An Algorithm for the Estimation of the Signal-To-Noise Ratio in Surface Myoelectric Signals Generated During Cyclic Movements

In many applications requiring the study of the surface myoelectric signal (SMES) acquired in dynamic conditions, it is essential to have a quantitative evaluation of the quality of the collected signals. When the activation pattern of a muscle has to be obtained by means of single- or double-threshold statistical detectors, the background noise level e_noise of the signal is a necessary input parameter. Moreover, the detection strategy of double-threshold detectors may be properly tuned when the SNR and the duty cycle (DC) of the signal are known. The aim of this paper is to present an algorithm for the estimation of e_noise, SNR, and DC of an SMES collected during cyclic movements. The algorithm is validated on synthetic signals with statistical properties similar to those of SMES, as well as on more than 100 real signals.