In this study, we investigated how the movements of upper torso and upper extremity joints (shoulder, elbow, and wrist) contribute to a ball velocity during overhand throwing. Ten healthy young males threw a rubber baseball towards a target. At first, they performed overhand throwing with maximal effort. Then, they pitched a ball at 80%, 60%, 40% and 20% of the maximal ball velocity in this order. We recorded their throwing motions using three high-speed digital video cameras (200 fps), and calculated angular velocities of wrist palmar flexion, elbow extension, shoulder horizontal adduction, shoulder internal rotation, and upper torso left rotation from three-dimensional coordinate data obtained by the DLT (Direct Linear Transformation) method. Based on these data, we computed the relative contributions of joint movement to ball velocity using cross products between each joint angular velocity vector and relative position vector from each joint to the center of the ball. As relative ball velocity increased, the angular velocities of wrist palmar flexion, elbow extension, and upper torso left rotation at the moment of peak ball acceleration increased (F(4, 36)= 12.1, p < 0.001 ; F(1.4, 12.6)= 28.0, p < 0.001 ; F(1.7, 15.2)= 46.6, p < 0.001, respectively), but the contribution of the angular velocity to ball velocity only increased in upper torso left rotation (F(4, 36)= 6.5, p < 0.001). At instance of ball release, although angular velocity of wrist palmar flexion, elbow extension, shoulder internal rotation, and upper torso left rotation increased with increasing relative ball velocity (F(1.5, 13.4)= 12.3 ; p < 0.01 ; F(1.5, 13.2)= 19.8, p < 0.001; F(1.4, 12.6)= 28.0, p < 0.001; F(4, 36)= 4.9, p < 0.01, respectively), the contribution of the angular velocity to ball velocity only increased in shoulder internal rotation (F(2.0, 17.5)= 7.1, p < 0.01). These results suggest that the human body mainly modulates angular velocities of upper torso left rotation and shoulder internal rotation to control ball velocity during overhand throwing.

View full abstract

For mental stress assessments, we conduct basic research to determine the relationships between stress and time-variant biosignals fluctuation triggered by visual tasks (named trigger task). Our goal is to devise a suitable and quantitative index of mental stress for individual users based on an understanding of the fluctuation of time-variant signals of heart rate and respiration with their dependence on the trigger task. Experiments elucidate the relationships between subjective scores of mental stress (GHQ28) and an objective index derived from two biosignals, RR intervals of ECG and respiration signals, under a trigger task. The experiments use a visual search task and a maze task as the trigger task because the stress raised by these tasks can be controlled. The results show, except for a few non-sensitive subjects, a positive correlation between subjective score and biosignal index, which is LF/HF ratio of PCA score from both biosignals with small user dependency. We also clarify that the indices of the subjects whose GHQ28 score is under half are sensitive (in terms of stress) to the trigger task used. This paper demonstrates the possibility of quantitatively measuring mental stress by using the proposed biosignal index. In addition, for accurate stress measurement it is necessary that the trigger task suit the individual.

View full abstract

In this research, we developed a novel ankle-foot assistive device for rehabilitation. This device uses a Stewart Platform Mechanism (SPM) to measure and assist the movements of a human ankle joint in six Degrees Of Freedom(DOFs). The SPM adapts to the displacement of the rotation axis of a human ankle joint during the movements of a human foot. In our previous work, we have proposed methods of measuring and controlling the movements of a human foot. In this paper, we propose a method of a force control of the device. The force can be controlled by changing the pneumatic pressures in two chambers of each cylinder without using any force sensors. The force control method is tested in experiments for one cylinder and for the developed six DOFs assistive device. These test results show that the mean squared error of one cylinder is 0.19[Nm] or less, respectively. The mean squared error about the force and torque control of this device is 0.49[N] and 0.06[Nm] or less, respectively. The experimental results showed that desired force can also be realized in a high enough accuracy using the proposed method. For stiffness and water rehabilitation, we have proposed methods of controlling stiffness rate and viscous damping coefficient of one cylinder. The mean squared error of stiffness controlling is 3.4[N/m] or less, and that of viscous damping coefficient controlling is 1.4[Ns/m] or less. These results suggest that the developed six DOFs assistive device could be used during ankle-foot rehabilitation without force sensor.

View full abstract

Calibration for calculating the external parameter of camera-projector system is important, to make high-precision 3-D measurement by 3-D endoscope using the grid active stereo. In this paper, we propose a new calibration method with a sphere calibration object in a geometrically flexible camera-projector system, and we examine the validity of the proposed method. As the result of comparing using conventional calibration with a plane calibration object, it is clarified that the proposed method resolves the scaling imprecision, that is a problem of the conventional method. Thus, the proposed method has higher potential of measurement accuracy. In addition, it is the advantage of the proposed method that precision estimation of calibration parameters with a low number of input images is possible.

View full abstract

Wearable technology for measuring biomedical signal and momentum has been advanced for the purpose of physiological monitoring system during daily life. In measuring ground reaction force, a method to estimate plantar pressure from output of compact force sensors on the underside of footwear has been used. However, compact sensors used in this method are expensive, it is hard to use these sensors in daily monitoring of sports activity and rehabilitation. In this paper, the authors adopted low-priced film force sensors instead of the compact force sensors, and measured load on plantar and made it possible to estimate vertical ground reaction force in walking. In addition, the authors confirmed that this monitoring system can be used in different pace and footwear.

View full abstract

In recent years, the demand for more sophisticated medical technologies has prompted the development of many advanced medical equipment, sensors, etc. In particular, implantable sensors are superior to other measurement methods in terms of sensitivity, resolution, precision, etc. and are expected for further technological development in the future. We are currently creating nerve-regenerative biochips that are specifically targeted for application to brain-machine interface (BMI), for which high mechanical adaptability, medical functionality and advanced biocompatibility are all essential. Carbon nanotube (CNT) is a representative material used in nanotechnology, application of which to implantable sensors appears promising due to its high conductivity and outstanding mechanical strength. However, there is a strong concern about the fact that there is little knowledge about its toxicity to living organisms and the environment, or interaction among them. Therefore, in this research, we have embedded the CNTs, which we actually use in creating our devices, subcutaneously in rats and examined the effect of functional group-added CNTs, which were formed by the plasma activation method (P-A method), to the living organisms. As for the evaluation method, in addition to weight measurement, leukocyte estimation test, and observation of change in organizational form, semi-quantitative analysis of expressed genes using molecular biological method was performed to assess the existence/non-existence of inflammatory reaction and the biocompatibility under each set of conditions. Consequently, it was suggested that, by applying the P-A process, the biocompatibility improved compared to the CNTs unprocessed by the P-A method. This result indicates the possibility that CNTs, to which functional group modification is applied, may be directly implanted in living organisms for a relatively long period of time.

View full abstract