The Feasibility of Wearable Sensors for the Automation of Distal Upper Extremity Ergonomic Assessment Tools

Abstract

Work-related distal upper limb musculoskeletal disorders are costly conditions that many companies and researchers spend significant resources on preventing. Ergonomic assessments evaluate the risk of developing a work-related musculoskeletal disorder (WMSD) by quantifying variables such as the force, repetition, and posture (among others) that the task requires. Accurate and objective measurements of force and posture are challenging due to equipment and location constraints. Wearable sensors like the Delsys Trigno Quattro combine inertial measurement units (IMUs) and surface electromyography to solve collection difficulties. The purpose of this work was to evaluate the joint angle estimation of IMUs and the relationship between sEMG and overall task intensity throughout a controlled wrist motion. Using a 3 degrees-of-freedom wrist manipulandum, the feasibility of a small, lightweight wearable was evaluated to collect accurate wrist flexion and extension angles and to use sEMG to quantify task intensity. The task was a repeated 95º arc in flexion/ extension with six combinations of wrist torques and grip requirements. The mean wrist angle difference (throughout the range of motion) between the WristBot and the IMU of 1.70° was not significant (p= 0.057); but significant differences existed throughout the range of motion. The largest difference between the IMU and the WristBot was 10.7° at 40° extension; this discrepancy is smaller than typical visual inspection joint angle estimate errors by ergonomists of 15.6°. All sEMG metrics (flexor muscle root mean square (RMS), extensor muscle RMS, mean RMS, integrated sEMG (iEMG), physiological cross-sectional area weighted RMS) and ratings of perceived exertion (RPE) had significant regression results with the task intensity. Variance in RPE was better explained by task intensity than the best sEMG metric (iEMG) with R2 values of 0.35 and 0.21, respectively. Wearable sensors can be used in occupational settings to increase the accuracy of postural assessments; additional research is required on relationships between sEMG and task intensity to be used effectively in ergonomics. There is potential for sEMG to be a powerful tool; however, the dynamic nature and combined exertion (grip and flexion/ extension) make it difficult to quantify task intensity