Selecting Muscles for Detection of Upper-Limb Compensatory Movements Using s-EMG Sensors

Patients with upper-limb injuries often use compensatory movements to overcome limitations in range of motion, which can lead to additional injury if not corrected early within a rehabilitation program. Although automatic detection of compensatory movements has been studied in the literature, the impact of sensor locations on detection performance has not been previously explored. To investigate how sensor locations affect the ability to automatically detect compensatory movements of the upper limb, sixteen surface electromyography sensors were placed on key muscles involved in these movements. Thirty-one healthy participants performed a door-opening task in three conditions: without elbow restrictions (healthy pattern), and two conditions with limited elbow range of motion (60° of flexion-full flexion and 30°–80° of flexion to simulate injury). Statistical analyses identified sensor locations with significant differences between the conditions. Support vector machine classifiers demonstrated notably higher performance using data from six sensors on the middle deltoid, the upper trapezius, the latissimus dorsi, the external obliques, and the erector abdominis. This study highlights the importance of thoughtful muscle selection for effective automatic detection and correction of upper-limb compensatory movements, which is crucial for a wearable mechatronic device to be effective in improving the movement quality of patients.