Characterization of spatiotemporally complex gait patterns using cross-correlation signatures.

We hypothesize that spatiotemporal joint coupling patterns during gait are closely associated with musculoskeletal injury mechanics. Previous studies examining joint coupling, have primarily focused on coupling between single pairs of neighboring body segments or joints; thus falling short of characterizing the full spatiotemporal complexity across the entire gait apparatus. This study proposes the reliance on properties of the temporal cross-correlation of distinct joint variables as a means to characterize and detect differences in multiple segmental coupling pairs and to quantify how these couplings change between different gait conditions or test groups. In particular, for each subject, a characteristic diagram array is obtained whose entries include the maximum values of the cross-correlation between all pairs of joint variables as well as the associated phase shifts at which these maxima are recorded. Paired t-tests are then used to highlight significant differences in the corresponding entries between two gait conditions. In the present study, this technique was applied to angular displacement and velocity histories across 12 lower extremity joint variables, for healthy subjects with and without a brace on the right knee. As expected, the statistical analysis indicated that the temporal cross-correlations associated with the right knee-angle variables differed the most between the two gait conditions. In addition, significant differences (p<0.01) were found in the coupling between other pairs of joint variables, establishing a characteristic spatiotemporal signature for the changes from normative gait that result from reduced mobility at the knee.