Elbow Extension Restoration Surgery in People With Tetraplegia: Evolution of the Muscular Co-activations of the Upper Limb During the Post-operative Rehabilitation

There is a need for a detailed and quantified follow-up during rehabilitation after an elbow extension restoration in people with tetraplegia. Surface or intramuscular electromyography (EMG) gives information on muscular activation, and can detect small but clinically relevant modifications in muscular activation of patients with tetraplegia. To this day, no study measured the evolution of muscular activation in the upper limb of people with tetraplegia who underwent an elbow extension restoration surgery. The objective of this study is to investigate for the first time the reorganization of muscle coordination after surgical restoration of elbow extension through the identification of muscle synergies and the quantification of muscle co-activations. Four participants with tetraplegia will take part to this study. The experimental procedure will be conducted before their surgery and once a month during 6 months after their surgery. The 1h procedure consists of performing consecutive elbow extension-flexion cycles with the shoulder abducted at different angles. Surface and intramuscular EMG measurements will be collected for several upper limb muscles. Kinematic markers will be placed in order to measure the upper limb range of motion and movement velocity. Surface EMG signals will be band-pass filtered (10-450Hz), full wave rectified, and smoothed with a low-pass filter (10Hz). Onset and offset times will be defined through the calculation of an activation threshold. For each session and each muscle, EMG amplitude will be normalized to its peak value across all trials. Each elbow extension-flexion cycle will be interpolated to 100 time points: 0-50% was extension; 50-100% flexion. For each participant and each movement, all combinations of muscle synergies will be identified using non-negative matrix factorization (NMF) from the EMG data and data will then be averaged across the cycles. Co-activations will be directly identified from the EMG data.