Towards Understanding Of Human Cervical Spine With EMG Signal

Electromyographic signals (EMG) are of great significance to the recognition of cervical motion intention. However, due to the limitations of existing sensing hardware and the inherent noise from the cervical muscle movement, how to obtain accurate EMG signals to precisely identify and analyze human cervical vertebra status remains an open issue. In this paper, we propose a method for human cervical vertebra state recognition using EMG signal. First, we used OpenSim simulation to select the locations for EMG sensor placement, which can provide more informative signal for sequential tasks. Secondly, the EMG signal is pre-processed by notch filtering and wavelet decomposition, and the relevant features are extracted by time-frequency domain method. Then we construct three classification models to identify motion intention and fatigue detection of neck with the data after dimensionality reduction. A large number of experiments and simulations demonstrate the effectiveness of our method.