Biomedical Engineering Reference
In-Depth Information
8.2.3 Gait Asymmetry and Regularity
As stated above, walking is a quasi-cyclic and almost symmetrical motion. Measuring
gait asymmetry and loss of regularity is widely used for diagnosis [
17
], especially in
double-tasks protocols with cognitive loads (such as counting down while walking).
There are two main methods to measure symmetry. The simplest one consists in
computing a ratio between left and right values [
30
]:
(
R
−
L
)
×
(
R
+
L
)
×
100
0
.
5
where R and L are respectively measurements performed on the right and the left
respectively (such as step length or step frequency). The other approach consists in
computing auto-correlation of the studied signal [
2
]. This signal is generally twice
the frequency of the stride. Hence auto-correlation of this signal between one step
and the following will provide information on asymmetry. Auto-correlation between
several strides will provide information on regularity.
As described above irregularity and asymmetry mainly occur when the user
has to perform a cognitive activity while walking. In immersive systems based on
metaphors, the cognitive load of the user may be different compared to natural walk-
ing. Regularity and symmetry may then be affected even if it has not been demon-
stratedinVRyet.
8.3 Joint Angles, Torques and Muscle Activity
In the previous section human body was modelled as a point (his center of mass). We
have seen that it could provide relevant information about the main gait pattern, such
as speed, step length and frequency. In VR this type of information could be used to
globally adapt the motion of the virtual camera or to capture basic gait parameters
to drive the simulation. However more accurate information could be required in
order to drive an avatar or capture more accurate information about the user's gait.
To this end, it is necessary to have multi-point measurements and to capture more
complex parameters, such as joint angles and position, but also to compute dynamic
parameters such as joint torques or muscle activation patterns.
8.3.1 Measuring Joint Displacements
The most popular method to get joint position depending on time is to track visual
markers placed over standardized anatomical landmarks. The International Society
of Biomechanics (ISB) has standardized this markers' placement in order to enable
