Citation

Weiske F, Böhme M, Jäkel J, Zentner J, Witt M. J. Biomech. 2021; 119: e110316.

Copyright

(Copyright © 2021, Elsevier Publishing)

DOI

10.1016/j.jbiomech.2021.110316

PMID

unavailable

Abstract

Understanding gait differences in context of group differences is dependent on statistical testing methods and time normalization techniques (TN). The method induces a relationship of both with one another. As to our knowledge, there has been no investigation into their relation so far. To show empirically what effects may be of importance, we use SPM with linear time interpolation (LI) and Dynamic Time Warping (DTW) separately for data of a study on stair ascent kinematics between two groups. There is a slight difference in statistical significance for the comparison of LI and DTW. LI-uniquely significant time highlight differences due to in-group time-variations, whereas DTW-uniqueness is tied to qualitative differences of homogeneous events. The comparison of stair ascent kinematics with DTW shows more pronounced evidence for backlift-like strategies for the older group, although trunk angles are kept more extended as to ensure stabilty. Thus, the difference in SPM from TN is slight but important, if there is need to mirror said effects methodically.

Language: en

Keywords

Dynamic time warping; Stair ascent biomechanics; Statistical parametric mapping