slope-walking was calculated from gait data using a mathematical model and compared with that calculated from O_2 consumption data. The method of gait measurement and the mathematical model have already been described by the authors. The values were in good agreement. The positive and negative work involved in level, stair, and slope walking were calculated. The amounts of positive and negative work were equal during level walking. Positive work overcame negative during ascension of stairs or slope, and negative work overcame positive during descent. More energy was consumed going up than during level walking because it required more positive work, and more energy was consumed going down than during level walking because it required more negative work. The energy per unit distance was minimized by adequate stride length and cadence during ascent of a slope, but the energy expended per unit of time was very large, so this was not considered an optimum condition. The condition at which the energy per unit time equals that of the optimum condition of level walking might be the most adequate condition during ascent of a slope. These results should be reflected in the design of stairs.

抄録全体を表示

It is known that nonhuman primates differ from most nonprimate mammals in the predominant use of the forward cross type locomotion, as opposed to the backward cross type, while walking. In this study, we investigated the development of the sequence of movenent of the four limbs, both on and off the ground, and the effect of weight load on it. We also tried to account for the sequence of movenent of the four limbs in human quadrupedal walking. Through the comparative study of animal and human quadrupedal walking, we clarified the functions of fore and hind limbs during quadrupedal walking, and the evolution of locomotion. The subjects were 15 infants under one year old, 90 children aged from 1 to 5 years, and 20 adults. We observed two kinds of quadrupedal locomotion: "creeping on hands and knees" and "walking on hands and feet". The children over 3 years of age and the adults were requested to perform these two kinds of locomotion on a downward slope (15°), which would necessitate the support of the body by the forelimbs more than on level ground. Body weight was measured on each of the fore and hind limbs while the subjects assumed various postures. Cinematograms were recorded on VTR. The sequence of movement of the four limbs was determined after repeated viewings of the VTR. We found that on a flat surface the bb-pattern (the backward cross type of limb movement that takes place both on and off the ground) was obvious in the creeping of infants and in the walking of subjects of all ages. The fb-pattern (the sequence of limb movement characterized by the forward cross type of locomotion off the ground and the backward cross type on the ground) was also observed in the creeping of children and adults. In walking, the body weight fell more heavily on the forelimbs than on the hindlimbs in static standing posture, especially in the creeping of children and adults. On the slope, twice as many children showed the bb-pattern while creeping than displayed it on the level ground. From these results, the following can be concluded concerning the movement of the four limbs in human beings. Humans, by nature, possess the motor program for the backward cross type of locomotion. In the pre-bipedal period, fore and hind limbs seem to have the same function. However, after humans master bipedal walking, the function of limbs in creeping may differentiate: forelimbs function to orient, and hindlimbs are for propulsion. However, this functional division between the fore and hind limbs is not apparent in those subjects walking on hands and feet after the achievement of bipedal walking. There may be an evolutionary continuum between walking on hands and feet in human beings and the quadrupedal walking of nonprimate mammals.

抄録全体を表示

Questionnaire surveys and motor performance tests were performed on 37 female students in the Department of Dance and Physical Education, Ochanomizu University, to investigate lateral differences in dance movements. Some typical movements employed for the classical ballet-balance (Developpe), jump (Grant jete) and turns (Pirouettes en dehors, Fouette en tournant, Tour piques)-were examined. The questionnaire survey showed that students tend to prefer the left lower limb as the supporting leg for each movement. Especially pirouettes en dehors were studied in detail by means of 3-dimensional graphical analysis (Ariel Performance Analysis System) of 11 students who had been trained for classical ballet. They were divided into advanced or beginning groups according to their skillfulness. The ratio of time required for each phase of pirouettes en dehors to the whole motion was analyzed with-the aid of the 3-dimensional graphics. Single rotations were broken down into 6 phases and double serial rotations into 8 phases. The advanced group showed large lateral differences: more for double serial rotations than for single rotations. The beginning group showed the opposite tendency. Large differences were found at the end of the rotations in both groups. Precessional rotation-which unavoidably accompanies pirouettes en dehors-was analyzed for each group with stick figures. Large lateral differences indicated in the double serial rotations for the advanced group could neither be recognized in the double serial rotations nor in the single rotations in the case of the beginning group. Large differences in the displacement of the centre of body mass are shown in the beginning group based on analysis of every movement. Larger lateral differences were shown in the double serial rotations than in single rotations in the advanced group. In the case of the advanced and begining group, the changes in acceleration of the centre of body mass indicate large lateral differences at the beginning and the end of rotations. It seems to be difficult for both groups to control the long axis during rotations, especially in the two phases when the swinging leg leaves the floor and touches it again. From the conscious and functional point of view, lateral differences in dance movements indicated between the advanced and beginning groups are reasonably explained from their level of skill.

