Kenta Nomura

Development and performance evaluation of parallel link type human ankle rehabilitation assistive device

This paper presents a novel rehabilitation assistive device called a “parallel link type human ankle rehabilitation assistive device (PHARAD).” It can accurately measure foot motions in six degrees of freedom and reproduce the ankle joint motions. By reproducing the input motions of the ankle joint, PHARAD conducts passive exercises for ankle rehabilitation. To measure and reproduce complex foot motions, we adopted a parallel link mechanism using six pneumatic cylinders with displacement sensors. In this research, we define the motions of a foot plate attached to a foot sole as the foot motions. A posture of the foot plate, i.e., the three-dimensional (3D) position (x, y, z) and rotation angle (θ, φ, ψ), is numerically calculated by solving the forward kinematics of the PHARAD. We conducted two kinds of experiments to evaluate the performance of the PHARAD. First, verification experiments for the accuracy were implemented by comparing the measured motions of the foot plate by the PHARAD with those obtained by a motion capture system. The experimental results showed that the maximum root mean square error (RMSE) values of the 3D position and rotation angle were 2.6 mm and 1.5°, respectively. Then, verification experiments for the reproducibility were implemented by comparing the reproduced motions with the input motions. The experimental results showed that the RMSE values of the 3D position and rotation angle were 5.6 mm and 6.1°, respectively. Moreover, after reproducing the input motions ten times, the standard deviations of the 3D position and rotation angle were 1.4 mm and 0.7°, respectively. These experimental results show that the PHARAD can precisely measure and reproduce complex ankle motions, and has the potential to reproduce the exercise therapy presented by physical therapists.

Development of Stewart platform type ankle-foot device for trip prevention support

This paper presents an ankle-foot device using a Stewart platform, which is a type of parallel-link mechanism, for trip prevention support. The developed device can reproduce the input motions of the ankle joint in six degrees of freedom by controlling six pneumatic cylinders at the same time. The root mean square errors of the 3-D position and rotation angle of the reproduced motions with the input motions (dorsiflexion and plantar flexion) were 6.3 mm and 3.0°, respectively. Verification experiments for trip prevention support performance were conducted by comparing motions in each walking condition measured using a motion capture system. The experimental results showed that the minimum foot clearance during mid-swing and initial swing increased significantly by the trip prevention support offered by the developed device. The developed device can perform passive exercises for ankle rehabilitation and support walking for trip prevention.

Measurement of the passive stiffness of ankle joint in 3 DOF using stewart platform type ankle foot device

This paper presents a method to measure the passive stiffness of an ankle joint in three degrees of freedom (DOF) under two motion speeds (1 Hz and 5 degree/s) using a developed Stewart platform-type device. The developed device can reproduce input motions of the foot in 6 DOF by controlling six pneumatic linear motion actuators. We used the device to measure the passive stiffness of an ankle joint undergoing three kinds of motion, namely dorsi-plantar flexion, inversion-eversion, and adduction-abduction. The measured values of the passive stiffness of the ankle joint in dorsiflexion that we obtained agreed well with that obtained in a previous study, indicating that the developed device is useful for measuring the passive stiffness of ankle joint. In addition, the developed device can be used to measure the stiffness in inversion-eversion and adduction-abduction motions as well, parameters that have never been measured. The results we obtained demonstrated certain interesting features as we varied both the direction and pace of motion (e.g., there were significant differences in the stiffness not only between adduction and abduction during the faster pace, but also between these and the other motions).This paper presents a method to measure the passive stiffness of an ankle joint in three degrees of freedom (DOF) under two motion speeds (1 Hz and 5 degree/s) using a developed Stewart platform-type device. The developed device can reproduce input motions of the foot in 6 DOF by controlling six pneumatic linear motion actuators. We used the device to measure the passive stiffness of an ankle joint undergoing three kinds of motion, namely dorsi-plantar flexion, inversion-eversion, and adduction-abduction. The measured values of the passive stiffness of the ankle joint in dorsiflexion that we obtained agreed well with that obtained in a previous study, indicating that the developed device is useful for measuring the passive stiffness of ankle joint. In addition, the developed device can be used to measure the stiffness in inversion-eversion and adduction-abduction motions as well, parameters that have never been measured. The results we obtained demonstrated certain interesting features as we varied both the direction and pace of motion (e.g., there were significant differences in the stiffness not only between adduction and abduction during the faster pace, but also between these and the other motions).

Evaluation of venous return in lower limb by passive ankle exercise performed by PHARAD.

This paper presents evaluation of venous return, i.e., blood flow volume of vein (BF), in the lower limb after passive exercise performed by our developed “parallel link type human ankle rehabilitation assistive device (PHARAD)”. The PHARAD can perform complex passive exercises (plantar flexion/dorsiflexion, inversion/eversion, adduction/abduction, and combination of these motions) by reproducing input motions of a foot plate that is attached to a sole of foot. The passive exercise can be performed for not only rehabilitation but also prevention of deep vein thrombosis (DVT). In this study, we measured the concentration of Total hemoglobin (Total-Hb) using multi-channel near infra-red spectroscopy (NIRS)-based tissue oximeters and calculated a gradient of Total-Hb during a venous occlusion. We defined the gradient as BF and evaluated BF after 3 min passive exercise performed by the PHARAD comparing to BF of resting. Seven healthy young adult people were recruited for the experiment and we assessed passive exercise, active exercise, and walking. Experimental results show that BF after the passive exercises significantly increases compare to BF of resting and this indicates that passive exercises performed by the PHARAD increases BF and has a potential to prevent DVT.

THREE-DIMENSIONAL POSTURE ESTIMATION OF FOOT BONES BY USING PLANTAR PLATE

Purpose: This paper proposes a method to easily and quantitatively estimate the changes in the foot bone three-dimensional (3D) posture from the 3D posture of a plantar plate without using X-ray or computed tomography (CT). Methods: The estimation functions from the posture of the plantar plate attached to the sole of a foot to the posture of the each bone are calculated using multiple regression analysis (MRA). Because we assumed that the posture of the plantar plate is related to each bone posture. Each bone posture can be estimated by substituting the plantar plate posture into the estimation function. Results: The adjusted coefficient of determination of the linear regression model (estimation function) of more than 90% was obtained by the estimation function, which was higher than 0.70. The estimation accuracy root mean square error (RMSE) of the translation and rotation were approximately within ±2.8mm and ±2.9∘, respectively. The RMSE/range of motion (RoM) values of the translation and rotation were approximately within ±22.1% and ±7.1%, respectively. Conclusion: The experimental results suggest that the 3D posture of almost all types of foot bones can be easily estimated using plantar plate posture and the linear regression model. This is an inexpensive, easy-to-apply method that can perform real-time measurement.