Sixty young participants were involved in this study (30 male
and 30 female, age 24.2 ± 2.1 years; height 169.1 ± 9.2 cm; weight
58.5 ± 10.3 kg). They self-reported to be free of any medical conditions that may affect their gait patterns. All participants signed an
informed consent form approved by the NTU Institutional Review
Board.
Apparatus and experimental protocol were the same with our
previous work (Hu and Qu, 2013). In brief, reflective markers were
placed on 26 anatomical landmarks. This marker setup can help
establish a 12-segment body model, which includes the head,
upper arms, lower arms, trunk, thighs, shanks and feet. Standard
footwear was provided to ensure the same floor coefficient of friction across participants. The participants were asked to wear a full
body safety harness to prevent injuries due to fall impacts. Walking
trials were conducted on a linear walkway (12 m1.5 m) which is
covered by vinyl tiles. Slips were induced using a removable vinyl
tile sheet (1.0 m1.2 m, lengthwidth) covered by water-detergent mixture (85% Dynamo Powereactive
plus 15% water). The vinyl material of the removable sheet was exactly the same as that of
the walking surface.
The participants were instructed to walk on the linear walkway
at their self-selected speed. To avoid the participants’ anticipation
of the slips location, the removable sheet was randomly placed at
one of four possible locations along the walkway. The participants
were instructed to look straight forward during walking. Meanwhile, the light was dimmed to prevent the awareness of changes
in the walking surface. For each participant, falls would be identified if the midpoint between the left and right hip joint centers
(estimated by the ASIS markers) dropped below 95% of its minimum height measured during normal walking (Beschorner and
Cham, 2008). Otherwise, balance recovery was considered to be