抄録全体を表示

Many investigations have been performed to analyze postural sway. However, few investigators have analyzed the Center of Foot Pressure (CFP) of the individual foot (ICFP). In this paper, our special interest is in the ICFP. In the first study, 85 non-handicapped subjects 20 to 60 years of age took part. They were asked to stand with eyes open and feet parallel, 0cm apart. The relations between foot length and the calculated items were analyzed. There were three different groups with respect to the correlation coefficient and its statistical significance. Strong correlations were determined in mean values of CFP and ICFP to some extent, and weak correlation were determined in those SDs and ranges. No statistical significance was found in the trajectory length. No strong correlation was found with respect to age. In the second study the principal component analysis (PCA) using ICFP was employed to analyze standing sway in both non-handicapped subjects and patients with lower extremity dysfunctions in an integrated manner. In addition to the 85 healthy subjects, 69 patients participated in this study. So a total of 154 data sets were analyzed. Standing posture was identical to that in the previous study. Three principal components (PCOMs) were found to be valuable: magnitude of postural sway, symmetry, and dependence on the sound side. Even though there was no PCOM which could differentiate data with respect to diseases, the three PCOMs clearly illustrated subjects' physical condition in regard to lower extremity functions related to standing ability. The analytical method is expected to be applicable as an auxiliary diagnosis so that any information on conditions which are not apparent from clinical observations can be revealed. We have to find a subjective method to analyze postural sway in each patient because medical rehabilitation procedure must be applied to each individually. The method employed in this research can provide basic knowledge for such purpose.

抄録全体を表示

Simulations of human walking have been actively performed in recent years. Most, of however, treat only normal walking. The abnormal walking is analyzed mainly by experiments due to dificulties associated with modeling abnormalities, but the simulation approach is expected to give much information about patients who have similar conditions. This paper presents a model which can simulate normal and abnormal walking. Our model has a Zero Moment Point, abbreviated as ZMP, at a joint of one leg out of the ankle, knee, and hip joints. The idea of the ZMP, in which the fact that moments about a point of application of floor reaction force are zero is utilized, is effective for the simulation of the human walking and expanded further in our model. The joint having a ZMP can transmit force to adjacent connected bodies but cannot generate the moment during the stance phase to control the motions of the bodies. The joint is called the Zero Moment Joint, abbreviated ZMJ. Based on the ZMJ, three modified models are also proposed: a Passive Element Joint (PEJ), which has a spring at the joint; a Partial Moment Joint (PMJ), which can produce a partial amount of the moment in about the joint in normal walking; and a Constrained Range Joint (CRJ), the angle of which stays within some constrained relative angle. Considering that many experimental results on normal walking have been obtained but of abnormal walking are very few, differences in characteristics between normal and abnormal walking simulated for the same walking conditions are evaluated and discussed as compensatory actions at the remaining sound joints. The actions became greater when the diseased joint is located in a higher position. The locus of the ZMP exceeded the supporting area when the ZMJ was at the knee or the hip. The walking pattern of the infants or the aged was effective to reduce the actions. Other joint models can also be used to reduce the actions. To use a cane on the opposite side to the abnormal leg was effective in making the posture of the upper torso to be upright. The simulation results suggested that what the clinical side do a patient having the lost function can walk in ways resembling the normal walking pattern with acceptable compensatory actions at the sound joints. Such an approach will be promising for the control of autonomous bipedal robots.

