Detection of movement cycles from multi-channel SEMG without kinetic and kinematic signals

Published: 7 December 2020| Version 1 | DOI: 10.17632/ggzyf42cvr.1
Contributor:
Jan Sedlak

Description

MATLAB_pcode.zip (main class + MC detection algorithm) - can check the algorithm performance on attached emg data by on one click solution Matlab pcode of proposed algorithm for movement cycle identification. Averaged SEMG profiles estimation is performed only by using multi-channel SEMG signal. function [x_segment, measure_conf, details] = MC_detection(signal_raw,fs,mode), REAL_sEMG_MC_KNETIC_REF_database.zip Dataset I.a Nordic walking uphill and downhill, self-selected pace on naturally bumpy terrain with inclination 15 degrees for 2 subjects Muscle (m.) biceps brachii (BB), m. triceps brachii (TB), m. latissimus dorsi (LD), m. pectoralis major (PM), m. deltoid anterior (DA), m. deltoid posterior (DP), m. trapezius descendens (TD), m. gluteus medius (GM) were measured on the right side of the body. M. trapezius transversalis (TT) and m. serratus anterior (SA) were monitored bilaterally. Probands were instrumented by one three-axis accelerometer for gait cycle identification. 16 records with average length 14.88 ± 1.78 MCs, 238 cycles, average period 1.107 ± 0.040 s Dataset I.b walking and running for two probands on smooth terrain at self-selected pace Signals were measured bilaterally for m. tibialis anterior (TA), m. peroneus longus (PL), m. gastrocnemius lateralis (GasL), m. gastrocnemius medial part (GasM), m. rectus femoris (RF), m. vastus medialis (VM). Pair of foot-switches were fastened under the heel and tip of right and left leg. 20 records with average length 26.75 ± 13.06 MCs in total amount 530 cycles with average period 0.902 ± 0.170 s Dataset I.c rowing on a single skiff in natural conditions and ergometer (CONCEPT2) on 7 probands Muscles BB, GM, m. gluteus maximus (Gmax), RF, m. biceps femoris (BF) and m. semitendinosus (SEM) monitored bilaterally Stroke was identified by using a three-axis accelerometer. The accelerometer was attached on the torso with the belt from an elastic fabric. 32 records with average length 17.22 ± 5.64 MCs in total amount 545 cycles with average period 2.760 ± 0.508 s SYNTHETIC_sEMG_MC_REF_database.zip Synthetic EMG data for walking,running,rowing, Nordic-walking with reference of MC and muscle activity are attached. Details are described in related article.

Files

Institutions

Ceske Vysoke Uceni Technicke Fakulta elektrotechnicka

Categories

Locomotion, Movement, Gait Analysis, Musculoskeletal Function, Biomechanics of Gait

Licence