Skip to main content
Fig. 3 | Journal of NeuroEngineering and Rehabilitation

Fig. 3

From: Individualized feedback to change multiple gait deficits in chronic stroke

Fig. 3

Inputs for the calculation of real-time principal component for visual feedback display. (a) Loadings for each dimension were calculated by submitting each individual’s baseline deficit to principal component analysis (PCA). These loadings were used to calculate a real-time principal component (PC1) which is a vector multiplication of the normalized real-time deficit (defined as the difference between the z-scored real-time kinematics and the prescribed goal kinematics) and the loadings. (b) Stroke-to-control group average baseline kinematics (solid traces) plotted against goal kinematics (dashed traces) for the four input dimensions. The shaded gray regions denote the rewarded time window during swing phase for that particular dimension (c) Control-to-stroke group average baseline kinematics (solid traces) plotted against goal kinematics (dashed traces) for the four input dimensions. The shaded gray regions denote the rewarded time window during swing phase for that particular dimension. The 10 % width of these windows is only approximate as the time window was fixed at 100 ms for each individual (d) Stroke-to-control group average loadings for the four input dimensions. (e) Control-to-stroke group average loadings for the four input dimensions. (f) Percent variance explained by the first principal component. All shaded errors and error bars denote SEM

Back to article page