Individualized Optimization of a tuneable prosthetic foot to lower the metabolic cost of walking for transtibial amputees

People after lower-limb amputation often experience reduced quality of life.
The reduction in quality of life is to a great extent the consequence of the
diminished mobility. To improve mobility, people after lower-limb amputations
are provided with an artificial limb that imitates the biological ankle-foot
function. However, current prostheses do not succeed to fully compensate for
the normal foot-ankle system and people with an amputation prefer a lower
self-selected walking speed. Altogether, this results in a 35% increase of
metabolic cost of walking compared to able bodied controls at similar walking
speeds.
To lower the metabolic cost of walking and improve mobility of lower-limb
amputees, the natural foot roll over function and push-off needs to be
restored. However, problems with the timing of energy return of the prosthetic
foot have been observed and the importance of prescribing individually-tailored
prosthetic components was already highlighted. Due to the difficulties in
individual tuning of prosthetic components, the effectiveness of current state
of the art assistive devices stays behind expectations.
A potential method to overcome these challenges is human-in-the-loop
optimization, in which the human is included *in vivo* in the control loop and
device parameters are systematically varied using an intelligent optimization
algorithm in response to a defined and directly measured cost function to
optimize human (-machine) performance. By means of this method, the parameters
can be tuned to biomechanical properties of the individual while taking into
account the natural adaptive behaviour of humans who simultaneously optimize
coordination patterns with respect to locomotor performance.

The aim of the current project is to evaluate the effect of a prosthetic foot,
which can be individually optimized using human in the loop optimization, on
metabolic cost of walking for transtibial amputees. We want to investigate both
the effect of the change in foot properties (i.e. joint stiffness and
alignment) as well as the effect of motor learning during this process on the
final reduction in metabolic cost after human-in-the-loop optimization.

Pre-posttest quasi experimental repeated measures design.

Participants will use a passive prosthetic foot with a manually tuneable joint
stiffness and alignment. During the optimization period, we will aim to
optimize the joint stiffness and alignment of the prosthetic foot to minimize
the metabolic cost of walking. The joint stiffness and alignment will be
changed manually by changing the position of the joint slider and the 'tendon'
length of the prosthesis. The settings that resulted in the lowest metabolic
cost of walking will be selected as optimal settings and will be compared with
the tuneable prosthetic foot with standard settings (reference joint stiffness
from participant*s personal prosthetic foot and neutral alignment), and the
participant*s personal prosthetic foot.

Participants (transtibial amputees) will be invited twice to the GRAIL lab of
the Center for Rehabilitation of the UMCG, Beatrixoord. During the visits the
participants will be asked to walk on the treadmill on self-selected walking
speed, while data for breathing gas exchange, kinematics and kinetics are
collected. The burden for the participants is that they have to spend about 4
hours in the lab and have to walk for almost one hour (with intermediate
brakes) during each session at self-selected comfortable submaximal walking
speed.
During the visits, the participants will wear a prosthetic foot with manually
tuneable joint stiffness and alignment. Although fitted by a certified
prosthetist, use of this prototype can cause some discomfort. Walking with the
new prototype can feel uncomfortable, mostly at the beginning of practice with
the prosthesis. Getting used to a new prosthesis, like this prototype, may also
cause muscle strain. However, these effects are not different from when the
patient is provided with a new prosthesis as part of regular clinical
management. Although the prosthetic prototype has been extensively tested for
safety and use, the unlikely event that the prototype may break could cause a
participant to fall. However, when using the prototype, the participants will
always wear a safety-harness, which will prevent them from harm in the event of
a fall. In addition, the researcher will always be present and next to the
participant to support the participant and prevent a fall from happening.