An erythrocyte can be deformed very easily by even very small amounts of stress, enabling it to flow smoothly through approximately 3 µm diameter microvessels despite its larger size of 8 µm in diameter and 2 µm in thickness. Reduction of erythrocyte deformability leads to poor circulation particularly in microvessels, which may result in hypertension and various circulatory diseases. To clarify blood microcirculatory properties, it is very important to understand the deformability of erythrocytes. However, a standard method for quantitative evaluation of erythrocyte deformability has not yet been established. In this study, an apparent Young's modulus and a time constant for the recovery of erythrocyte shape following uniaxial stretching are proposed as deformability indices. When an erythrocyte was pulled horizontally on opposite ends using two micropipettes, tensile stress and strain were measured, and an apparent Young's modulus was calculated based on Hooke's law. On returning to its initial shape after the erythrocyte was released from stretching, the elapsed time and strain were measured, and a time constant for shape recovery was calculated based on the Kelvin model. The average apparent Young's modulus was 16.0Pa, and the average time constant was 113.4ms, with a slight tendency toward an inverse relation between the two physical quantities.

抄録全体を表示

Muscles do not only generate contractile force but also behave like a spring. This spring-like property, or elasticity, has been studied by applying perturbation to the muscle and measuring its responses such as the force and the evoked mechano-myogram (MMG). Muscle elasticity increases as the muscle activity increases. The system identification technique has been applied to an evoked MMG, but the evoked MMG was measured in the resting state. Many motor units are activated synchronously or asynchronously in a voluntary contraction. Forces generated by the motor units fuse. As a pilot study, the muscle elasticity in a twitch contraction was investigated. The purpose of this study was to identify the mechanical property of the muscle in a twitch contraction, using the system identification technique applied to MMGs evoked by single- and double-pulse stimulations. In double-pulse stimulation, the first pulse was applied to stimulate the tibialis anterior muscle, and before the muscle returned to the resting state, the second pulse was applied to the common peroneal nerve. We assumed that the MMG signal evoked by the second pulse stimulation reflects the viscoelasticity of the muscle in a twitch contraction. The MMG system was identified, then the natural frequency of the transfer function was calculated as an index of elasticity. The MMG system with single-pulse stimulation was approximated with a third-order model, and its undamped natural frequency was 4.0±0.4 Hz. On the other hand, the MMG system with double-pulse stimulation was approximated with a fourth-order model, and its high natural frequency was 8.2±0.5 Hz. The increase in natural frequency may be caused by increased elasticity in an activated muscle. These results suggest that the proposed method is a novel technique for estimating the elasticity of a muscle in twitch contraction.

抄録全体を表示

In this study, we investigated the possibility of extending our equilibrium point control model by functional electrical stimulation (FES). While FES has been studied extensively in an isometric environment, the goal of the present work is to extend it to an unconstrained environment. Our model uses the electrical agonist-antagonist muscle (EAA) ratio, which is closely related to the joint angle corresponding to the equilibrium point. In addition, EAA activity has a close relationship with joint stiffness. We represented FES control by a model that has a neuromuscular system and a musculoskeletal system. We verified whether our model is reasonable through analysis of the frequency characteristics. We modeled the musculoskeletal system for human elbow joint movement, and verified our model by studying the frequency characteristics between input and output. We assumed that the EAA ratio is an input while the elbow joint angle is an output. We showed that by coupling the model of neuromuscular system and the model of musculoskeletal system, we were able to explain the frequency characteristic of the total model. From these results, we showed the validity of our model and indicated the possibility of achieving FES control under an unconstrained environment.

