Rehabilitation
is the process by which patients or elderly individuals who have
experienced loss or impairment of their motor ability try to regain it
through training. A major element that rehabilitation attempts to
recover is muscle strength, but the restoration of proprioception, which
provides sensory information on limb position, joint speed and
acceleration, and posture perception, is also important [1]. Proprioception decreases with aging [2, 3] and may also deteriorate with injury or disease such as stroke [4, 5].
It is also involved in movement stability; when proprioception of the
ankle joint decreases or worsens, postural sway increases, and gait
stability deteriorates, which increases the risk of falling [6, 7].
Various
rehabilitation systems are being developed to prevent fall, and powered
ankle–foot orthosis (PAFO) is one of them. PAFOs are wearable robots
that helps rotate the wearer’s ankle joint [8].
Usually, they have a talocrural joint as an only axis, which is
responsible for rotation in the sagittal plane. And they assist in
providing the propulsion required for a walker to move forward. The
purpose of these 1 degree-of-freedom (dof) PAFOs are to aid in
propulsion, and many studies are being conducted to optimize the
controller and minimize the objective functions related to muscle
activity or metabolic rate [9,10,11,12].
However, as the performance of rehabilitation devices improve enough to
be applied in the clinic and they are becoming sufficiently safe,
studies on PAFOs which are focused on improving the wearer’s gait
stability rather than providing simple assistance is becoming more
common [13,14,15].
The goal of these qualitative rehabilitations is to restore the walking
ability of the patients in a way that minimizes the risk of falling,
rather than focusing only on moving forward, when the wearer’s ability
to walk is impaired due to an accident or disease.
To properly
restore the function of the ankle joint, it is necessary to consider the
configuration of the joint. The ankle has a subtalar joint that rotates
in the frontal plane in addition to a talocrural joint and its function
is closely related to stability, the restoration of which is a goal of
rehabilitation [16].
The subtalar joint preserves gait stability by controlling the rotation
of the center of mass (CoM) in the frontal plane, which is known as the
ankle strategy or foot tilt strategy [17, 18].
It generates a stabilizing moment that compensates for the tilting
moment caused by the misalignment of the projection of the CoM, which is
located in the trunk, on the ground and the plantar center of pressure
(CoP) through eversion so that an individual can walk in a stable state.
Approximately 80% of the gait cycle, excluding initial and terminal
double support, involves a one-leg stance [19],
so the ankle strategy performed during this period is very important in
reducing the risk of falls of walkers. In fact, the reason for the
deterioration of gait stability with aging is the decrease in the range
of motion of eversion along with the weakening of the plantar flexion [20]. And this shows that ankle movement in the frontal plane considerably contributes to the prevention of fall risk.
However,
the PAFOs that have been developed thus far have focused just on the
talocrural joint. Rehabilitation performed in the absence of the
subtalar joint does not enable complete recovery of proprioception, and
learning of ankle strategy is not easily achieved, so there is a
possibility that the patient won’t return to the pre-injury state after
rehabilitation. To solve this problem, we fabricated a PAFO with both
talocrural and subtalar joints [21]
and used two pneumatic artificial muscles (PAMs) to simultaneously
assist plantar flexion during propulsion and eversion when creating a
stabilizing moment. Additionally, through comparative experiments with a
1 dof PAFO, we proved that eversion has a positive effect on reducing postural sway during walking with PAFOs [22].
However, since we measured simple postural sway only, we could not
analyze how the power provided by the PAMs affects gait stability of the
wearers.
Gait stability cannot be fully evaluated by assessing
outward fluctuations such as postural sway, and it is usually assessed
by examining cycle variation or local dynamic stability. Cycle variation
is quantified by calculating the kinematic variability between cycles
during cyclical movement [23].
Local dynamic stability is a method used to evaluate stability that
involves measuring how the magnitude of a deviation increases as a
cyclical movement proceeds after an initial external perturbation to the
system [24].
Both are actively used to evaluate gait stability and the performance
of exoskeleton robots, and both are directly related to the risk of
falling, which is caused by deterioration of stability [25, 26]. Although we have already confirmed that the developed 2 dof PAFO affects postural sway in positive ways with previous study [22], we wanted to know whether the subtalar joint, which is included in the 2 dof
PAFO, actually assists the ankle strategy of the wearers and generates a
stabilizing moment to compensate for the tilting moment and preserve
gait stability which were evaluated with not only outward fluctuations
with postural sway but also indices for gait stability like kinematic
variability or local dynamic stability.
It was proved that the
characteristics of assistance from PAFO affect the wearers’ stability
and that it might be more important than proper foot placement [27,28,29]. Although these studies are about experiments or simulations conducted with prosthetic foot or 1 dof
PAFO, but they were enough to show the importance of assistance
characteristics in wearers’ gait stability. So, considering that the
fall risk is related to the weakening of the evertor [20],
it is also meaningful to examine the wearers’ gait stability according
to the characteristics of the assistance provided by 2 dof PAFO
with subtalar joint as the rotation axis. In fact, in order to preserve
stability during steady state gait, it is necessary to correct errors
that occurred during foot placement, which can be implemented by an
active ankle strategy by shifting the CoP in the medial–lateral
direction [30]. So, if the assistance of 2 dof
PAFO is conducted by designing proper controller, it helps to shift the
CoP in a direction that compensates the tilting moment by assisting
ankle strategy, or to provide a stabilizing moment so that it is
expected that wearer can perform a stable gait.
In this study, we examined how the power provided by the 2 dof
PAFO affects the gait stability of wearers using indices such as local
dynamic stability and kinematic variability. In general, the trajectory
of the CoP in the global coordinates is said to be positioned in more
laterally to improve gait stability. However, in the case of in-shoe CoP
measured in the local coordinate system of the sole during one-leg
stance, the trajectory tends to move in a more medial direction during
eversion because of the decrease of tilting moment [31]. We found in a previous study that the power generated by the PAM of a 1 dof
PAFO for assisting plantar flexion causes the in-shoe CoP to move in
the lateral direction, resulting in an increase in the tilting moment
and subsequent postural sway, and that the power of the PAM in a 2 dof
PAFO that is used to strengthen eversion compensates for that
phenomenon. We used phase-based controllers (PBC) for previous study
which is normally used for PAFOs but in this paper with these findings
we developed controllers for PAMs based on the lateral deviation of the
in-shoe CoP trajectory caused by propulsion assistance. Through the
clinical experiments, we wanted to prove that the PAMs of 2 dof
PAFO mitigate the deterioration of stability when stabilizing moments
are appropriately provided to the wearer based on this algorithm. These
experiments aimed to understand whether the ergonomic properties of the 2
dof PAFO could sufficiently support the wearer’s ankle strategy in the aspect of the gait stability.