抄録全体を表示

Previous studies showed that three reflex EMG responses were recorded from the stretched muscle, when the wrist flexors were stretched by sudden angular displacements of the wrist joint in the extension direction. The first response, with short latency is called the M1 component and is thought to be a monosynaptic spinal reflex. The second and third responses, with long latency, are labelled M2 and M3 components and are assumed to be transcortical reflexes. It has been reported, however, that the M3 component is not reliable in occurrence. The amplitudes of reflex components are known to be modified by voluntary movement. Premotor time (PMT) in the reaction time movement is not studied in relation to aspects of the reflex EMG responses including long latency components. The present study was performed in order to investigate whether the PMT under conditions of stretch stimulus (SS-PMT) differs from that under conditions of light stimulus (LS-PMT) and touch stimulus (TS-PMT), and how the PMTs are related to the aspects of the appearance of the reflex EMG responses. Nineteen healthy males, ranging in age from 21 to 28, participated in this study. The DC torque motor was used to evoke the reflex EMG activities of the wrist flexors. The results are summarized as follows: 1) In all subjects SS-PMT was significantly shorter than TS- and LS-PMT. The results suggest that input from the muscle spindle to cerebral cortex shortened the PMT. 2) The short and long latency reflex components appeared preceding the voluntary EMG burst. The latency of M1, M2, and M3 was 19.1ms, 49.2ms, and 71.7ms, respectively. 3) The subjects were classified into three groups (A, B, C) according to the presence or absence of reflex components (M1, M2, M3) and these EMG amplitudes. SS-, TS- and LS-PMTs in group A were significantly shorter than the PMTs in group B, and each PMT in group B was significantly shorter than in the group C. 4) In group A the amplitude of M2 was significantly higher than the M1 amplitude. The M3 component was not identified, since the M2 component was followed by voluntary EMG burst. In group B the M2 amplitude was significantly higher than the M1, and the M3 amplitude appearing before the voluntary EMG burst was significantly smaller than the M2 amplitude. In group C the amplitudes of M1, M2, and M3 were larger in that order. These results suggest that the PMT is related to the presence or absence of the long latency components and those amplitudes.

抄録全体を表示

The central nervous mechanisms by which sensory input becomes associated with motor output can be termed sensory-motor integration, and this plays an important role in the adaptive behavior of animals. We have attempted to describe the neural mechanisms of sensory-motor integration by which behaviorally relevant sensory input is translated to the motor activity necessary for the execution of a movement. We have investigated the neuronal basis of this translation process in part of a neuronal network for linking sensory input to motor output, i. e., S 1, area 5, area 6, and area 4. Single neuron activity was recorded in these four cortical areas during the performance of a task in which a monkey received a brief vibratory stimulus to either the fingers or the thumb and had to respond with an isometric force change with the stimulated digit/s. Data were also recorded when the monkey did not respond to the cue (no response trials). Quantitative methods of data analysis were used to characterize the temporal organization of changes in neuronal activity in the different cortical areas. A method of analysis was developed to estimate the timing of the different peaks in activity of each neuron using an estimation of the probability of spike occurrence and a non-linear template matching procedure. The variance ratio of two time periods-onset of stimulus to a peak in activity and peak activity to the onset of movement-was used as an index of the functional significance of the peak in activity on a sensory to motor gradient. When the behavioral response did not follow presentation of the tactile cue, there was still a change in activity of S 1 neurons, whereas the pattern of change in activity in area 5 neurons and area 6 neurons in no-response trials was modified compared to the pattern in response trials. There was no change in activity for area 4 neurons during no-response trials. The distribution of timing of peak activity in the different cortical areas showed that there was an overall sequential order. Initial peaks in activity were observed first in S 1 and area 5 neurons, then in the activity of area 6 neurons. Around the onset of the behavioral response, there were peak activities in area 4 neurons and a second peak in the activity of area 6 neurons. After the onset of the behavioral response, again, there were peaks in the activity of area 5 and S 1 neurons. Although there was a large overlap in the distribution of variance ratio for the different cortical areas, many of the initial peaks of activity in S 1, area 5 and 6 neurons were functionally interpreted as being "sensory" where-as the peak activity in area 4 and the second peaks in activity of area 6 neurons were interpreted as being "motor". On the basis of the variance ratio data, the activity of some area 6 neurons appeared to be functionally between these "sensory" and "motor" groups. There were slight differences in the timing and variance ratio of the peaks in activity of area 5 and S 1 neurons that occurred after the behavioral response. This late activity in area 5 may have been due to an efferent copy from area 6, whereas that in S 1 was most likely due to sensory feedback from the periphery evoked by the behavioral response. Based on the results of a combined qualitative and quantitative analysis, we propose that area 5 and area 6 integrate sensory information and motor information in the cerebral cortex during the performance of sensory-motor tasks.

