Enhancing the Measurement of Clinical Outcomes Using Microsoft Kinect

Abstract

Abstract-There is a growing body of applications leveraging Microsoft Kinect and the associated Windows Software Development Kit in health and wellness. In particular, this platform has been valuable in developing interactive solutions for rehabilitation including creating more engaging exercise regimens and ensuring that exercises are performed correctly for optimal outcomes. Clinical trials rely upon robust and validated methodologies to measure health status and to detect treatment-related changes over time to enable the efficacy and safety of new drug treatments to be assessed and measured. In many therapeutic areas, traditional outcome measures rely on subjective investigator and patient ratings. Subjective ratings are not always sensitive to detecting small improvements, are subject to inter-and intra-rater variability and limited in their ability to record detailed or subtle aspects of movement and mobility. For these reasons, objective measurements may provide greater sensitivity to detect treatment-related changes where they exist.In this review paper, we explore the use of the Kinect platform to develop low-cost approaches to objectively measure aspects of movement. We consider published applications that measure aspects of gait and balance, upper extremity movement, chest wall motion and facial analysis. In each case, we explore the utility of the approach for clinical trials, and the precision and accuracy of estimates derived from the Kinect output.We conclude that the use of games platforms such as Microsoft Kinect to measure clinical outcomes offer a versatile, easy to use and low-cost approach that may add significant value and utility to clinical drug development, in particular in replacing conventional subjective measures and providing richer information about movement than previously possible in large scale clinical trials, especially in the measurement of gross spatial movements. Regulatory acceptance of clinical outcomes collected in this way will be subject to comprehensive assessment of validity and clinical relevance, and this will require good quality peer-reviewed publications of scientific evidence.

Suitability of the Kinect Sensor and Leap Motion Controller—A Literature Review

“…Breedon et al [ 74 ] did a comparison between the two Kinect versions, the Intel RealSense SR300, Xtion Pro Live. The latter did not offer many improvements over the Kinects, but it has a better depth camera.…”

“…Breedon et al [ 74 ] compared the LMC and the Creative SENZ3D. The SENZ3D has multiple advantages over the LMC: it can be controlled by voice as it has a dual array microphone.…”

Supervised Classification of Motor-Rehabilitation Body Movements with RGB Cameras and Pose Tracking Data

“…In literature, there are many implementations of systems for rehabilitation using Kinect devices (Breedon et al, 2016), which were introduced in 2010 by Microsoft, and other devices with depth sensors, such as Intel RealSense and Leap Motion for hand tracking. Cameras with depth sensors can be more precise, but their costs and availability also increase.…”

Correlation between MOVA3D, a Monocular Movement Analysis System, and Qualisys Track Manager (QTM) during Lower Limb Movements in Healthy Adults: A Preliminary Study

“…The majority of studies carried out in the field of rehabilitation aim to engage patients in an exercise regime proposed by the physiotherapist. These systems enable patients to receive real-time feedback on the prescribed activity, assist in the correction of joint positions during the execution of the movements, and aid in remote monitoring and adjustment of the exercise prescription between each clinical visit to the patient [14].…”