Biomedical Engineering Reference
In-Depth Information
device [
41
,
52
] is based on the idea that the user has to lean his whole body to indicate
direction and speed of the navigation task in virtual environments. The key idea is
to naturally stimulate the vestibular system by simply leaning the body in order to
maintain the perception-action coupling that occurs when walking.
3.4 Comparison Between Ground and Treadmill Walking
One of the most popular methods used to walk through virtual environments is to
use treadmills [
32
,
67
]. It physically engages the user in a very natural manner and
the proprioceptive and sound feedbacks are naturally generated through walking.
However, it is important to check if treadmill walking is really similar to ground
walking to ensure that users are engaged in a natural walking experience in VR.
Some authors [
66
] suggest that walking on the CyberWalk, a large omnidirectional
treadmill, is very close to normal walking, especially after some initial familiariza-
tion. This statement was mainly based on comparing kinematic gait parameters, but
it would be interesting to evaluate to what extent this statement is true from the
dynamic point of view and if this difference in ground reaction force could affect
presence in navigation tasks. Some authors have demonstrated that even if kinematic
gait parameters could become similar in ground and treadmill walking after some
training, the joint torques and muscular activity remains different [
38
].
The main difference between ground and treadmill walking is that some variations
of the belt speed occur in treadmill walking (some researchers reported a 5 % change
in treadmill speed within a walking cycle [
58
]). These variations seem to be corre-
latedwith the subject's mass, gait speed, and the intrinsicmechanical properties of the
treadmill itself. The general consensus in the literature is thattreadmill walking leads
to higher step frequency and lower step length [
14
,
80
]. It seems that these variations
lead to an increase in variability duringwalking [
60
] and a decrease of the casual speed
compared to ground walking [
7
]. Globally, variations of the treadmill speed lead to
instability with a longer stance phase compared to ground walking [
14
,
44
]. Percep-
tion of walking speed is also affected and vision is globally perturbed [
81
]. Recent
studies have confirmed that one's perception of speed is influenced by the treadmill,
such that individuals were unable to match their corresponding self-selected ground
running speed [
36
]. The unmatched perception of speed is likely due to the distor-
tion of normal visual inputs resulting from the discrepancy between observed and
expected optic flow [
36
]. Some authors have artificially induced greater step vari-
ability through visual perturbations from a VR display [
47
]. Perturbations generally
induced greater variability in both step width and (to a lesser degree) step length.
However, it seems that most of the reported differences decrease with some train-
ing period. Four to six minutes of training seem to significantly decrease kinematic
differences between ground and treadmill walking [
68
], but the adaptation period
differs from one user to another. For example, elderly people have more difficulty
adapting to treadmill walking compared to younger adults [
61
,
82
].
