Figure 1

Methods for kinematic analysis during transspinal stimulation. (a) Experimental set up and placement of transspinal stimulation electrodes and reflective markers for tracking body segments. Subjects walked on a treadmill for 10 min while transspinal stimulation was delivered with a DS8R stimulator at 15, 30 and 50 Hz with 10 kHz carry over frequency at sub- and supra-threshold stimulation intensities. 3D kinematics were recorded for the lower limbs with an 8-camera optoelectronic system at 150 Hz. (b) Step length during walking was defined as the length between the left and right footprints at the time of heel contact. Step width during walking was defined as the mediolateral distance between two subsequent footprints. (c) Estimation of changes in step length and step width fluctuations as a function of timescale for the detrended fluctuation analysis (DFA). The demeaned and integrated spatiotemporal time series was divided into sequences of non-overlapping windows. A fitted linear regression line was subtracted from the data in each window (detrending), and the average of the local squared residuals was calculated for each window. This process was repeated for several different timescales or windows. A regression analysis was performed between the squared root of the average squared residual and the timescale to estimate the scaling exponent α-DFA. (d) Phase space reconstruction for the calculation of the largest Lyapunov exponent (LyE) of the joint angles using the method of delayed embedding. Optimal time delay (τ) and embedding dimension (m) parameters were calculated using the average mutual information and the false nearest neighbor algorithms. The 3D projection of the reconstructed phase space obtained with optimal embedding parameters τ = 24 and m = 3 for the right knee joint angle of a subject during control walking is shown. (e) Analytic signals were computed from the centered segment angles via Hilbert transformation. The continuous relative phase was calculated as the arctangent of the product of the analytic signal of the proximal angle with the conjugate analytic signal of the distal angle. The continuous relative phase estimation between the right thigh-shank coupling is shown. (f,g) Definition of the parameters used to compute margin of stability (MOS) at heel contact (f) and at toe-off (g). The margin of stability was calculated as the antero-posterior distance from the extrapolated center of mass (xCOM) to the lead heel (HEEL). Negative margin of stability means that the xCOM is ahead of the heel, while positive margin of stability means that the heel is ahead of the xCOM. COM: center of mass; xCOM: extrapolated center of mass; MOS: margin of stability.