抄録全体を表示

Microbubbles are widely used as contrast agents in ultrasound diagnosis. Microbubbles may also has therapeutic uses in the heat amplification of high-intensity focused ultrasound ablation or as carriers of acoustic targeted drug/gene delivery therapy. However, microbubbles injected into a blood vessel are diffused throughout the whole body;therefore, their efficiency is still limited. If microbubbles could be controlled in vivo, their efficiency and efficacy would be significantly improved. To address this issue, we have proposed a technique that controls microbubble behavior in blood vessels using ultrasound emitted from the body surface. To apply the technique in vivo, robotic ultrasound transducer positioning on body surface is required. For this purpose, we have developed a robotic system and confirmed that microbubble can be manipulated by the system. In more practical condition, focal length of an ultrasound transducer has to be considered. To address the issue, we propose a control system considering the focal length in this study. The system consists of a parallel-link robot for ultrasound transducer positioning, a robot controller, and an optical tracking device. The robot has three arms, and a transducer holder, and a six-axis force sensor. The robot controller generates ultrasound emission plans using body surface position measured by the tracking device, and manipulate the robot. As for validation of the system, we performed following experiments;1) positioning accuracy evaluation without contact, 2) evaluation of contact forces control, and 3) in vitro ultrasound emission tests. From the first experiment, positioning accuracy was less than 1mm. As for the contact force control validation, the system could keep required reaction force for ultrasound emission on a phantom surface within 1.5mm errors. In the third experiment, the errors in the perpendicular direction of the ultrasound axis and the direction of the axis were 0.71mm and 5.52mm, respectively. From the results, we confirmed that the system could emit ultrasound to a target by using a hydrophone in a poly(ethylene glycol) monomethacrylate (PEGMA) phantom. Consequently, the results demonstrated that the proposed system could generate appropriate plan and manipulate an ultrasound transducer on body surface considering contact condition with body surface.

抄録全体を表示

Walking and balance tests are used for evaluating the effects of rehabilitation and the time taken to perform the test. The Four Square Step Test (FSST) is a useful test for evaluating balance in patients with stroke and orthopedic diseases. For the FSST, 4 bars are used to divide a floor surface into 4 areas. Subjects have to step from one area into the neighboring area in a sequence, and their movements and stability are evaluated. However, no detailed evaluation has been conducted on the transition of the forward, backward, right, and left movement phases. This study aimed at detailed evaluation of the FSST in post-stroke hemiplegic patients and elderly subjects, using wearable motion sensors comprising an accelerometer and angular velocity sensors capable of identifying the direction of the subjects' movements. Twelve post-stroke hemiplegic patients (6 left hemiplegic patients, 6 right hemiplegic patients; Brunnstrom stage of the lower limb: V) and 6 normal elderly subjects were studied. The sensor was attached to the waist and both thighs of a subject. Then FSST was administered and a video was recorded simultaneously. Using the angular velocity signals of the thigh, the four movement phases in FSST were classified. The following items were analyzed: root mean square (RMS) acceleration, RMS angular velocity, amplitude of the thigh angle, and amplitude/performing time. The time measured by the wearable motion sensors correlated with the time recorded by the monitoring video (r = 0.88, p < 0.05). The total performing time did not differ significantly between normal elderly subjects and hemiplegic patients. The RMS acceleration and RMS angular velocity for the waist were significantly different between normal elderly subjects and hemiplegic patients (p < 0.05). Furthermore, the changes in the thigh angle of patients when they stepped over the bars while performing the FSST were different between hemiplegic patients and normal elderly subjects. We conclude that the FSST using wearable motion sensors is useful for the evaluation of balance in post-stroke hemiplegic patients.

抄録全体を表示

Picture books are one of the elements of childcare that are indispensable for emotional development. The reading of picture books to children or grandchildren by their parents or grandparents plays an important role in enhancing their imaginative power. However, visually impaired parents often experience trouble in communicating the contents and details of picture books to their children, since they have difficulties not only in reading the stories but also in describing the pictures. In this study, we developed a storytelling assistive system to use with picture books for the visually impaired, which is based on a tablet terminal using speech sounds.

抄録全体を表示

People with upper limb disabilities often need to use auxiliary equipment or alternative input methods to operate personal computers. Instead of using a traditional mouse, they may employ methods that use a numeric keypad supported by a function of MouseKeys, manipulate a joystick or trackball using their residual abilities, or use a mouth stick or head stick to control the mouse. Each of these methods requires a lot of time and effort. In this report, the authors propose a new pointing method that uses a high-precision visual marker (ArrayMark) developed in the robotic field. The marker uses a microlens array and overcomes the primary problem confronted by conventional markers: reduced pose-estimation precision in frontal observations. A user mounts the marker on the head and turns the face toward the display. The proposed system estimates head position and pose, and synchronizes a cursor with facial direction. In precision experiments, when an able-bodied subject fixed his face toward a target for 3.3 seconds at a distance of 700 mm, the standard deviation error for the cursor coordinates was within 1.7 mm. The times required for pointing when using the proposed method, mouse, and MouseKeys were compared. The proposed method took 1.8 times more time than the mouse and 0.36 times less time than the MouseKeys.

抄録全体を表示