Three-dimensional shape and leg-heel alignment of the right foot were measured for 8 male subjects under the following two conditions: 1) standing on a flat plane and 2) standing on a wooden base whose surface simulates the lower surface of shoe last (20mm of heel height). The coordinates of 23 sections were calculated for the data obtained under the first condition. All the sections were vertical to the reference plane defined by MT (metatarsale tibiale), MH and LH (medial and lateral heel points: medial and lateral points 30mm forward from heel point at the height of MT), which is parallel to the horizontal plane. For the second condition, the coordinates of 5 sections were calculated vertical to the surface of the wooden base for the part anterior to and crossing MT, and those of 18 sections vertical to the reference plane for the part posterior to and crossing MT. The differences in shape between the two conditions were analysed by using geometrical properties of each cross-section. Also the shape of the foot represented as a set of 17 equidistant cross-sections (sections 6-22) was described by the Fourier descriptors, and the shape difference was evaluated by the pseudo-distance measures, which are defined corresponding to the 5 independent global features of the 3-D object: elongatedness, horizontal strain, section shape, displacement, and torsion. There was no significant difference in the breadth of each cross-section. The height (sections 7-11), direction angle of principal axis (sections 9-13 and 20-22), and flatness index (sections 7-10) in condition 2 were larger than those in condition 1. Cross-sections 7-13 showed a swollen lateral part of the lower border of the outline and a pronounced plantar arch. The differences in leg-heel alignment between the 2 conditions indicate foot supination in condition 2. The differences in 3-D shape between the 2 conditions were clear only in displacement. These differences were probably due to the obliquity of the ankle in the coronal plane; movement of the leg over a fixed foot produces lateral rotation of the lower leg on planter flexion, which in turn produces supination of the foot. Other possible causes are change in the weight distribution on the metatarsal heads and tension in ligaments of the sole such as plantar aponeurosis.

View full abstract

Inertia properties of the body segments such as segment mass, location of the center of mass, and moment of inertia can be measured and predicted in a number of ingenious approaches. They can be classified into a) direct measurements on cadavers, b) indirect measurements on living subjects, and c) mathematical modelling. However, there is little information upon which complete inertial estimates for Japanese people, especially male and female athletes, can be based. The purposes of this study were to determine the mass, center of mass location, and moments of inertia of the body segments for Japanese male and female athletes using a mathematical modelling approach, and to develop a set of regression equations to estimate inertia properties of body segments using simple anthropometric measurements as predictors. Subjects were 215 male and 80 female athletes belonging to various college sport clubs. Each subject, wearing swimming suit and cap, was stereo-photographed in a standing position. Ten body segments including the upper and lower torso were modelled to be a system of elliptical zones 2cm thick based on Jensen and Yokoi et al. Significant prediction equations based on body height, body weight, and segment lengths were then sought, and some prediction strategies were examined. The results obtained were summarized as follows: 1) Table 2 provides a summary of mass ratios, center of mass location ratios and radius of gyration ratios for males and females. There were many significant differences in body segment parameters between the two sexes. This suggests the need to develop different prediction equations for males and females. 2) Close relationships were noted between segment masses and segment lengths and body weight as predictors for all body segments. Table 5 provides coefficients of multiple regression equations to predict segment masses. 3) No close relationship was noted between independent variables and estimates of the center of mass location. This indicates that the variance in the center of mass location in proportion to the segment length was very small, and that location of centers of mass could be estimated by the mean ratio provided in Table 2. 4) Close relationships were noted between segment moments of inertia and segment lengths (except hand and foot), and body weight as predictors. Tables 6 and 7 provide coefficients of multiple regression equations to predict segment moments of inertia from segment lengths and body weight.

View full abstract

Objective evaluation of the patient's condition is important in a clinical diagnosis. Since the frequency characteristics of the respiratory impedance represent the airway constriction, they are often used to diagnose the obstructive lung diseases. Forced oscillation technique is appropriate to measure the respiratory impedance, and has been often used. But due to the limit of the frequency response of conventional device, the frequency characteristics of human respiratory impedance could be measured only up to 20Hz. In this study, by differentiating displacement signal of the piston which excites the flow, the frequency characteristics of human respiratory impedance could be measured over the frequency range of 2-120Hz, and found to be characterized by two resonances at 8-10Hz and at 80-100Hz. Conventionally, to obtain the frequency characteristics, FFT technique was frequently used, and it required long measurement time. But since it is difficult to confine patients with advanced obstructive lung diseases to a measurement instrument for a long time, we tried to shorten the measurement time by employing the Auto-Regressive Moving Average model (ARMA model) parameter estimation technique, which made it possible to measure the frequency characteristics with a data of one second, corresponding to about one sixteenth of the conventional FFT method. In addition, by using the estimated ARMA model parameters, a 4-element equivalent electrical circuit model is derived for the human respiratory system. Four element values were measured for ten normal males, and found to be slightly lower than those of the previous reports, which had noticed only the resonance at 8-10Hz. Since the 4-element model represents the condition of the human respiratory system, the estimated values can be applied to the clinical diagnosis of lung diseases. The respiratory function test was carried out to determine the degree of constriction of human airway passages induced by cigarette smoke particle irritant aerosols. Nine normal male smokers were asked to inhale cigarette smoke deeply 15 times over 5 minutes (nicotin 1.35mg, and tar 21mg). After smoking, increases in resistance, which indicated the acute effect of airway passage constriction, were observed.

View full abstract

This paper describes motion analyses during running which used distance-velocity meter and bioelectrical impedance method. The distance-velocity meter is constructed of a rotary encoder, a string of high strength and high elasticity, and a guide wire. It gives information on the distance and velocity of the runner in every running point with high precision. A lower leg electrical impedance is used in this bioelectrical impedance method. This method gives information on landing point, take-off point, ankle angle and so on during running. Impedance wave form is able to observe intuitively the motion of running and to divide the running period. Using these results, we can analyze running in relation to stride length, pitch,running speed and running form. First, the measurement system is described in terms of construction, performance and measure ment method. The accuracy of measurement in distanceis ±(0.03%+0.8cm). The accuracy of measurement in velocity of running is 1% for an average distance over 48cm. The non-linearity of impedance measurement by telemetry is ±0.5% to the extent of ±10 ohms. The measuring area is about 100m from the antenna. Next, the method of running is analyzed. Impedance wave form during running is discussed in relation to the running motion. The results show that the characteristics of various running forms appear in the wave form of lower leg impedance. The stride length is also determined with the accuracy of about 1%. Last, this method is applied for 100m and middle-distance races, and discussed. This measurement system is simple and easy to handle. Hence, it can also be used for analysis of other field and track events-hurdle races or running broad jump, for example.