抄録全体を表示

The purpose of this paper is to summarize our recent work concerning development of the capacity to generate coordinated isometric finger forces used for lifting and holding a small object using a precision grip in human children. One hundred and eight children aged from 10 months to 13 years and 15 adults participated in the study. Using a force transducer attached grip apparatus (wt., 200g or 400g; surface, suede), grip force and lifting force acting on the object and vertical position of the object were measured. Two separate experiments were carried out in this study. In the first experiment, the subjects (93 children and 15 adults) lifted the apparatus weighing 200g 5 times. In the second experiment, the subjects (79 children and 15 adults) lifted the apparatus weighing either 200g or 400g 20-25 times. The weight was changed in a random manner without informing the subjects. In the first experiment, it was found that younger children (aged below 5 years) were slower than adults in sequencing between touching and gripping actions and between lifting and holding actions. The grip force for these children was greater than that in adults. A higher between-trial variability was also noted in force parameters for these young children than for the adults. A general trend in development of controlling the timing and force application during griplift actions showed that 0-1 year old children could not adequately program temporal and spatial components of the isometric grip and lifting forces in relation to the object weight. The capacity progressed rapidly up to 5 years of age, and progressed slowly from that age on. These results suggest that to acquire well-developed force control capacity in precision grip, as seen in adults, nearly 10 years of learning as well as development of neural and motor systems are required. Poor capacity in young children is believed to be due to immature functioning of cortico-spinal descending pathway, an inhibitory input to the motoneuron pool activating the finger muscles, and of the central nervous system responsible for a feedforward strategy of motor action.

抄録全体を表示

The purpose of the present study was to make it clear how the mechanical parameters (viscous and elastic coefficients) of muscle change with the activation level (contractile force) and to examine whether the estimated parameters were valid or not, by comparing them with previously reported values. Experiments were made from flexor pollicis longus muscle in five healthy subjects. The left hand was placed on a horizontal plate, and the subject was asked to maintain constant isometric force by watching the developed force displayed on a CRT. When the muscle force was kept constant, the thumb was ramp-stretched or shortened by a small amount (approximately 4.5 degree) by means of a servo-controlled motor. Time courses of the force (torque) and the length (angle) observed during the period ranging from the onset of length perturbation to 50ms were used for estimating the mechanical parameters. It was showed from examining EMG activities that the contractile force during this period was almost the same as the constant isometric force maintained before the length perturbation. Individual parameters were estimated by searching the best fit of the model response with the measured force and by analyzing the sensitivity of estimated values. The obtained results are summarized below. 1) Both the series and the parallel elastic coefficients increased almost linearly with the contractile force (isometric force). 2) The viscous coefficient could be estimated only at the range of slow velocity, while dynamic constants of the force-velocity relation (a and b) could not be obtained. The viscous coefficient increased linearly with the contractile force. 3) Linear increases of these mechanical parameters with the contractile force were shown in both stretching and shortening of the muscle. 4) Obvious differences were found between estimated values of each mechanical parameter obtained from the five subjects; that is, the mechanical property of the muscle was subject-dependent. 5) Meanings of the simultaneous modulation of the muscle force and the muscular activity in regulating limb movement and mechanical impedance are discussed from the control point of view.

