Comparison of Five Kinematic-Based Identification Methods of Foot Contact Events During Treadmill Walking and Running at Different Speeds

Abstract

This study involved a comparison of 5 kinematic-based algorithms to detect heel strike (HS) and toe-off (TO) events during human locomotion at different speeds. The objective was to assess how different running and walking speeds affect contact event determination during treadmill locomotion. Thirty male runners performed walking at 5 km/h and running at 9, 11, and 13 km/h on a treadmill. A kinematic system was used to capture the trajectories of 2 retroreflective markers placed at the subject’s right heel and second metatarsal. A footswitch device was used to determine the “true” times of HS and TO compared with 5 kinematic-based algorithms. The results of the current study illustrated that speed influences the HS error in the vertical position and horizontal velocity algorithms, and the TO error in the vertical position and horizontal velocity algorithms. This difference was found in the transition from walking to running; however, higher running speeds did not affect the error estimation. Higher accuracy was found with combined algorithms, namely, one using vertical acceleration and position and another using horizontal and vertical position with no influence from different locomotion speeds. Therefore, these algorithms are recommended in studies where speed is self-selected because they work well for a broad range of locomotion velocities.

A kinematic algorithm to identify gait events during running at different speeds and with different footstrike types

“…The use of the Heel vertical acceleration peak within the touchdown algorithm is consistent with a number of previous studies (Alvim et al, 2015;Leitch et al 2011;Maiwald et al, 2009). However, in the current investigation the touchdown algorithm also referenced the Met vertical acceleration peak which successfully extended this algorithm to include FFS runners.…”

“…The first toe-off algorithm was also based on that proposed in a number of previous studies (Alvim et al, 2015;Leitch et al 2011;Maiwald et al, 2009) and was given by the peak in vertical acceleration of the Toe marker within the toe-off window (TOA). The second toe-off algorithm followed the same procedure but used the vertical jerk peak of the Toe marker (TOJ) (Figure 1c).…”

“…The first touchdown algorithm was an adapted version of that proposed in a number of previous studies (Alvim et al, 2015;Leitch et al 2011;Maiwald et al, 2009) and involved identifying the vertical acceleration peaks of the Heel and Met markers within the touchdown window. The instant of touchdown was then defined by the first of these peaks (TDA).…”

“…Several authors have proposed algorithms for walking (De Witt et al, 2010;O'Connor et al, 2007;Zeni et al, 2008), and for running using only foot-based markers (Leitch et al 2011;Maiwald et al, 2009), only markers on more proximal segments (Milner et al, 2015), as well as both foot-based and more proximal segment markers (Fellin et al, 2010;Hreljac & Stergiou, 2000;Smith et al 2015). One algorithm has been proposed for both walking and running, although the running speeds were relatively low (2.5-3.6 m•s -1 ) (Alvim et al, 2015). Existing running algorithms have typically been limited to either only rearfoot strike running (Alvim et al, 2015;Hreljac & Stergiou, 2000) and / or single speed or multiple slow running speeds (Alvim et al, 2015;Fellin et al, 2010;Milner et al, 2015;Smith et al, 2015) and few have reported the accuracy of ground contact time (GCT) as well as the individual touchdown and toe-off GCEs (Hreljac & Stergiou, 2000;Smith et al, 2015).…”

“…One algorithm has been proposed for both walking and running, although the running speeds were relatively low (2.5-3.6 m•s -1 ) (Alvim et al, 2015). Existing running algorithms have typically been limited to either only rearfoot strike running (Alvim et al, 2015;Hreljac & Stergiou, 2000) and / or single speed or multiple slow running speeds (Alvim et al, 2015;Fellin et al, 2010;Milner et al, 2015;Smith et al, 2015) and few have reported the accuracy of ground contact time (GCT) as well as the individual touchdown and toe-off GCEs (Hreljac & Stergiou, 2000;Smith et al, 2015).…”