View full abstract

A system is developed to measure multi-axis joint angles by using improved flexible electrogoniometers. The weight and shape of the goniometer is smaller than the usual one which was developed in our laboratory a few years ago. Two pieces of specially designed strain gauge which are extra long are plastered on the two opposite sides of a metal thin beam with uniform section, to form a Wheatstone bridge circuit. The total amount of elongation of each strain gauge changes in proportion to the angle between the two ends of the beam and does not depend on the deformation curves of the beam. Thus, the output voltage of the bridge circuit changes in proportion to the angle between the two ends of the beam. The weight of the measuring part is under 1g; its width is under 3mm, thickness is under 0.1mm, and the length is variable up to 100mm. It is very easy to flex and to attach on the joint without considering alignment. It has no convex part to side direction, and can be used without taking off the clothes. These are great advantage for application in the clinical fields. The two goniometers are attached on the multi-axis joint perpendicular to each other to detect the flexed angle of the joint in the corresponding plane. The outputs of the goniometers are stored in a microcomputer through A/D converter, the joint angles on the two measuring planes are calculated with correction of the interference effect between the outputs. The correction of the interference is rather simple compared with other multi-axis electrogoniometers. These are processed in a very short time, and the result is displayed graphically on the CRT of the microcomputer in real time. This measuring system is applied to the human wrist joint of a normal subject. The result is displayed in real time as the relative motion of the hand segment against the forearm segment in three-dimensional space on the CRT display. The precise phenomena of motion can be observed, and the range of motion to the direction of dorsal, palmar, radial and ulnar flexion can be easily grasped quantitatively.

View full abstract

In this experiment the measurements were executed by specially developed apparatus, i. e., a force plate, a tactile image sensor system which measures the pressure distribution of the foot contact area, and an automatic stepmeter which measures the stride length and stride frequency. Video cameras were also used for the kinesiological analysis. All the apparatus were synchronized. We conducted experiments with 30 male and 30 female subjects to process the results statistically. For the walking speed we set up three categories: slow, ordinary, and fast. Walking speed was determined according to each person's subjective point of view. Ten trials in each category of walking speed were recorded and analyzed. The vertical component of the foot-ground reaction force measured by the force plate generally has first maximum amplitude, first minimum amplitude, and then second maximum amplitude during the stance phase. Walking speed has a high positive correlation with the first maximum amplitude and a high negative correlation with the first minimum amplitude. These results are similar to those in previous studies. But there was a problem with the relation between walking speed and the second maximum amplitude. One study pointed out a positive correlation, while another indicated no correlation between them. Previous studies, however, made the calculation for the sample as a whole. Our experiment reveals that there are differences between individuals. There is a positive correlation in some subjects, a negative correlation in others. In the remaining subjects there was no correlation between them. The automatic stepmeter measures the temporal change in the length of the sole contact area for each millisecond at a resolution of 5mm. The measured data reveal that more than 80 percent of the sole length (without toes) has already made contact with the floor at the moment of the first maximum amplitude of the vertical component of foot-reaction force. Trajectories of the point of application are calculated from the data on the force plate. The trajectory goes from the heel straight to the point between the head of the second and third metatarsal bones. There its speed slows down; then it goes toward the big toe with an angle of approximately 35 to 45°. This corresponds to the triangle structure of the supporting plane of the foot. The "isobarogram" and "maximum pressure isobarogram" were drawn based on the data from the tactile image sensor. From these illustrations, we see that the pressure at the heel, at the head of the second and the third metatarsal bones, and at the big toe or the second toe increases while walking. The pressure at these points is compared for each category of the walking speed. The differences were smallest at ordinary speed. It suggests that the construction of the sole is best adapted to this walking speed.

View full abstract

Recently, various motion measurement systems are used as a basic technique for motion analysis. These systems include electrogoniometric method and optical method by land markers put on the body surface. However, the former methods restrict human motion, and the land markers are missed and dislocated by deformation of the body surface. Additionally in clinical application, more simple method is required, which can reduce troubles with walking and fitting up the markers. In this study, we developed a new technique of motion measurement to fit the human geometrical model into the video image, in order to reduce the difficulties of measurement by land markers. The geometrical model contained 16 segments. Each segment had 1 degree of freedom of rotation and 2 degrees of freedom of translation. Human motion was recorded to a LASER disc recorder (TEAC: LA200A). Then the video image was played frame by frame and sent to a personal computer (TOSHIBA: J 3300/50) using a frame memory (HAMAMATSU: C1900). Subsequently, the body image was picked up using the thresholding method in image processing, and the geometrical model was fitted into this image. As the first step, the model was corrected roughly using a distance-transformed image which corresponded nearly to the center line of the body image. After rough fitting, the segment of the model was divided into 4 or 6 parts, and the model was corrected finely based on the direction and the value determined from the overlapping area between those divided parts and the body image. Repeating the fine correction, the joint angles can be estimated to about 1 degree. Using this method, we measured the normal gait motion with a CCD camera set at the distance of 3 meters. It takes about 20 seconds per frame to calculate, or about 10 minutes per 1 cycle of walking (30 frames). The amplitude of joint angle measured by this method agreed with what was measured by land markers simultaneously; the angles calculated by land markers had an offset because of the dislocation of the markers put on the joint. There was a tendency to decrease amplitude of shoulder joint angle with land markers and their cables. The influence of noise of image and model size error within 5% of its segment length are almost neglected in calculation of joint angles. This is because this method is using the information of overlapped area for fitting models. There still remain several problems in this method, such as automatic determination of the model size and detection from normal scene which includes background, clothes and tools. However, the developed method is characterized by unrestricted motions and simplified measurement.

View full abstract

Furi, mimic gestures in theatricals, are based on daily movements. In this study, daily movements and mimic gestures of the movement of hands and arms were compared. Subjects were 10 adult females and 30 infants. This test involved four experimental gestures: (1) the pot is empty; (2) the pot is half full of water; (3) the pot is full of water; (4) the pot is covered, and the subject does not know how much water it contained. Each experiment had three phases: (1) transport empty (reaching), (2) grasping, and (3) transport loaded (pulling back). The results were as follows. For daily movements, when there was an increase in the volume of water, the time required was longer, especially in the second phase. When subjects could not estimate the volume of water, the time required in the first phase was longer. When there was an increase in the volume of water, the area covered by hand movements was smaller, and the hand tended to move in a straight line. When there was an increase in the volume of water, the trunk of the body rotated around the longitudinal axis. The proximal parts of the body tended to be fixed. Movements of the distal parts of the body were smaller in the real action than in the mimic action. For mimic gestures, when there was an increase in the volume of water, the time required was longer in the third phase. The area covered by movement of the hand and elbow was larger, and the hand tended to move in a curved line. The proximal parts of the body moved with the distal parts. When there was an increase in the volume of water, the trunk of the body leaned anteriorly. As for infants under 4 years, there is no significant differences the time and spaces of upper limb movements effected with object they estimate. The results indicated that the time required was in proportion to the volume of water in both mimic gestures and daily movements. However, in daily movements the time required in the second phase is larger, while in mimic gestures that in the third phase is larger. The area covered by hand movements is even larger. In mimic gestures of theatricals these characteristics were clearly seen. It can be said that actors "overact" or omit some of the daily movements, according to the dramatic requirements of the scenes.

