Evaluation of s-EMG Sensor Locations for Upper-Limb Compensatory Movement Detection

Area of Science:

• Biomedical Engineering
• Rehabilitation Science
• Wearable Technology

Background:

• Musculoskeletal rehabilitation is vital for upper limb recovery post-injury or stroke.
• Unsupervised rehabilitation can lead to compensatory movements, hindering patient progress.
• Current wearable devices may not adequately assess movement quality, missing compensatory patterns.

Purpose of the Study:

• To determine optimal surface electromyography (sEMG) sensor locations for detecting compensatory movements in upper limb rehabilitation.
• To enhance the design of wearable systems for more effective home-based rehabilitation monitoring.

Main Methods:

• Collected sEMG data from 40 healthy individuals performing simulated healthy and impaired movements.
• Utilized statistical analysis and feature importance to identify sensor combinations differentiating movement patterns.
• Evaluated classification performance of identified sensor placements.

Main Results:

• Eleven sEMG sensors on specific upper body muscles (trapezius, deltoids, biceps, triceps, latissimus dorsi, erector spinae, rectus abdominis, external oblique) achieved 81.43% accuracy and 0.8549 F1 score.
• Sensor count can be reduced to seven with minimal impact on overall accuracy and F1 score.
• Specific tasks may show performance degradation with fewer sensors.

Conclusions:

• Optimal sEMG sensor placement is crucial for wearable devices to detect compensatory movements.
• A reduced sensor set of seven can maintain detection efficacy, improving wearability and reducing demands.
• Findings support the development of smarter wearable rehabilitation tools to improve patient outcomes.