Elsevier

Gait & Posture

Volume 36, Issue 3, July 2012, Pages 372-377
Gait & Posture

Validity of the Microsoft Kinect for assessment of postural control

https://doi.org/10.1016/j.gaitpost.2012.03.033Get rights and content

Abstract

Clinically feasible methods of assessing postural control such as timed standing balance and functional reach tests provide important information, however, they cannot accurately quantify specific postural control mechanisms. The Microsoft Kinect™ system provides real-time anatomical landmark position data in three dimensions (3D), and given that it is inexpensive, portable and simple to setup it may bridge this gap. This study assessed the concurrent validity of the Microsoft Kinect™ against a benchmark reference, a multiple-camera 3D motion analysis system, in 20 healthy subjects during three postural control tests: (i) forward reach, (ii) lateral reach, and (iii) single-leg eyes-closed standing balance. For the reach tests, the outcome measures consisted of distance reached and trunk flexion angle in the sagittal (forward reach) and coronal (lateral reach) planes. For the standing balance test the range and deviation of movement in the anatomical landmark positions for the sternum, pelvis, knee and ankle and the lateral and anterior trunk flexion angle were assessed. The Microsoft Kinect™ and 3D motion analysis systems had comparable inter-trial reliability (ICC difference = 0.06 ± 0.05; range, 0.00–0.16) and excellent concurrent validity, with Pearson's r-values >0.90 for the majority of measurements (r = 0.96 ± 0.04; range, 0.84–0.99). However, ordinary least products analyses demonstrated proportional biases for some outcome measures associated with the pelvis and sternum. These findings suggest that the Microsoft Kinect™ can validly assess kinematic strategies of postural control. Given the potential benefits it could therefore become a useful tool for assessing postural control in the clinical setting.

Highlights

► We examined the validity of the Microsoft Kinect for assessing postural control. ► Data were collected during functional reach and standing balance tasks. ► Results from the Microsoft Kinect and a 3D motion analysis system were compared. ► Comparable inter-trial reliability and excellent concurrent validity were observed.

Introduction

The assessment of postural control is commonly undertaken in laboratory and clinical settings for a wide range of pathologies, and the ability to perform well on many of these tests has been linked to factors such as physical function and falls risk [1], [2], [3], [4]. Measurement tools for assessing postural control range from simple, time-based assessments through to full-body kinematic and kinetic examinations [5]. In regard to clinic-based assessments, three of the most commonly performed postural control tests are single leg stance time, the Berg Balance Scale and timed up and go [6]. Although the Berg Balance Scale includes visually determined assessments of quality of movement, a common component of these tests is their limitation to either timing or reach-based outcome measures. While providing useful information to the clinician, they are prone to ceiling effects and often cannot accurately quantify the postural control strategies being used by the patient [5], [7]. Adding more advanced data collection and analysis tools such as force platforms and three dimensional (3D) camera systems allows for the same tests to be analyzed in finer detail. For example, in addition to measuring hand displacement during a functional reach test, a 3D camera system can be incorporated into the testing protocol to measure spatiotemporal factors such as trajectories of movement, which have been shown to discriminate between neurological and healthy populations [8]. In addition, determining patient specific movement and stability techniques may uncover maladaptive strategies for maintaining balance. For example, someone with disease or injury-related deficiencies at their feet and legs may successfully maintain their standing balance for the entire duration of a clinical postural control assessment test. However, they may achieve this by utilizing a hip-based postural control strategy in the absence of adequate proprioceptive input from the ankle and knee [9].

While inexpensive devices such as the Nintendo Wii Balance Board™, a clinically feasible alternative to a force platform [10], can provide postural control information related to function [11], it cannot accurately differentiate joint movements. This is most commonly achieved using systems that require multiple cameras and tracking markers placed on the skin, making them cumbersome to house and transport, expensive and requiring extensive technical expertise to operate and interpret. This combination of factors precludes their use in all but the major clinical centers and research laboratories. However, a recent development in computer gaming technology – the Microsoft Kinect™ – is inexpensive, portable and does not require markers to determine anatomical landmarks, and consequently may overcome the limitations associated with laboratory-based movement analysis systems.

The Microsoft Kinect™ incorporates infra-red light and a video camera to create a 3D map of the area in front of it [12], and uses a randomized decision forest algorithm to automatically determine anatomical landmarks on the body, such as joint centers, in close to real time [13]. The results of previous studies are promising, and have shown that the depth sensor itself is accurate for assessing 3D position in a workplace environment [14], and that joint centers derived from the Microsoft Kinect™ can be used to classify dance gestures [15]. If the positions of these reported anatomical landmarks are found to be accurate during the assessment of postural control, this could facilitate advanced analysis of these tests to be performed in the clinical setting. Consequently, the aim of this study was to assess the concurrent validity of the anatomical landmarks collected using the Microsoft Kinect™ with a well-established kinematic assessment tool (i.e. a 3D camera-based motion analysis system) during three commonly performed standing postural control assessment tests – single leg standing balance and the forward and lateral reach tests.

Section snippets

Subjects

Twenty young, injury free individuals (age: 27.1 ± 4.5 yr, height: 173.7 ± 10.3 cm, mass: 71.7 ± 11.0 kg, male = 10) with no history of neurological conditions or medication use that would have influenced their postural control volunteered to participate. This study was approved by the institution's Human Research Ethics Committee and all subjects provided informed consent.

Procedures

Subjects were required to wear tight-fitting shorts and an upper body garment (t-shirt, singlet or for males the option to perform

Results

The inter-trial reliability measures for each system are provided in Table 1. Overall, the outcome measures from the Microsoft Kinect™ and 3D camera methods were comparable in terms of absolute and relative test–retest reliability (ICC difference ≤0.16; ratio CV difference ≤11.6%), with pairwise comparison of ICC values showing no significant differences between devices (95% bootstrap confidence intervals included zero). The mean results for each system, and the correlation and significant bias

Discussion

The ability to differentiate postural control strategies using an inexpensive, portable and widely available system could provide clinical and research benefits in a variety of patient populations. Our results suggest that the Microsoft Kinect™ provides anatomical landmark displacement and trunk angle data which possesses excellent concurrent validity when compared to data obtained from a 3D camera-based motion analysis system. Given that a number of previous studies have shown a link between

Source of funding

No external funding was received specifically for this project. No support was received from Microsoft other than the use of the freely available official software development kit. The first author's research fellow position is funded by ASICS Oceania.
Conflict of interest statement

There is no conflict of interest.

References (26)

  • K.M. Sibley et al.

    Balance assessment practices and use of standardized balance measures among Ontario physical therapists

    Physical Therapy

    (2011)
  • L. Blum et al.

    Usefulness of the Berg Balance Scale in stroke rehabilitation: a systematic review

    Physical Therapy

    (2008)
  • C.T. Bonnet et al.

    Proximal postural control mechanisms may be exaggeratedly adopted by individuals with peripheral deficiencies: a review

    Journal of Motor Behavior

    (2011)
  • Cited by (0)

    View full text