The main goal of the FP7 project BALANCE (balance-fp7.eu) is to explore the possibility to support postural stability of a human, through the use of exoskeletons, focusing on control approaches.
An important aspect of such an exercise is how to measure the balance-performance of the human, and of the human supported by an exoskeleton, that is: benchmarking balance. A first source of inspiration, we thought, can be the way balance-performance is assessed in humans, especially through the clinical scales that are used to assess balance ability in patients with different impairments.
The existing clinical tests generally start from a functional perspective and examine how well a person can perform a variety of functional tasks that require control of balance. Many relevant tests and their components related to balance performance can be found in the ‘rehabmeasures’ database, with a search on balance. The fact that so many clinical scales exist and are being used, may be due to the infinite number of potential functional tasks that involve balance as a component of the task. There seems to be no clear advantage of one scale over the other, nor is there an intrinsical
ly obvious method of scoring the underlying balance performance component of mobility tasks, as separated from other aspects of that task. The clinical tests typically assess the ability to perform specific activities while balanced, and are scored through qualitative, subjective scoring of the ability to perform the tasks in a stable way, or by measuring speed or duration of task completion. A general problem with such assessment procedures is that the reduced ability to perform functional tasks other than quiet standing, such as walking, may be the result of a broad range of sensory, motor and cognitive deficits that are not necessarily related to balance skills alone.
We observed that several more complete clinical assessment procedures, such as the BESTest procedure, involve providing pushes or perturbations to the human and judging the ability to deal with those, but that these pushes are typically provided by a therapist in a subjective way, and are only provided during standing, most probably because of practical reasons.
To overcome such limitations we focused on building devices that could provide perturbations during both standing and walking, and moreover provide the perturbations in a very precise way. These perturbations are precise in size (force profile over time) as well as precise in timing (at which moment in the gait cycle), and do not hinder the human reaction to the perturbation (transparent operation mode). One device was developed by University of Twente, The Netherlands; this is a pusher that is mounted over a treadmill that can provide defined force-profile pushes at pelvic level to a standing or walking subject. Another device was developed by University Rehabilitation Institute; Slovenia, this is a mobile over ground device with comparable abilities, but that additionally also can provide rotational perturbations, and allows turning gait.
Until now, both devices were mainly used to study human reactions to specific perturbations in standing and walking. In BALANCE, they will also be used to benchmark balance performance. The simplest metric is to evaluate experimentally until which size of perturbation the human, or the human with exoskeleton is able to handle the perturbation (without falling, or without making a corrective step), but more precise metrics are also being studied, in order to quantify and normalize the specific behavior in response to a perturbation. These metrics are based on properties like the total body Center of Mass, the Center of Pressure, the Base of Support or also on the Centroidal Dynamics of the total body and describe the quality of the balance performance in a qualitative way.
About the author: