Estimating forces at multiple degrees of freedom from surface EMG using non-negative matrix factorization for myoelectric control

A novel signal processing approach for surface electromyogram (EMG) is proposed to estimate the forces produced at multiple degree of freedom (DOF) simultaneously during dynamic voluntary contractions. The method is based on a generative model of surface EMG, and the estimation algorithm is a modified non-negative matrix factorization (NFM). The proposed method has potential to provide simultaneous and proportional control information for multi-function powered prosthesis.