View full abstract

It has been recognized that atherosclerotic lesions tend to occur preferentially in certain areas of the arterial vessels, and that they are related to the elevated and disturbed hemodynamic forces in regions of nonuniform or complex conduit geometries sush as branch points. This finding has led many investigators to consider the relation of local hemodynamic flow patterns to the development of atherosclerosis. However, the explanation of the mechanism of development of atherosclerosis still requires reliable information on hemodynamic factors under physiologically relevant flow conditions. In the present study, flow visualization and measurements of wall shear stress have been carried out for the model conduit of a renal artery branch with a 90°take-off angle. The renal artery is characterized by side branching, and its branched shape may cause significant flow distortions and influence the wall shear stress profile. Flow visualization in the branching vessel revealed that the geometry of the branched entrance plays a significant role for the formation of a vortex in the side branch; a branched vessel, in which the proximal lip is rounded and the distal one squared, and a branch with sharp edges were employed as model renal arteries. The nature of flow in a side-branch is particularly affected by the entrance geometry. In the branch with rounded edges, no separation occurs in the proximal wall, although a secondary flow is induced due to the curved strean line. However, a separated region was generated in the branch with sharp edges, and double pairs of secondary flow appear in the side branch. We have measured the wall shear stress profile in the model renal artery branch by means of electrochemical technique. The shear stress profile in steady flow at the Reynolds number of 1000 was obtained, and the wall shear stress becomes high at the inner wall of the main branch and the distal wall of the side branch. A significant non-uniformity of wall shear stress found in the curved and branched vessels due to the three-dimensional flow structure plays an important role in the localization of atherosclerotic lesions.

View full abstract

Many researchers have attempted to measure the change in velocity of the center of gravity (CG) for speed skaters in stroking. However, there have been some difficulties in measuring the velocity change in speed skating; they include the very long stride length (about 10m) and the three-dimensional behavior of the skater's CG. The purposes of this study were to investigate the push-off technique for top-level Japanese speed skaters and the change in velocity during the push-off phase using 3D cinematography, and then to examine acceleration theory during the push-off phase in speed skating. Skaters participating in the 500m race of the All Japan student championship (1989) were videotaped (60 fields/s) by 10 VTR cameras over 20m at the crossing zone of the back straightaway. Twenty-two male skaters were selected as subjects and classified into two groups on the basis of the performance of the competition. 3D coordinates of the segment endpoints were obtained on five sub-areas (each 4m in length) using a DLT method. Displacement and velocity of the CG and the angles of the hip, knee and ankle joint were calculated. The results obtained are summarized as follows: 1) Push-off movement for the top group skaters placed the CG further forward than that of the second group. 2) The vector derived from the push-off movement for the top group skaters was directed forward, and accelerated the CG of the skaters effectively. 3) Increase in the velocity in skating direction for all subjects seemed to contribute more than expected to the acceleration of the CG. It has been proposed that acceleration in speed skating occurs by push-off of the leg in a direction perpendicular to the gliding direction of the skate, since the force applied to the opposite direction of gliding cannot contribute to acceleration of the CG due to very small frictional force. However, this theory cannot thoroughly explain the findings obtained for the top skaters in this investigation. Therefore, the acceleration theory should be modified to reflect the fact that the CG of the skater during speed skating is accelerated not only by the push-off perpendicular to the gliding direction but also by an increase in velocity vector in a gliding direction.

View full abstract

In recent years techniques of performance in international-level gymnastics have shown rapid development. Risk, originality, and virtuosity are required to win higher evaluations for routines. New airborne twist-somersaults have increased the risk of injuries. Coaching methods for these skills have been developed through experience rather than scientifically. Computer simulation enables the coach to isolate each component of a skill and to evaluate their overall contribution to performance without risking injury, and computer graphics can be used to explore new or better ways to perform twist-somersaults. The construction of the model itself may provide a deeper understanding of the mechanics of a performance. In this paper a computer simulation capable of completing the new airborne twisting-somersaults from a vaulting horse are described, and the points for coaching them are developed. Equations of motion for the simulation were based on the model of the human body presented by Tang (1989). The path of the gymnast's mass center was established by the application of Newton's laws of motion. Since air resistance was negligible, the angular momentum of the gymnast about his mass center remained unchanged throughout the flight phase. The equations governing the gymnast's body segment parameters and his translation and rotation during a twisting-somersault were programmed for solution on a personal computer. Inputs to this program included the gymnast's anthropometric measurements (BSP), the initial conditions of the technique and its boundary values. The computer then processed this information and supplied numerical output which completely defined the angle of twisting and somersault as a function of time. The data (velocity vector of mass center at the instant of takeoff from the long horse, the angular velocities of the twist and somersault, the time of arm swing, the duration of flight and the landing angle) for input were obtained from measurement of actual twist-somersaults with three-dimensional cinematography.

View full abstract

The shoulder complex is composed of the scapula, clavicle and humerus. Shoulder motion is the result of cooperation of these elements. So far motion analysis of the shoulder complex has been carried out by means of X-ray photography. However, it is difficult to measure the three-dimensional motion of each element by this method. Using a magnetic sensor system (3 SPACE), we tried to measure three-dimensional motion of the scapula and humerus in elevation in the sagittal, scapular and frontal planes. This magnetic sensor system is composed of a source coil which generates a magnetic field and a sensor coil which detects electromotive force corresponding to attitude and distance of the sensor coil with respect to the source coil. The source coil was attached on the chest and the first sensor coil was attached to the humerus. We tried to fix the second sensor coil to the scapula by grasping the spine of the scapula together with the sensor. This method is not so accurate but it is safe and easy. We measured Euler angles of the scapula (scapulothoracic angle) and humerus and excursion of the scapula with respect to the reference coordinate attached to the trunk. And we calculated relative Euler angles between the scapula and humerus (glenohumeral angle). The ratio of glenohumeral to scapulothoracic angle during elevation was not constant, as has been reported so far. Instead this ratio varied considerably. In elevation in the sagittal and scapular plane, the scapula moved posteriorly and inferiorly at maximal elevation. However, in abduction the scapula moved medially and posteriorly.

