Date Published: February 14, 2017
Publisher: Public Library of Science
Author(s): Behdad Dehbandi, Alexandre Barachant, Anna H. Smeragliuolo, John Davis Long, Silverio Joseph Bumanlag, Victor He, Anna Lampe, David Putrino, Jeffrey M. Haddad.
The objective of this study was to determine whether kinematic data collected by the Microsoft Kinect 2 (MK2) could be used to quantify postural stability in healthy subjects. Twelve subjects were recruited for the project, and were instructed to perform a sequence of simple postural stability tasks. The movement sequence was performed as subjects were seated on top of a force platform, and the MK2 was positioned in front of them. This sequence of tasks was performed by each subject under three different postural conditions: “both feet on the ground” (1), “One foot off the ground” (2), and “both feet off the ground” (3). We compared force platform and MK2 data to quantify the degree to which the MK2 was returning reliable data across subjects. We then applied a novel machine-learning paradigm to the MK2 data in order to determine the extent to which data from the MK2 could be used to reliably classify different postural conditions. Our initial comparison of force plate and MK2 data showed a strong agreement between the two devices, with strong Pearson correlations between the trunk centroids “Spine_Mid” (0.85 ± 0.06), “Neck” (0.86 ± 0.07) and “Head” (0.87 ± 0.07), and the center of pressure centroid inferred by the force platform. Mean accuracy for the machine learning classifier from MK2 was 97.0%, with a specific classification accuracy breakdown of 90.9%, 100%, and 100% for conditions 1 through 3, respectively. Mean accuracy for the machine learning classifier derived from the force platform data was lower at 84.4%. We conclude that data from the MK2 has sufficient information content to allow us to classify sequences of tasks being performed under different levels of postural stability. Future studies will focus on validating this protocol on large populations of individuals with actual balance impairments in order to create a toolkit that is clinically validated and available to the medical community.
It is estimated that each year, fall-related injury costs the United States $34 billion in direct costs . For elderly individuals alone, fall-related visits to the emergency department have been estimated to cost an average of $2,823 per visit, with each fall-related hospitalization estimated to cost an average of $25,465 . Methods for stratifying fall risk and focusing preventive measures on those who are at highest risk remains one of the most important management strategies for saving resources and preventing suffering [1–3]. However, sensitive and reliable measures for assessing balance and fall-risk are difficult to develop and implement. One widely used balance measure, the Berg Balance Scale (BBS), relies on subjective assessment, and most of its domains require the ability to stand, thereby limiting scoring range and applicability toward non-ambulatory subjects [3–5]. By contrast, computerized posturography is highly sensitive, but requires costly specialized equipment only found in institutional settings [6–8].
The MK2 is leading the field of affordable, whole-body markerless motion capture technology that is appropriate for home use. Several studies have utilized the MK1 or the MK2 for balance training and assessment and found reliability between both high-end motion capture and clinical scales [40–43], establishing that data from the MK1 and the MK2 is viable for such an approach. Here, we have proven that the MK2 can also make inferences about an individual’s postural stability. Building on this initial work, we will create a platform that can perform computerized balance assessments to identify individuals who are at risk of falling. Stratification of fall risk is related to a decrease in falls, and subsequently, a decrease in fall-related injuries . An automated and computerized fall risk assessment that can be completed safely in a home environment will significantly aid attempts to stratify each user’s fall risk, and devote resources appropriately to fall prevention. This work is a necessary first step in developing such a platform.