Stationary versus bifurcation regime for standing wave central pattern generator

• Published in: Biomedical signal processing and control Vol. 32; pp. 57 - 68
• Main Authors: Martin del Campo, R., Jonckheere, E.
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.02.2017
• Subjects: Bifurcation; Bootstrapping; Central pattern generator; Coherence; Surface electromyography; Wavelets; Bootstrapping; Bifurcation; Wavelets; Surface electromyography; Central pattern generator; Coherence
• ISSN: 1746-8094
• DOI: 10.1016/j.bspc.2016.08.011

•Coherence between wavelet subbands of spinal sEMG signals is “robustly” reproducible.•Reproducibility spans across a large population of subjects of different background.•Reproducibility spans across raw signals of significantly different visual appearance.•Cross power spectral density mean shift is demonstrated with a=0.5 significant level.•Shift in the cross power spectrum density is dynamically a period halving bifurcation. The purpose of this research is to show that the correlation analysis on surface electromyographic (sEMG) signals that originally confirmed existence of a standing wave central pattern generator (CPG) along the spine are reproducible despite evolution of the entrainment technique, different hardware and data collection protocol. Moreover, as major novelty of the present research, it is shown that this CPG can undergo “bifurcations,” here revealed by signal processing extrapolated towards the period-halving dynamical interpretation. The visually intuitive manifestation of the bifurcation is statistically confirmed—using bootstrap analysis—by a shift in the cross power spectral densities, consistently with the standing wave occurring on different subbands of the Daubechies DB3 wavelet decomposition of the sEMG signals.