Resolution of Superpositions in EMG Signals using Belief Propagation: Results for the Known Constituent Problem

Abstract — The problem of resolving superpositions in elec-tromyographic (EMG) signals is considered. The shapes of the motor unit action potentials that make up each superposition are assumed to be known a-priori (known constituent problem). Two different and novel belief propagation algorithms have been developed to solve this problem. These algorithms and simulation results are presented in this paper. Index Terms — EMG signal decomposition, factor graphs, belief propagation I