抄録全体を表示

A two axis flexible electrogoniometer made of electroconductive rubber has been developed. The goniometer is highly flexible and does not break even if deeply deformed. Four pieces of electro conductive rubber string are plastered on a silicone rubber column of squaresection, and both ends are molded by hard plastics. A slide mechanism is constructed on one side of each end to prevent excessive deformation of the measuring part. Each pair of the two opposite sides of electroconductive rubber construct a half Wheat-stone bridge circuit to detect only the angle made by both ends of the goniometer. Two different sizes are made. The section of the measuring part is 4mm×4mm or 3mm×3mm, and its length is 150mm or 100mm. The weight of the sensing part is 10g or 5g, and the weight of the amplifier is 170g. The method of making the measuring part using a specially designed jig is described. The goniometer is calibrated to set correct bending angle separately in the directions perpendicular to each other from -90°to 90°. The nonlineality, hysteresis, and interference between the two directions are very small, and there is no necessity to compensate for clinical use. This goniometer is applied on the human wrist joint and ankle joint of a normal subject. The result is displayed in real time as the relative motion of the distal segment against the proximal segment in three-dimensional space on the CRT display. The minute phenomena of the motion can be observed, and the range of motion in the direction of dorsal, palmar, radial, and ulnar flexion of the wrist and planter, dorsal, invert, and evert flexion of the ankle can be easily grasped quantitatively.

抄録全体を表示

Human motions during steady-state gait have been extensively investigated and their average characteristics are well understood. In recent years, studies of transient gaits, such as initiation and stopping or turning during steady gait, have attracted the interest of some researchers. But most of these studies have been done with level walking. Going up and down stairs is a common activity of daily living. From the mechanical viewpoint, stair gait is quite different from level walking, because the vertical motion of the body is added. A study of the characteristics of the ground reaction forces during stair gait is an important step toward the understanding of the motion of the stair gait. In this study, the characteristics of both the reactions and the points of application during ascending and descending stairs were investigated, and the differences between them are discussed. The initiation and stopping of the stair gait were investigated also. These results are compared with the data on level walking. In experiments, two large force plates and two staircase models composed of five steps were used. Each model had standard dimensions in height and depth for an indoor staircase, and both were set on the large force plate. The results of this study showed many interesting characteristics. In the steady stair gait, the peak-to-peak transition in the longitudinal reaction force diagram was not monotonous, as it is in level walking. The value of the peak in the vertical force diagram was generally larger than that for level walking. It was produced in the double support phase in going up stairs, and in the single support phase in going down stairs. The point of application in the longitudinal direction moved backward once in the single support phase. The reason for this is that the first contact with floor is not the heel but the toe. In the initiation and stopping of the stair gait, the most remarkable characteristics was that the magnitude of the peak of the vertical force component during the first single support phase in initiation differed from that of the last single support phase in stopping. This difference was closely related to the vertical driving and breaking forces, which were produced at the time of initiation and stopping, respectively.

抄録全体を表示

The functional role of the moments acting at the Metatarsophalangeal joints of the foot (MP moment) during biped human locomotion is discussed in relation to the time sequence of the foot contact events of both feet and to the controllability of the angular momentum of the whole body. Theoretical consideration of the foot contact sequence indicated that the MP moment is critical in controlling the angular momentum of the whole body during the period between heel-off of the ipsilateral foot and heel-contact of the contralateral foot, so that the body does not fall down and excessive loading on the forelimb is prevented. The MP moment was measured in normal bare-footed level walking for two adult subjects both in steady state walking and at initiation of walking, each under slow, normal, and fast conditions. The results showed that the MP moment is as large as one fifth to one third of the maximum plantar flexion ankle joint moment, and increased with an increase in walking pace. Thus, it was demonstrated that the MP moment plays an important role in controlling the angular momentum of the whole body in the late stage of the single support and the early stage of the double support, and cannot be neglected.

