Biomedical Engineering Reference
In-Depth Information
of optic flow has been shown to influence the perceived speed of participants and
to elicit changes in walking speed [
37
,
40
,
43
,
62
]. Similarly, shifting the pattern
of optic flow to the left or right influences the perceived heading direction [
70
],
and elicits compensatory postural [
3
,
69
] and steering adjustments when walking to
a goal [
59
,
68
]. Using VR to manipulate optic flow thus has the potential to alter
the interaction space and provide salient information about locomotion speed and
heading to the patient.
Lamontagne et al. [
28
] used such a manipulation to examine the perceptual-
locomotor adaptability of patients suffering from post-stroke hemiplegia. During
two experiments patients and unaffected control participants walked on a human-
driven treadmill while virtual corridors provided optic flow information through a
head-mounted display (HMD). The first experiment required participants to walk
at a comfortable speed as the optic flow rate was varied continuously in an open-
loop sinusoidal pattern at 0.017 Hz. This resulted in a compensatory out-of-phase
relation between gait and optic flowspeed for all participants (i.e., participantswalked
faster during slower optic flow conditions and vice versa),
4
although this was less
pronounced for the patients and their phase relation was much more variable. In the
second experiment the walking speed of participants during a baseline optic flow trial
(1:1 mapping between walking pace and optic flow) was compared to their walking
speed in a series of trials in which optic flow was discretely manipulated above or
below the comparison trial. Again, walking speed was inversely related to rate of
optic flow, but the patients were equal to the healthy controls in their gait response to
optic flow. Taken together, the results of these two experiments provide evidence that
patients with hemiplegia following stroke are influenced by optic flow in a similar
way to healthy controls. This indicates, preliminarily, that virtual optic flow might
be useful in training these patients to increase their walking speed over the course of
a training intervention.
VR has also been used to manipulate visual cues to modulate the gait character-
istics of patients with Parkinson's disease through both continuous optic flow (e.g.,
[
49
]) and continuous information paired with discrete visual stimuli (e.g., [
60
]).
Similar to Lamontagne et al. [
28
], Schubert et al. required patients with Parkinson's
disease and control participants to maintain a preferred walking speed on a human-
driven treadmill while they viewed an optic flow pattern that varied at a constant
speed perceived to be either faster or slower relative to the preferred walking speed
of each participant. The results indicated that the patients with Parkinson's disease
were more susceptible to changes in optic flow speed (i.e., their preferred walking
speed was more variable) compared to control participants. The researchers con-
cluded that the patients were more reliant on visual information, perhaps due to their
4
The relation between optic flow and gait speed has been studied extensively (see text). While the
findings of Prokop et al. [
43
] and Mohler et al. [
37
] parallel those of Lamontagne et al. [
28
], it is
unclear why, exactly, the out-of-phase relation was observed. One possibility, as suggested by the
authors, is that a sinusoidal change in optic flow speed may lead to a more pronounced time lag
between the change in stimulus and the behavioral response. Another is that when the flow rate
decreases, the participant walks faster to compensate for a perceived decrease in speed, in order to
maintain a constant or preferred speed [
37
].
