Umbral de detección de la intención de movimiento de los músculos involucrados en la flexo-extensión de la rodilla durante la marcha

Abstract

Three motion intention detectors of the muscles involved in flexion-extension of the knee during gait were studied, based on surface electromyography (sEMG). The purpose of this study is to estimate the detection threshold of the movement intention of nine specific muscles. We used a sEMG acquisition system formed by an EKG / EMG shield based on Arduino Mega. Participants were 21 subjects (14 men), age (21.52 ± 2.4 years) without pathologies at the level of the knee. sEMG of 9 muscles were recorded at 1 KHz and 58 ± 3 seconds in duration. Subsequently, a study of three detectors was carried out: Single Threshold (D1), Double Threshold (D2) and Double Statistical Threshold (D3). The comparison was made based on the distance to the perfect detection point (DDP) and time delay (Td), using a cross-validation technique. D3 presented a better performance (SEN: 85.88 ± 2.18%, ESP: 86.11 ± 3.55% Td: 6.24 ± 2.42 ms). Additionally, the minimum / desirable value of the SNR was determined for a reliable measurement of the Medium Frequency (MNF) of the sEMG. The signals were contaminated with different noise levels. The thresholds were determined using the K-means algorithm (Minimum: 5.51 dB, Desirable 12.28 dB). The beginning of muscle contraction oversees activating an orthopedic device, therefore, an accurate detection of this episode of muscular contraction is of crucial importance to ensure the proper operation of the device. The D3 detector fulfills these characteristics and this finding constitutes a contribution to the development of the exoskeleton prototype proposed by the DEET.