Post-Stroke Gait Analysis Using Motion Capture

Post-processing of motion capture data was performed using Vicon Nexus software (version 2.10, Vicon Motion System Ltd., Oxford, UK) with 5 trials in which the animal maintained a consistent walking pace reconstructed for each session (8-, 5-, and 1-days pre-stroke and 3 days post-stroke). Markers were labelled and spline and cyclic algorithms used to fill all visible gaps (Vicon Motion System Ltd., Oxford, UK). A fourth order, zero lag, low pass Butterworth filter was applied with a cut-off frequency of 10 Hz. Data were exported to C3D format and further processed with custom MATLAB^® code (Mathworks, Natick, MA, USA).
Motion capture parameters selected for analysis sought to capture post-stroke gait, asymmetry and general apathetic behaviour observed, such as lowering of the head and shoulders and inability to extend the fetlock joint contralateral to the stroke affected hemisphere. Parameters of interest were subsequently classified into global and limb-specific. The final parameters selected for analysis and their purpose are provided in Supplementary Table S1 (global parameters) and Supplementary Table S2 (limb-specific parameters). Global parameters correspond to the outcome measures pertaining the entire trial, for example forward velocity. These outcomes were calculated as the mean value across the entire trial. Limb-specific parameters were computed from the observation of the kinematic data of each limb within its corresponding gait cycle. For each trial, the gait cycles were identified following the method from Ghoussayni et al. (43 (link)). Changes in the velocity of each limb's hoof marker (DPHAL, Figure 2) were detected in the vertical and progression directions, determining when the marker stopped moving (entering stance phase) or started moving (entering swing phase). One complete gait cycle per limb was extracted from each trial to calculate kinematic measures of interest using two-dimensional (2D) planar analysis. Planar analysis was independently performed in the vertical (sagittal) and lateral (left/right) directions. Joint angles were defined between two vectors in the sagittal plane for the fetlock, carpus, and elbow of the forelimb, and fetlock, tarsus, and stifle of the hindlimb (Figure 2).