Ground reaction force (GRF) data captured directly by force plate has been used as a representative kinetic parameter to describe running performance. The purpose of this study was to evaluate the reliability of GRF estimates using kinematic data obtained with high-speed cameras, in relation to differences of running speed, foot strike pattern, and cut-off frequency of digital filtering. Participants were instructed to run at 2.0, 3.0, and 4.0 m/s, using rearfoot and forefoot strike patterns on an instrumented treadmill that has force sensors. These running motions were captured using 13 high-speed optical cameras at 250 Hz. 27 reflective markers were attached on the whole body, in accordance with the standard Plug-in-Gait model. The raw kinematic data were passed through a 4th order-zero-delay Butterworth digital filter using five different cut-off frequencies (6, 10, 15, 25, and 50Hz). The error of the mean values of antero-posterior and vertical GRFs was smaller than the error in the peak values. The error of the mean values of the vertical GRF was smaller than that predicted for the propulsive or braking GRFs in the antero-posterior component. Running speed and foot strike pattern did not affect the error in mean vertical GRF. Smaller cut-off frequencies in filtering resulted in smaller prediction errors (6 Hz: 0.4 ± 1.4%). These results indicated that the mean value of the vertical GRF estimated using a kinematic approach is a reliable parameter to describe the running performance in the absence of a force plate.
© Copyright 2014 International Calgary Running Symposium, August 14-17, 2014. All rights reserved.