View full abstract

The human shoulder joint has a very wide range of motion for movement of the hand, which performs one of the most important functions in the human activities, to anywhere in the space. The shoulder joint also has a very complex mechanism for stabilizing the upper extremity in any condition. Because of the above factors, it poses interesting biomechanical problems. In addition, the dynamic conditions around the shoulder cannot be disregarded clinically. However, there are very few biomechanical studies of its behavior, except for measurements of scapular motion and/or electromyographical activity, because its structure is too complex to analyze, thus there are few accurate data about the shoulder complex. In this paper, we develop a biomechanical model of the shoulder joint and measure some important parameters of the model. The model has four rigid bone segments connected by 19 ligaments and two joint capsules, and 37 muscles control the bone segments. In order to better understand the shoulder complex, we developed the generation method of ordinary shoulder behavior using a computer simulation technique. In this simulation, we assumed that shoulder behavior is used to minimize the energy expended through the motion, and the muscle force was determined so as to minimize the sum of the stress of each muscle. By using non-linear minimization strategy, we can generate shoulder behavior and estimate the internal force around the shoulder joint, for example the muscle force. Compared with actual human shoulder behavior, the simulated result of this method was very similar. By simulating the results of elevating the upper extremity using actual human parameters and hypothetical ones, we found the following: The form of the humeral head elevates the scapula naturally depending upon the elevation of the upper extremity. This behavior is highly energy-efficiency. The structure of the rotator cuff has an effect both upon the dynamic stabilizing mechanism and in improving the energetic efficiency. Additionally, in this simulation the musculoskeletal system can be changed at will. Thus, it is very useful for the evaluation of surgical methods and for the development and/or evaluation of prostheses, because the behavior under hypothetical conditions can be simulated by this method.

View full abstract

The anterior-posterior displacement of the tibia elicited by the loading of the quadriceps and hamstring muscles was determined as a function of joint angle and muscle load using the data collected from five fresh cadaver knees with a highly accurate computerized radio-graphic technique. A two-dimensional mathematical model, taking into account movements and forces of the patellofemoral and tibiofemoral joints in the sagittal plane, was described and a computer simulation was performed to verify the experimental results. The simulated and experimental results closely coincided. Both the results demonstrated that quadriceps contraction can result in an anterior displacement of the tibia in the range of 0°to 40°of flexion, and in a posterior displacement in the range of 80°to 120°of flexion. However, hamstrings contraction always causes a posterior displacement of the tibia, irrespective of knee flexion angle. Thus it was concluded that quadriceps contraction has a direct impact on ACL stress, as hamstring contraction does on PCL stress. It was further concluded, however, that the absolute magnitudes of both the cruciate ligaments were not so much influenced by the thigh muscles' contraction as they were influenced by knee flexion angle, a conclusion that throws into question the assessment of cruciate ligament stresses by anterior-posterior displacement of the tibia. Some useful parameters that serve as a function of knee flexion angle were also introduced through the simulation: contact force and slipping ratio of the patellofemoral and tibiofemoral joints, variation of the patellar ligament force, and thigh muscle length. There is a linear relationship between quadriceps muscle force and patellofemoral contact force, whereas there is little relationship between quadriceps muscle force and tibiofemoral contact force. A semilinear relationship is observed between knee flexion angle and patello-femoral contact force. The tibiofemoral contact force shows a bell-shaped pattern against knee flexion angle. Variation of the patellofemoral slipping ratio shows a complex form in which the glide and roll of the patella on the femur take place in the opposite direction for 0°to 95°of knee flexion while glide and roll take place in the same direction for more than 95°of flexion. Variation of tibiofemoral slipping ratio shows that the femur mainly glides on the same position of the tibia between 30°and 90°of knee flexion. Linear relationships do exist between knee flexion angle and, respectively, quadriceps length and hamstrings length. Patellar ligament force varies from a similar value to about 60% of the quadriceps force as the knee flexes.

View full abstract

In order to improve human performance in competitive sports, human motions have been analyzed from various viewpoints. One of the well-known methods is to compare the patterns of joint torques with those of top athletes during the motions. However, such comparisons are not necessarily sufficient to evaluate human motion, since individual differences in muscular forces and anatomical characteristics are not taken into consideration. The aim of this study is to investigate the relationship between the optimal vertical jump (squat jump) motion and the mechanical characteristics of muscular forces. The vertical jump motions of four male subjects were analyzed; their anatomical characteristics are similar, but the muscular force characteristics are different. Each subject performed a squat jump, in response tothe command "jump ashigh as possible," from an initial squat position with 90°hip and knee angles. The reaction force from the platform, the motions of the limb and the body, and the electromyographic (EMG) data were recorded. In order to estimate the optimal control for vertical jump motion under various conditions, a simulation system is applied which is based on the musculoskeletal model with the mean characteristics of muscular forces and the anatomical parameters of the four subjects chosen. In order to investigate the relationship between the optimal vertical jump motion and the mechanical characteristics of muscular force, a series of simulations was carried out by varying the parameters of the musculoskeletal model, such as the force-velocity relationship and the maximum contraction force. The following conclusions are derived from the results of the experiments and the simulations. (1) Change in the force-velocity relationship of human muscles results in a change in the optimal vertical motion and the sequence of the firing pattern of muscles, so that the contraction velocity of muscles does not become extremely large. (2) Changes in maximum contraction forces of some muscles result in changes in the optimal vertical motion and the sequence of the firing pattern of muscles, so that each muscle contracts under the optimal condition for the vertical jump. (3) The firing pattern of muscles is governed by the relationship between the anatomical characteristics and the muscular forces, and the timing of firing is determined by the relationship between the firing patterns and the maximum muscular forces. (4) The maximum contraction forces and the force-velocity relationship have to be improved in order to improve vertical jump performance.

View full abstract

There are numerous combinations of cadence and stride length that result in a walk with a certain speed. However, natural walking has high predictability, and is done within a limited range of cadence and stride for each individual. The purpose of this study is to clarify the determination criteria of this daily motion from a biomechanical viewpoint. One of the factors that determines the criteria is the efficiency of energy consumption during walking. Conventionally, the energy consumed by body exercise was estimated from the oxygen consumption volume at the respiration. In this study, consumed energy in muscles was calculated using a musculoskeletal model and measured articular motion in walking. With this simulation, the cause and the mechanism in energy change can be analyzed. Articular moments in walking were calculated with a rigid link model of seven two-dimensional segments which were constructed independently in sagittal and frontal planes. The link system was attached to eleven muscles, and the muscle activities were assumed to be controlled in order to minimize muscle fatigue. Using these assumptions, muscle tension was calculated under the restriction of equilibria between the articular moment and the muscular moment by an optimization technique. Furthermore, supposing a metabolic model of muscle that has two elastic elements and two viscosity elements, the force of the construction element in each muscle was estimated from the calculated muscle tension and the measured articular motion. The metabolic energy calculated from the mechanical work of the contraction element and the heat loss from each element of the muscle model agreed well with the experimental results by gas analysis. Articular motion in walking was measured by a semiconductor camera using infrared markers. External force on foot was measured by a force platform. Electromyograms of six surface muscles were measured for the evaluation of the calculated muscle tensions. Totals of 600 various kinds of cadences and stride lengths were obtained from ten male subjects. From the results of the calculated metabolic energy in these walking experiments, it was found that there was a combination of cadence and stride length which minimized the metabolic energy per unit distance and unit body weight. The walking speed at the minimum energy point agreed with the average speed of free walking. The cause of the minimization was as follows. The energy expended by muscle activities in the sagittal plane for promotion increased in high-speed walking. In contrast, the energy expended by muscles to support body weight and fundamental metabolic energy increase in low-speed walking. Accordingly, it was concluded that cadence and stride length in natural walking were determined so as to minimize total metabolic energy per unit of locomotion distance.

