Identificación de Patrones en Tiempo y Frecuencia de Señales sEMG Usando la Transformada de Hilbert-Huang
(Time and Frequency Patterns Identification of sEMG Signals Using Hilbert-Huang Transform)
Alvaro Altamirano (firstname.lastname@example.org)1, Arturo Vera (email@example.com)1, Roberto Muñoz (firstname.lastname@example.org)1, Lorenzo Leija (email@example.com)1, Didier Wolf (firstname.lastname@example.org)2
1Centro de Investigación y de Estudios Avanzados del Instituto Politécnico Nacional2Centre de Recherche en Automatique de Nancy, Université de Lorraine CNRS UMR 7039
This paper appears in: Revista IEEE América Latina
Publication Date: Oct. 2017
Volume: 15, Issue: 10
This article reports the identification of two groups of patterns in time and frequency, also a muscular intensity characteristic associated to hand movement per user, all this results were obtained by superficial myoelectric signal analysis using Hilbert-Huang transform. Muscular signals were acquired from five muscles of the right forearm from five healthy subjects. Using a four channels acquisition system with differential configuration myoelectric signals were recorded for six movements of the fingers in flexion and extension. Electrodes were placed over five forearm anterior, posterior and finger linked muscles. Time pattern corresponds to sinusoidal oscillations with average length of 24.5 ms, corresponding to voluntary contraction and relaxation of the muscles. These oscillations are present into all acquisition channels and there is evidence about groups of two consecutive oscillations with a range of 30 ms between them. Also, there found a pattern of three frequencies: 83.3 Hz, 96.7 Hz and 113.3 Hz, present into contraction and relaxation timeslot. Using Hilbert transform were identified the instantaneous frequencies that indicate the state of the muscles from doss to active and vice versa. There is an intensity characteristic identified in the recordings about behavior of muscles per movement for each subject being repetitive and unique. These processes could decrease time computing to obtain intrinsic characteristics for myoelectric signals.
myoelectric signals, Hilbert-Huang transform, patterns, multichannel signals, muscular intensity
Documents that cite this
This function is not implemented yet.
[PDF Full-Text (640)]