抄録全体を表示

An experimental ankle-foot orthosis in which the flexibility and initial angle of the ankle joint can be changed arbitrarily was made in order to assess the effect of mechanical property of ankle-foot orthoses on hemiplegic gait. The flexibility of the experimental ankle-foot orthoses was adjusted by changing the elasticity of the spring, and the initial angle was adjusted by changing the length of the spring. The gaits of 15 hemiplegic patients were measured continuously using the experimental ankle-foot orthosis. Temporal factors, the ankle and knee joint angles, and the ankle joint moment were calculated, and the data were analyzed statistically. The following results were obtained. 1) Drop foot during the swing phase was prevented by a more elastic spring. 2) A less elastic spring caused an increase of the dorsiflexion ankle joint moment and a decrease of the plantar flexion angle of the ankle joint at foot contact. 3) Adjustments in dorsiflexion moment by the ankle-foot orthosis that were adequate for the individual patient's disability at foot contact improved the knee joint angle throughout the gait cycle and for all temporal factors. The gaits of 6 hemiplegic patients using the experimental ankle-foot orthosis were measured by force plates and a position-sensitive semiconductor camera system. Results showed that the flexibility and the initial angle of the ankle-foot orthosis affected the knee joint moment. These results are useful in obtaining the criteria to design ankle-foot orthoses which are appropriate for individual hemiplegic patients.

抄録全体を表示

This paper reports on a virtual training environment developed using virtual reality technology with force and visual image feedback capability. In our system: (1) A light and safe force-display implemented using a pneumatic rubber actuator is available. It has a wide range of motion and large driving force comparable to those of human joints and muscle. (2) The trainee's muscle characteristics can be measured. (3) The parameters of the training environment (e. g. spring constant, weight of dumbbell) can be changed easily so as to obtain an environment best fitting the characteristics of human muscle. In order to accomplish the above purposes, a system must be capable of freely generating and controlling the physical and psychological elements of a training environment. For the first step, applying virtual reality techniques, we are developing a computer controlled training system which can generate and control various audio/visual images and forces to be applied to the trainee. Currently, however, we have decided to limit the scope of implementation to the upper extremity as the training target, and to visual images for the environmental information. The trainee using this system wears a force-display which can apply force to his/her upper extremity and a head mount display through which he/she can see the virtual world, a room with wall, windows, etc., in which a spring and a dumbbell are placed in the room. The trainee can "use" these sporting goods and can feel forces on his/her upper extremity as if he/she were actually exercising using them. By measuring the trainee's muscle characteristics and setting them in the system's computer before starting a training session, an improved training environment results. In addition, in the case of rehabilitation, the system can provide information such as video images of rehabilitation history data, which can help increase the trainee's motivation for attending the exercise. In our system, an important role is played by the actuators which are attached to the force-display to generate various reaction forces. An actuator serving those purposes should be safe, small, light, and capable of high force output. As human muscle of the upper extremity is much stronger than the muscle of the fingers, an actuator with high output is very desirable. It is also important that the apparatus not feel unpleasant to the trainee when he/she wears the force-display. For these reasons, we have chosen a pneumatically controlled rubber actuator.

抄録全体を表示

For the purpose of the restoration of hands-free paraplegic standing using functional electrical stimulation (FES) and an artificial feedback postural control system, electromyogram (EMG) was analyzed in normal subjects during standing, and postural sway from foot pressure was estimated. In the EMG study, activities of 24 muscles, 22 on the left and two on the right, were simultaneously recorded during eight motions: forward, backward, right-leaning, and left-leaning and their reverse motion. The t-test, results showed that the activity of the muscles of the posterior shank, hamstrings, erector spinae, and quadratus lumborum increased in forward leaning. In contrast, the activity of the gluteus maximus did not show an obvious increase. In backward leaning, the activities of the tibialis anterior and quadriceps femoris increased. In coronal plane motion, the gluteus medius, tensor fascia latae, and shank muscles of the weight-bearing leg and the quadratus lumborum of the free leg increased their activities. Four thin film pressure sensors using conductive rubber were put on the big toe, metatarsal head I, metatarsal head V, and heel. The pressure of the heel decreased, and the pressures of other points increased in forward leaning for a normal subject. For paraplegic standing between parallel bars using FES, no sensor output change was observed with leaning because there was less ankle joint stiffness. Therefore, a plastic ankle-foot orthosis was applied to increase the ankle joint stiffness. It was found that postural sway can be estimated by the measurement of foot pressures where there is sufficient ankle joint stiffness. It is expected that these findings will lead to the use of FES to facilitate practical hands-free standing.