View full abstract

In skeletal muscles which have muscle fibers running parallel to each other, motor unit action potentials (MUAPs) arise from the innervation zones (motor end-plate regions) and propagate bidirectionally along the muscle fibers. Using this property, it is possible to estimate the location of the innervation zones by recording the EMG signals simultaneously with multiple-contact surface electrodes. This method for defining the innervation zones has been successfully applied to the muscles in the upper and lower extremities, because these muscles have long fibers running parallel to each other and consequently the propagation of MUAPs is easily detectable. However, if this technique were applicable only to the muscles in the extremities, it would have a limited application. The purpose of this study is to clarify the applicability of this technique to the back muscles: the trapezius, the latissimus dorsi, and the erectores spinae. These muscles are related to ergonomic problems such as a stiff neck or the low back pain. Five healthy male subjects were studied during an isometric contraction with a 15-channel surface electrode array. In the muscles which show the propagation of MUAPs, a topographical map of the innervation zones was obtained by moving the electrode array over the surface of the muscle. In the trapezius, which is separated into four portions, the MUAPs showed a clear propagation pattern, and the map of the innervation zones was clarified. The innervation zones lay in each portion at the middle of the muscle, forming a line perpendicular to the muscle fibers. In the latissimus dorsi the propagation was detected, but it was not so clear as in the trapezius. The innervation zones were scattered concentrically around the axillary portion, and the distribution of the innervation zones differed between the subjects. In the erectores spinae the bidirectional propagation of MUAPs was not observed except in one subject. This is perhaps because the muscle fibers have a complex configuration in this muscle.

View full abstract

The average conduction velocity of human skeletal muscle fibers measured with multicontact surface electrodes has been reported to increase with the muscular contraction force. This increase of the conduction velocity is considered to be a manifestation of the size principle of motor units: the larger motor units with higher recruitment thresholds tend to have higher conduction velocities. However, this increase was not observed in all the muscles studied. The present study is designed to further clarify this relationship between the contraction force and the average muscle fiber conduction velocity. The subjects were requested to track the force trajectory presented on a CRT screen. The force trajectory had three plateaus, which lasted 2s each and were connected by straight lines. The first and third plateaus of the force trajectory corresponded to the sustained contraction of 50% of the maximal voluntary contraction (MVC), which was used as the reference. The second plateau corresponded to a variable target contraction force that was set at the four levels of 30%, 50%, 70% and 90% MVC. The biceps brachii, the tibialis anterior and the vastus lateralis of five healthy males were studied. The average muscle fiber conduction velocity was calculated by the cross-correlation between two myoelectric signals derived from positions 15mm apart along the muscle fibers. In some recordings, the conduction velocity at the 50% MVC after the target contraction decreased significantly from the velocity at the first 50% MVC reference contraction. This decrease was most prominent in the biceps brachii and is considered to be a manifestation of muscular fatigue. The recordings that showed no decrease of the conduction velocity after the target contraction indicated a positive correlation between the contraction force and the conduction velocity; that is, the conduction velocity had higher value during the contraction of larger force. This result indicates a tendency for the conduction velocity to increase with the contraction force, as previously reported, but the decrease of the velocity caused by muscular fatigue may hide this tendency.

View full abstract

This paper describes the importance of an adjustable impedance control for contact task robots from the vantage point of human arm control. We know that the human arm operates especially well in contact tasks. The reason is most probably that the joint impedance of the human arm is adjusted directly by muscle viscous friction control and elasticity control. Incidentally, elasticity change is obvious though viscous friction change is not yet obvious in continuous motion pertaining to in vivo muscle. To estimate the muscle viscous friction change and elasticity change in a contact task, we carried out some visual target tracking experiments on the human wrist joint with an elastic motional load. In the experiments, the target moves at a constant speed and the subject tracks a target against the elastic torque of the load. As a result of the experiments, we found that oscillation frequency of the tracking motion of the human wrist joint shifts continuously to a higher region that corresponds with loaded torque increase. This frequency shift suggests that the human arm neuromuscular system changes its viscous frictional coefficient and/or elasticity coefficient continuously by muscle activity change. However, it is not obvious which affects the frequency shift more, the viscous friction change or the elasticity change. Therefore, we carried out computer simulations of dynamic motion on the neuromuscular system model with elastic load. In the model, active muscle was represented by a force generating unit, parallel elastic and viscous frictional components, and a serial elastic component. The coefficients of these components were supposed to change with muscle activity. The simulations were made in three different conditions. In the first condition, the muscle elastic coefficients and viscous frictional coefficient were changed according to the elastic torque of the load. In the second, the elastic coefficients were changed but the viscous coefficient was not changed, and in the third, the viscous coefficient was changed but the elastic coefficients were not changed. Our results showed that the frequency shift depends mainly on viscous frictional coefficient change. Needless to say, the muscle elastic coefficients have to change according to the loaded torque change. Because of the correspondence of the experimental results with the simulations, we can say that the human arm adjusts its joint impedance continuously by muscle viscous friction and elasticity control. We can also say that the compliant and stable robot performing contact task is best controlled by adjusting joint impedance continuously with actuator torque.

View full abstract

The way in which the multi-degree-of-freedom musculoskeletal systems are coordinated adaptively during skilled actions as diverse as walking, running, grasping, handwriting and reaching for a target is one of the fundamental issues in understanding human movement. The degrees of freedom of a system are "the least number of independent coordinates required to specify the position of the system elements without violating any geometrical constraints." Generally speaking, the larger the number of degrees of freedom in a dynamic system, the more difficult it is to make the system behave as desired due to the nonlinear interactions between degrees of freedom. It is known that skilled actions are planned in a corresponding-task spatial coordinate system. The planned actions must be transformed into body (joint) space form and muscle space form. However, generally, the number of degrees of freedom in the musculoskeletal system engaged in a task exceeds the number of degrees of freedom needed to specify the task to be carried out. The motor control problem is highly indeterminate with respect to most tasks. Therefore, it is necessary to find constraint conditions for transforming a planned trajectory or force in the task space into that in the joint or muscle spaces. In the present paper, we discuss the roles of motor impedance in reducing the redandunt degrees of freedom on the musculoskeletal system. First, we give the motor impedance transformations among the task, the body and the muscle spaces, and then inverse kinematics solutions using the impedance constraints. Further, it is shown that in crank rotation tasks, the subjects determine limb postures based on the hand and joint impedances.

