Inertial Measurement Unit Based Upper Extremity Motion Characterization for Action Research Arm Test and Activities of Daily Living.

Hyung Seok Nam,Woo Hyung Lee,Yoon Jae Kim,Sungwan Kim,Moon Suk Bang,Han Gil Seo

doi:10.3390/s19081782

Hyung Seok Nam, Woo Hyung Lee + Show 4 more

Open Access

https://doi.org/10.3390/s19081782

Copy DOI

Abstract

In practical rehabilitation robot development, it is imperative to pre-specify the critical workspace to prevent redundant structure. This study aimed to characterize the upper extremity motion during essential activities in daily living. An IMU-based wearable motion capture system was used to access arm movements. Ten healthy subjects performed the Action Research Arm Test (ARAT) and six pre-selected essential daily activities. The Euler angles of the major joints, and acceleration from wrist and hand sensors were acquired and analyzed. The size of the workspace for the ARAT was 0.53 (left-right) × 0.92 (front-back) × 0.89 (up-down) m for the dominant hand. For the daily activities, the workspace size was 0.71 × 0.70 × 0.86 m for the dominant hand, significantly larger than the non-dominant hand (p ≤ 0.011). The average range of motion (RoM) during ARAT was 109.15 ± 18.82° for elbow flexion/extension, 105.23 ± 5.38° for forearm supination/pronation, 91.99 ± 0.98° for shoulder internal/external rotation, and 82.90 ± 22.52° for wrist dorsiflexion/volarflexion, whereas the corresponding range for daily activities were 120.61 ± 23.64°, 128.09 ± 22.04°, 111.56 ± 31.88°, and 113.70 ± 18.26°. The shoulder joint was more abducted and extended during pinching compared to grasping posture (p < 0.001). Reaching from a grasping posture required approximately 70° elbow extension and 36° forearm supination from the initial position. The study results provide an important database for the workspace and RoM for essential arm movements.

Highlights

In the last decade, there have been dramatic improvements in rehabilitation robots and kinematic analyses of the upper extremities
The range of root mean square error (RMSE) for the elbow flexion/extension angle ranged from 2.11◦ to 4.75◦ (3.61 ± 1.32◦ ), and 0.42◦ to 1.22◦ (0.85 ± 0.40◦ ) for wrist dorsiflexion/volarflexion angle
The intra-subject coefficient of variation (CoV) ranged from 5.5% to 9.5%, whereas the inter-subject CoV ranged from 8.4% to 14.3%

Summary

Introduction

There have been dramatic improvements in rehabilitation robots and kinematic analyses of the upper extremities. Many types of multi-axis exoskeletons have been developed, as well as relatively simple end-effector type robots [1,2,3,4,5]. Even in exoskeletons, the angular movements of human and robot joints do not exactly match. This discrepancy comes from the fundamental difference in that exoskeleton joints have mechanical joint axes with their corresponding. The goal of exoskeleton rehabilitation robot development should not focus on perfectly resembling the human arm joint and structure, but rather on designing a modified structure based on a better understanding of human kinematics

Objectives

Methods

Results

Discussion

Conclusion