Digital filtering of kinematics of racing wheelchair propulsion

The purpose of this study was to examine the residual analysis technique described by Winter [1] in the digital filtering of the kinematics of racing wheelchair propulsion. Specifically, the selection of the type and cut-off frequency of the digital filter were examined. The residual analysis was performed on the second metacarpophalangeal joint (2MP) marker. Four different types of digital filters were examined: a 2(nd), a 4(th), and, 10(th) order Butterworth, as well as a 10(th) order Chebyshev type I with a passband ripple of 0.001 dB. The residual analysis was useful in determining the appropriate cut-off frequency for each individual filter. The cut-off frequency was between 7 and 8 Hz in the x-direction, between 9 and 10 Hz in the y-direction, and 7 Hz in the z-direction, depending on the type and order of the filter implemented. The residual analysis was not useful in determining the appropriate type of filter. Knowledge regarding the magnitude and phase characteristics of the filter, as well as the type of data to be filtered, should be used to determine the most appropriate type of filter.