View full abstract

The driving system of a manipulator is divided into two groups of movements. One group is driven directly by the actuator on each joint, and the other is driven indirectly by an actuator which is set apart from the manipulator. The second type of movement allows the manipulator to be made small. We have researched a mechanism and control method that can move as softly as the human arm, and we have adopted wire cables for the driving system of our human-type manipulator. The cable transmits the force from another, separate motor. This is similar to the human muscle system. But in most cases in which a cable is used to drive the manipulator, it is used as the means of transmitting the rotational torque from separated motors to each joint. But this is basically different from the action of a muscle. The muscles contend with each other, and one acts on the joints. Making the most of the characteristics of human mechanisms, we aim to achieve a skilled motion, like that of the human arm, that can control the position and force precisely. In the study reported in this paper, we looked at the human elbow joint and made a manipulator joint that has a similar mechanism that is characteristic of the human elbow joint. This joint consists of three links which correspond to the bones of the human arm and has 2 degrees of freedom (d. o. f.), flexion and rotation. This mechanism is driven by three cables which transmit the force from a separate motor. One cable has 2 d. o. f., and others work in contention. The basic control method is PI feedback control. Besides the main feedback loop, this algorithm has minor feedback loops that return the tension of the cables. We examined the position control. From the results of the numerical simulation and the experiment on this joint, we found that the algorithm of the contending force for 1 d. o. f. motion is able to extend to the control of 2 d. o. f. motions which are driven by three cables. The coefficient of the feedback loop does not influence the time response of the joint angle. This minor feedback of tension is restrained from being excessive, but excessive feedback makes the position control unstable.

View full abstract

A large part of the work performed daily by humans is so-called constraint work, in which the object of work is under some external constraint. Under constraint conditions, it is difficult to mechanically regulate the force of the hands and arms so as not to apply undue force to the object of work, but humans perform this with ease. This study aims to clarify the human hand and arm mechanism of force control by analyzing constraint work performed by humans from the standpoint of biomechanism, and to study a manipulator control strategy on the basis of it. This paper presents the results of an experiment and analysis of human motions in handling a bead on a guide shaft as an example of constraint work with 1 degree of freedom (d. o. f.), and observations of the roles that the hand and arm play in manual work. It also deals with applications to manipulator control. In the experiment conducted on the working the handling bead on a shaft by a seated human being, the change in translational force and rotational force applied to the shaft during the work and the shaft position and hand position were detected, combining the following four conditions in various ways: (1) Upper body d. o. f. (2) Hand d. o. f. (3) Forced displacement (4) Direction of motion The results of the experiments are summarized below. (1) Humans do not use the d. o. f. of the upper body to absorb the translational and rotational forces generated on a constraint surface. (2) Humans absorb the rotational force generated by constraint using the d. o. f. of the hand. (3) Humans absorb the translational force generated from the translational displacement of the shaft by the d. o. f. of the arm, and absorb translational forces by the arm regardless of the direction of motion. (4) There is a direction of motion or a posture in which translational force can be better absorbed because the degree of absorption of translational force varies depending on the direction of motion. As regards the application to robots of compliance control using finger-arm coordination, compliance control of the arm is discussed. Thus, the rotational component of compliance is disregarded, while only the relationship of translational force with translational displacement in the hand is dealt with. Simplifying the relational expression of force and displacement, the end-point compliance matrix can be expressed as a symmetrical matrix with 6 elements. A desired translational compliance can be set in the end-point of a manipulator having an arm with 6 d. o. f or more.

View full abstract

A new system has been developed which can estimate postural mechanism using a movable platform. An outstanding characteristic is that the center of pressure (COP) for each leg can be measured. Based on experimental measurement of COP in the static and dynamic conditions of the platform, the maximum error was set within the allowable regions. The main theme of the present research is to develop parameters to assess how to control postural stability under a preprogrammed response pattern, under the command of the central nervous system. By measuring the angular displacements mainly in the lower extremities (by PSD cameras), several experiments estimated postural response for (1) forward/backward movement and (2) lateral inclination stimulation. Subjects are 25 males (ranging from 24 to 72 years old). As a first stage, we conducted a power spectrum estimation (using an autoregressive model of the time seris analysis) of data on COP and angular displacements. As a pre-programmed response pattern in the foreward/backward movement, its response patterns were complex, but the ankle and hip strategy were found at a specified stage of the intensity of stimulation for each subject. The older group made use of the hip strategy to hold a postural stability under comparatively smaller stimulation than the younger group. In the case of hip strategy, the spectral curve often shows a peak around 1 Hz. In the lateral inclination stimulation, we found that the hip was one of the main parts involved in controlling a stable and smooth response. In a stronger stimulation, a spectral curve often shows a remarkable second peak from 2.0 to 2.5 Hz, besides a 1 Hz-peak similar to the forward/ backward movement. This suggests a more complex response pattern of the hip and trunk. Generally a vectral presentation of COP was useful to express the accuracy of the response pattern, giving the latency and the difference between two legs. Frequency analysis was also useful to represent some parameters of postural responses by a movable platform, and could be used to present a limited stable response under a specified frequency stimulation.

View full abstract

Muscle fatigue is one of the serious problems of functional electrical stimulation (FES), because possible standing time is limited by the fatigue. In this study, the three factors of muscle fatigue on FES standing in paraplegia is discussed. First, standing postures with a long leg brace (LLB) and two FES channels were analyzed. The subject was a male, 18-year-old, paraplegic patient with a T8 spinal cord injury. The knee joint flexion moment was calculated from the relationship between the zero moment point of the ground reaction force and the knee joint position. The estimated knee joint flexion moments were 16 Nm (FES) and 28 Nm (LLB). The average moment of three normal subjects was 14 Nm in the extending direction. It was found that the knee joint moment of a paraplegic is significantly greater than that of a normal subject. Knee flexion moment requires a large contraction force of quadriceps. This is one of the reasons for quadriceps fatigue. Second, the properties of muscle fatigue against electrical stimulation in paraplegia and normal subjects were measured. The contractile force decrease by 30 seconds' maximal intensity continuous stimulation varied from 44 to 10 percent in paraplegia. The contractile force decrease in normal subjects was 17 percent on average. The obvious difference between paraplegiacs and normal subjects was not found in the fatigue property. The high-intensity continuous stimulation is a serious problem, because it brings considerable muscle fatigue in both paraplegic and normal subjects. Finally, the fatigue rates caused by the 30 seconds' continuous stimulation and the 60 seconds' alternative stimulation, 3 sec on and 3 sec off, were compared. The alternative stimulation showed a lower fatigue rate. This result shows that stimulation stop or decrease can decrease muscle fatigue. According to this result, an FES standing device was developed. The device was used in the artificial reflex method, one of the stimulation sequences driven by knee joint flexion. In conclusion, three factors were pointed out. 1) Two-channel FES standing requires large quadriceps contractile force. 2) High-intensity continuous stimulation brings serious muscle fatigue. 3) Alternative stimulation decreases the muscle fatigue.

