Analysis of Fatigue-Induced Compensatory Movements in Bicep Curls: Gaining Insights for the Deployment of Wearable Sensors

TL;DR

This study investigates fatigue-induced compensatory movements during bicep curls to guide wearable sensor deployment for home-based rehabilitation. By recording eight upper-limb sEMG signals and joint kinematics from 12 healthy subjects across weight-free, standard, and fatigue conditions, the authors extract two muscle synergies using NNMF and analyze relative muscle contributions, activation patterns, and kinematics. The results reveal a fatigue-driven shift from forearm to shoulder muscle contributions and a substantial rise in activation amplitudes, with more pronounced shoulder motion under fatigue. These findings offer concrete targets for sensor placement—particularly around the shoulder and upper trapezius—and establish quantitative benchmarks (e.g., $VAF \ge 0.9$ with growth $< 3\%$, RMS increases up to $127\%$, and shoulder ROM changes) to detect fatigue-induced compensations in home-rehabilitation settings.

Abstract

A common challenge in Bicep Curls rehabilitation is muscle compensation, where patients adopt alternative movement patterns when the primary muscle group cannot act due to injury or fatigue, significantly decreasing the effectiveness of rehabilitation efforts. The problem is exacerbated by the growing trend toward transitioning from in-clinic to home-based rehabilitation, where constant monitoring and correction by physiotherapists are limited. Developing wearable sensors capable of detecting muscle compensation becomes crucial to address this challenge. This study aims to gain insights into the optimal deployment of wearable sensors through a comprehensive study of muscle compensation in Bicep Curls. We collect upper limb joint kinematics and surface electromyography signals (sEMG) from eight muscles in 12 healthy subjects during standard and fatigue stages. Two muscle synergies are derived from sEMG signals and are analyzed comprehensively along with joint kinematics. Our findings reveal a shift in the relative contribution of forearm muscles to shoulder muscles, accompanied by a significant increase in activation amplitude for both synergies. Additionally, more pronounced movement was observed at the shoulder joint during fatigue. These results suggest focusing on the shoulder muscle activities and joint motions when deploying wearable sensors to effectively detect compensatory movements.

Figures (7)

• Figure 1: Experimental scene with VICON motion capture cameras.
• Figure 2: Four joint angles during bicep curl in this study. (a) Sagittal view. (b) Coronal view. The red dotted lines represent the references; the purple line links the clavicle and the acromioclavicular joint; the blue, green, and yellow lines depict the upper arm limb, the lower arm limb, and the wrist, respectively. The arrows indicate the positive directions.
• Figure 3: Placement of sEMG sensors and VICON markers on subjects. (a) sEMG sensors (b) VICON markers
• Figure 4: The fatigue metrics. (a) The averaged median frequency of the subjects. (b) The averaged RMS amplitude of the subjects. The blue bars and error bars show the average median frequency and RMS amplitude of the sEMG collected under standard conditions. In contrast, the green bars and error bars represent the data under fatigued conditions.
• Figure 5: Averaged VAF values of different numbers of synergy for different stages. The orange graph represents the averaged VAF values for weight-free conditions; the blue graph represents the averaged VAF values for standard conditions, and the green graph represents the averaged VAF values for fatigue conditions.