Analysis of Respiratory EMG Signals for Cough Prediction

Respiratory muscle weakness can be one of the aftereffects due to cervical or high thoracic spinal cord injury (SCI). This problem affects the organism's capacity to maintain the correct levels of oxygen, and the cough events start to be difficult to perform. Studies have demonstrated that the application of synchronized Transcutaneous Functional Electrical Stimulation (TFES) to the respiratory muscles, diaphragm and abdomen, enhanced their resistance and strength. So, this study aims to evaluate if electromyography (EMG) sensors positioned on the sternocleidomastoid (SCM), omohyoid (OM), and external oblique (EO) muscles can be used to help in automatic cough prediction for a future TFES synchronization system. Voluntary coughs of nineteen volunteers were recorded using three EMG sensors and their potential use in cough prediction was evaluated. Results show that EO muscle presents better results (55 ± 31.37 V) in amplitude variations compared to the other ones. However, it is still not possible to standardize one single muscle for prediction due to the variation of activation in each person. Also, the use of EMG sensors can contribute significantly to cough prediction systems considering the muscle response can be perceived well before the sound.