View full abstract

Functional Electrical Stimulation (FES) is one of the ways to restore motor functions of paralyzed extremities in stroke or cervical cord injury patients. By using our multi-channel FES system, some activities of daily living were restored in a paralyzed hand and arm. Howerer, in order to apply this system in clinical cases, selection of adequate control commands obtained from the patients' residual functions is essential for easy and sophisticated operations. This paper describes control strategy of EMG processing and its application to a C7 quadri-plegic patient for restoring his hand movements. Electromyograms (EMGs) during wrist movements were examined in five normal volunteers by bipolar intramuscular electrodes in order to select the wrist muscles for detecting control signals. It was found that the wrist movements were mainly differentiated by EMGs of the following muscles: extension by the extensor carpi radialis brevis (ECRB), abduction by the extensor carpi radialis longus (ECRL) and flexion by the flexor carpi ulnaris (FCU) or flexor carpi radialis (FCR). Therefore, EMGs of ECRB during wrist extension and FCR during wrist flexion were utilized as control signals for hand grasping and opening movements, respectively. Elimination of the stimulating pulses and evoked potentials (Mwave, etc.) from muscle discharges during active movements was performed by establishing a window of 16 msec for detecting muscle discharges preceding each stimulating pulse. From the clinical application of this system, it was found that grasping and releasing of the patient's hand were successfully achieved by EMG commands of the wrist muscles. Since the patient had utilized tenodesis for grasping light objects, the EMG commands detected from the wrist muscles provided easy and natural control of the hand movement. Thus, EMGs of the wrist muscles are suitable as a control command for FES. Further investigations on the processing of the EMG and the selection of muscles are necessary to develop a more sophisticated FES system.

View full abstract

The application of Functional Electrical Stimulation (FES) to body control has been proposed for the restoration of the voluntary rolling-over motion of a paralyzed body. EMG analysis is performed in order to consider muscle activities for the realization of the rolling-over motion by FES. A normal subject performed the rolling-over motion from the supine position to the lateroabdominal position. The sequence of the rollingover motion was constructed of four phases as follows: 1) flexion of the right lower extremity, 2) internal rotation and adduction of the right hip joint, 3) left rotation of the pelvis, and 4) left rotation of the chest. Multichannel EMG signals measured simultaneously by the bipolar method using percutaneous electrodes were amplified, full-wave-rectified, and smoothed (time constant 1 sec). We obtained valuable data to achieve the rolling-over motion by FES through the analysis of the EMG signals. For example, the activity of the obliquus externus abdominis (left), the rectus abdominis (left), the erector spinae (right), and the quadratus lumborum (right) had a similar pattern during the rolling-over motion, i. e., they increased in the lower-extremity-flexion phase, had a high peak in the hip-rotation-adduction phase, kept a constant level in the pelvis-rotation phase, and decreased in the chest-rotation phase. These muscles are located at the same or symmetrical positions relative to the axis of the body. Hence these results indicated that they seemed to play a synergetic role in the twisting of the chest and the waist, as in a rolling-over motion. The characteristics mentioned above can be useful in the data compression technique for the creation of stimulus patterns of FES, and simplification of the stimulation system. The rolling-over motion by means of FES does not require the precise control which is an important factor in the restoration of the upper extremity motion, e. g., opening or grasping of fingers. The results of this study suggest the fundamental feasibility of the control of the body movement by means of FES. The knowledge will also be applicable to the strengthening of the muscles through Therapeutic Electrical Stimulation (TES).

View full abstract

We propose a new method for evaluating spasticity of ankle joint muscles by using microvibration (MV) measurement. In order to measure the MV, a piezo-electric pick-up sensor was used and set on the M. gastrocnemius. The MV signal was detected in both sitting and standing positions, and was recorded. Thereafter, the MV signal was analyzed by digital computer. The subjects consisted of two groups: (1) 11 male healthy individuals and (2) 10 male and 1 female hemiplegic individuals with spasticity of ankle joint muscles. The MV signal was subjected to Fast Fourier Transform (FFT), with sampling frequency of 500 [Hz] and 1024 data points. For estimating the difference of MV characteristics between healthy and spastic subjects, the following analyzing method was employed: First, FFT amplitude spectra of the MV signals in sitting and standing positions were obtained and the latter amplitude spectrum was subtracted by the former amplitude spectrum. Next, low frequency band component (L. F. B.) and high frequency band component (H. F. B.) were obtained from the subtracted spectrum. Ratio of both components to the total spectral power were calculated as the percent expression (L. F. B. C. [%] and H. F. B. C. [%]). Furthermore, L. F. B. C. [%] and H. F. B. C. [%] were plotted on x-y coordinates, and Euclid distance between 4 healthy and 6 spastic subjects' data was calculated. We obtained maximum Euclid distance when L. F. B. having the center frequency at 5 [Hz] with the band width of 10 [Hz] and H. F. B. having the center frequency at 35 [Hz] with the band width of 30 [Hz] respectively. With this discrimination method, MV characteristics of 11 healthy and 11 spastic subjects were plotted on x-y coordinates. The result showed a significant difference of MV characteristics between normal and spastic subject groups. Increase in H. F. B. and decrease in L. F. B. represented decrease of spasticity. This was also verified and correlated with a phyisio-therapist's evaluation. Furthermore, continuous change of spasticity in standing position on an oblique board was found by using this method.