抄録全体を表示

The propose of this study was to develop a new type of hand prosthesis which simulated dynamic properties of the neuromuscular control system of the human hand. This hand consists of a mechanical hand (one degree of freedom), a processing unit for surface electromyographic (EMG) signals, and a digital servo system of the DC motor. EMG signal was used as a control signal for opening and closing the hand prosthesis. Since both muscle viscoelasticity and gain of stretch reflex varied linearly with muscle activity, stiffness of the hand prosthesis was designed to change linearly with the amplitude of integrated EMG (rectified and smoothed electromyographic signal). The digital servo system contained two systems, a force feedback system that mimicked dynamics of the neuromuscular control system, and a position control system. Both systems had their own one-chip microprocessors that ran independently. The one-chip microprocessor had the fundamental functions for control of the DC motor. That is, it had three digital I/O ports, four A/D converter channels, two output ports for PWM signals, and two input channels for timer counters. The usefulness of this hand prosthesis was indicated by three myoelectric control experiments with a normal subject, (1) voluntary control of the finger angle, (2) measuring responses to the disturbance torque, and (3) handling hard and soft objects. In the first experiment, the subject easily controlled our hand prosthesis with a myoelectric signal. The second experiment showed that the subject could control the stiffness of this hand prosthesis by changing the contraction levels of two antagonistic muscles. The third experiment showed that subjects who were not trained to control of the prosthesis could handle a soft object with our hand prosthesis.

抄録全体を表示

People having severe cerebral palsy cannot move by themselves due to physical limitation of their limbs. This always constitutes a great handicap for them. The purpose of our work is to facilitate independent movement for such people. We developed an electric powered wheelchair for a severely disabled cerebral-palsied girl. Although she cannot control her limbs, she can move her head as she intends. Measurement of her three-dimensional head movements indicated that she can rotate and flex her neck very well. As a result, her rotation angle and flexion angle were adopted as the controlling parameters. This system consisted of a three-dimensional position and angle sensor (ISOTRAK system), head operating start switches, an emergency stop switch, sensors to prevent collision, and a handheld personal computer. The rotation angle and flexion angle are measured by the ISOTRAK system and processed by computer. The wheelchair turns to the right after a head rotation to the right, and similarly for the left. A nodding movement from the operator stop the wheelchair. Operating functions can be changed by altering the computer software, so it can be easily adapted for other disabled people. The new wheelchair also features a special seating system to reinforce both the ability to control head movements and the postural stability. The subject first drove the wheelchair in a gymnasium which was wide and had a flat surface. She was able to go straight, turn right/left and stop very well. She could also propel herself a previously selected small spot. Next she drove on a pathway in an institution. By delicately controlling the direction, she could turn right/left. And an examination of her driving speed showed that an adequate speed was 2.0km/h. Moreover, we discovered that the tilt angle, which is the angle at which the chair is inclined, affected her head movement. As a result of measuring the change of position during driving, the appropriate angle found to be 22°. In the next phase, she drove on a road and sidewalk on the grounds of our rehabilitation center. She could control the direction by degrees, so that she was able to overcome the narrow width of the sidewalk and the tendency to turn downhill on a side slope. Measurement of her head movement while driving over a curb showed a shock to the neck, but it had no serious effect on control of the wheelchair. In developing a device that facilitates independence the severely disabled people, it is important to grasp the features of the individual, to make it easy to change the control functions, and to evaluate the device's actual action. The possibility of independent movement is greatly increased if all these factors are considered.

抄録全体を表示