Identification of gait events without a force plate during walking, running and sprinting: A comparison of methods

Abstract

When analysing gait, the identification of the period of stance is often needed. Traditional methods of identifying gait event (foot-strike/toe-off) times have used force platform recordings, however as force plates only have a small surface area multiple strides cannot be analysed. Recent literature has used kinematic methods to identify stance and it has been suggested that event times could be accurately predicted within both walking (Hreljac and Marshall, 2000: Journal of Biomechanics, 33, 783-786) and running (Fellin et al., 2010: Journal of Science and Medicine in Sport, Article in Press) contexts. The purpose of this study was to investigate if kinematic methods could be used to accurately identify gait events times within a sprint running context in comparison to the ‘gold standard’ force platform recordings, and also clarify how kinematic methods of identification alter between walking, running and sprinting. Following ethical approval, six volunteer (males) healthy university level middle distance runners (age 19.3 ± 0.9 years; body mass 68.5 ± 3.9 Kg; height 177.9 ± 4.6 cm) performed three standing trials and five walking, five running and five maximal sprinting trials over an active force platform (1000) Hz). Kinematic data was collected using a CODA automatic motion analysis system (400 Hz) and marker coordinate information and peak vertical acceleration recordings from active markers located on the heel and fifth MTP were examined to identify gait events times. RMS differences between kinematic methods and criterion force plates method were calculated. Peak vertical acceleration algorithm most accurately identified touchdown and take-off within sprinting trials to between 0.003 and 0.006 seconds compared to criterion time. Marker coordinate data was able to identify gait events between 0.008 and 0.028 in the sprint trials. These findings revealed that kinematic methods could be used to identify events within a sprint running context to the same accuracy to that of walking and running. The study also highlighted that whilst methods of stance identification remain consistent for both walking and sprinting trials, this is not the case within a running context and the need to clarify landing techniques of different individuals is essential before using the proposed methods within this study.