View full abstract

Paraplegics cannot walk in the manner they used to before the onset of spinal cord injury, because they have lost the power in the lower extremities that is essential for a normal bipedal gait. Yet paraplegics can regain their walking ability by using two crutches and long leg braces after having had appropriate physical therapy. Here the problem arises. Most paraplegics choose not to continue crutch ambulation but prefer wheelchair ambulation, at some time after being discharged from rehabilitation hospitals. We postulated that the crutch swing-through gait is very unlike the normal bipedal gait. So it is very difficult for any adult to master this mode of locomotion. To explain this assumption, we conducted energy cost measurement of the crutch swing-through ambulation. We did not adopt the conventional O_2 consumption method, but used a mathematical link model analysis. We followed the method that we reported previously, in which we calculated leg muscle power and energy consumption in normal bipedal gait. This consists of measurement of trajectories of landmarks placed on the body, floor reaction forces, and foot and floor contact timing. For the link and source of power for crutch ambulation, we reduced human body to 5 segments and 9 muscles on one side of the body. Weight and center of the mass of each segment were measured by a technique that processes a graphical image of body shape, which we have reported previously. The maximum strength of 9 muscles was measured by a strain gauge appliance and a Cybex machine. All data were fed into an NEC9801RX computer to obtain contraction speed and length of each muscle during crutch ambulation. These parameters were used to compute muscle strength and power using Hill's equation. Energy consumption per unit of time was obtained by integrating muscle power. We used a non-paraplegic subject (23years old, 172cm height, 63kg body weight), who simulated paraplegic swing-through ambulation at 3 different speeds. The results were as follows. At 42.0m/min walking speed, energy consumption per minute was 76.5 (cal/kg/min), permeterwas 1.74 (cal/kg/m). At 57.6m/min walking speed, energy consumption per minute was 70.3 (cal/kg/min), per meter was 1.22 (cal/kg/m). At 61.2m/min walking speed, energy consumption per minute was 64.4 (cal/kg/min), and that per meter was 1.10 (cal/kg/m). These results agreed with those of other workers who measured energy consumption of paraplegic walk by conventional technique. Paraplegics use shoulder girdle depressors, elbow extensors and shoulder flexors for this mode of locomotion. Unless these muscles are strengthened, it seems difficult to master the swing-through ambulation.

View full abstract

The ankle joint moment in the saggital plane, dorsi and plantar flexion of the ankle joint, flexion and extension of the knee joint, and temporal factors of gait were measured continuously for 10 hemiplegic patients under two conditions: with and without AFO. The ankle joint moment was measured by a capacitive transducer which detected vertical floor reaction forces during the gait. All devices were attached to the subject's body, so that 8 to 20 strides could be measured continuously during 30 seconds. A specially disigned AFO was used in order to decompose the total ankle joint moment into two factors, the moment due to muscle forces and the moment due to the corrective forces generated by the AFO. The data were analyzed with special focus on the corrective moment of the AFO. The maximum dorsi-flexion moment due to the AFO in the early stage of the stance phase was less than 5Nm, and the maximum plantar-flexion moment in the late stage of the stance phase was less than 8Nm. These values were small in comparison with the total ankle joint moment, which ranged up to 70Nm. In regard to the effect of the AFO, the following general tendencies were statistically found. 1. The gait cycle decreased. 2. The single stance phase duration of the affected leg increased. 3. The swing phase duration of the affected leg decreased. 4. The maximum dorsi-flexion angle of the ankle joint during stance phase decreased. 5. The maximum plantar-flexion angle during swing phase of the ankle joint decreased. 6. The maximum dorsi-flexion of the ankle joint occurred in the ealier stage of the stance phase. 7. The maximum extension of the knee joint occurred in the later stage of the stance phase. 8. The maximum dorsi-flexion moment increased. 9. The maximum plantar-flexion moment increased. 10. The maximum dorsi-flexion moment occurred later. 11. The activities of dorsi-flexor muscles and plantar-flexor muscles decreased. However, hemiplegic gait showed extremely large variation depending on the subject. Thus, a case where the ankle joint moment increased with the AFO, a case where the ankle joint moment decreased with the AFO, and cases that showed a change in the knee joint angular pattern are reported and examined in more detail individually.

View full abstract

This paper deals primarily with 6-axis forcemoment such as two shearing forces (F_x, F_y), compression (F_z), two bending moments (M_x, M_y) and a tortional moment (M_z) that are acting on the stump of the during walking. An above/knee amputee walked while wearing his A/K prosthesis, which is equipped with the 6-axis force-moment sensor between the knee joint of the prosthesis and the socket, and the above-mentioned forces and moments were measured. As a result of the walking experiments, the values of F_x, F_y, F_z, M_x, M_y and M_z during walking were determined. In the case of fast walking, the peak of F_x and F_z at the beginning of the stance phase is bigger that those in the case of slow walking. This shearing force F_x prevents sudden flexion of theknee joint. Then two application studies of 6-axis forces and moments are described. The first study was concerned with the invention of an evaluative method for alignment of A/K socket with the above-mentioned 6-axis forces and moments. The toe of the A/K prosthesis is set inside (toe-in) or outside (toe-out) about 10°. An A/K amputee walked, and the 6-axis forces and moments were measured. As a result of the walking experiments, the differences between the adjusted alignment and the above-mentioned setting are identified, and the following is turned out. In the case of setting toe-in, it is difficult for the amputee to walk smoothly because stability in the side direction decreases. The second study is about an application of the forces and moments to controlling signals for ascending a staircase. The shearing force (F_x) and the compression (F_z) are utilized as triggers for extending or flexing the knee joint, respectively. An amputee wearing the A/K prosthesis which we had developped could walk on a staircase with a natural posture without the help of his arms. These results of the two application studies confirm the usefulness of the sensing system which is proposed in this study.

View full abstract

The objects in this study are the physically handicapped of C4, C5 and C6 levels. A robotic aid system has been developed to support highlevel quadriplegics who possess little or no manipulative function. Many handicapped persons are youths in the prime of life suffering damage from vertebrae cervicales on account of traffic accidents or sports. This study has established a system of technical support by robotic aid for young highlevel quadriplegics. The utility form of robotic aid enables the handicapped to turn over pages or play games using a bookshelf and a personal computer on a desk. The robot is utilized in the deskwork. Then works in different conditions, such as the cosmetic motions of combing the hair in a lavatory and shaving, are performed by the robot mounted on a wheelchair. We thought that a robotic aid system was needed in both partial help and nursing care at home, and proceeded with the following aims. (1) The robot can easily be mounted on the wheelchair. (2) The robot can pick up things on the floor by arranging a rotational mechanism with one degree of freedom on the wheelchair. (3) The robot can be placed on a desk or working table as its weight makes it portable with the assistance of an attendant (about 14 kgf). (4) When the robot is mounted on the wheelchair, it is long enough to reach the head of the handicapped person. (5) As a gripper, a cosmetic electrical prosthesis is utilized which can easily be changed for an electrical prosthesis of hook type. (6) The control system is stored in the robotic body and can easily correspond to the movements. (7) The robot can apply a 3-joint linkage mechanism, as it has a stable motion and rigid body when stretching arms horizontally. (8) The motor software servo is limited to the setting torque for the self-safety. (9) A voice device is utilized for the conversation between robot and the handicapped, which especially abstracts registered words for a synthetic voice. We have mainly developed keeping aims (1) to (8) in mind, because the robot for partially helping can be easily used at home.

View full abstract