A new strategy for multifunction myoelectric control

Abstract

This paper describes a novel approach to the control of a multifunction prosthesis based on the classification of myoelectric patterns. It is shown that the myoelectric signal exhibits a deterministic structure during the initial phase of a muscle contraction. Features are extracted from several time segments of the myoelectric signal to preserve pattern structure. These features are then classified using an artificial neural network. The control signals are derived from natural contraction patterns which can be produced reliably with little subject training. The new control scheme increases the number of functions which can be controlled by a single channel of myoelectric signal but does so in a way which does not increase the effort required by the amputee. Results are presented to support this approach.

Nuevo Enfoque para la Clasificación de Señales EEG usando la Varianza de la Diferencia entre las Clases de un Clasificador Bayesiano

“…Cuando se requiera controlar dispositivos, es posible mejorar los resultados de precisión alcanzada si se tiene en cuenta la detección desde el inicio del movimiento hasta 300 ms después del mismo, que es el período de tiempo en el que el usuario no puede percibir el retardo al usar un dispositivo (Hudgins et al, 1993). La ocurrencia de ERD es posible en el intervalo de dos segundos antes del movimiento hasta un segundo después de que el movimiento fue ejecutado (Pfurtscheller and Da Silva, 1999;Ibáñez et al, 2013;Kirchner et al, 2014).…”

Myoelectric control of artificial limb inspired by quantum information processing

“…Particular choice of the location of the electrodes strongly depends on particular amputation and remnant musculature, therefore it is discussed elsewhere. There are four classical features that can be extracted from the raw signals of each electrode channel: mean absolute value (an estimate of the mean absolute value of the signal), zero crossing (the number of times the waveform crosses zero), slope sign change (the number of times the slope of the waveform changes sign) and wave length (the cumulative length of the waveform over the time segment) [16]. Thus, for each of the four features the feature space is of dimension of n. Let one of these features, mean absolute value for example, be encoded into the states of a (discrete) n-dimensional quantum system, so that the component a i |i of the state |ψ = n i=1 a i |i represents the feature extracted from the ith electrode.…”

Vocal frequency estimation and voicing state prediction with surface EMG pattern recognition

“…Since we used 2 channels of sEMG this provided a total 14 features per segment. These features were chosen because of their prior use in the literature to estimate muscle tension [43,44].…”