Validity of the Kinect and Myo armband in a serious game for assessing upper limb movement

Sadeghi Esfahlani, Shabnam and Muresan, Bogdan and Sanaei, Alireza and Wilson, George (2018) Validity of the Kinect and Myo armband in a serious game for assessing upper limb movement. Entertainment Computing, 27. pp. 150-156. ISSN 1875-953X

[img]
Preview
Text
Accepted Version
Available under the following license: Creative Commons Attribution Non-commercial No Derivatives.

Download (7MB) | Preview
Official URL: http://dx.doi.org/10.1016/j.entcom.2018.05.003

Abstract

A cost-effective, easily-accessible neuro-motor rehabilitation solution is proposed that can determine the range of motion and the kinematic ability of participants. A serious game comprising four-scenarios are developed in which the players control an avatar that mirrors the rotations of the upper-limb joints through multi-channel-input devices (Kinect, Myo, FootPedal). Administered functional reach tests (FRT) challenge the player to interact with a 3D-environment while standing or sitting and using the FootPedal which simulates the action of walking whilst body movement is measured concurrently. The FRT’s complexity level is adapted using a Monte Carlo Tree Search algorithm which determines a virtual object’s position based on the proved ability of the user. Twenty-three volunteers were recruited to play the game in 45-minute sessions. The data show that the system has a more positive impact on players performance and is more motivating than formal therapy. The visual representation of the trajectory of the objects is shown to increase the perception of the participants voluntary/involuntary upper extremity movement, and the results show a comparable inter-session reliability (acceptable-good) over two repeated sessions. A high Pearson correlation demonstrates the validity of using Kinect and Myo devices in assessing upper-limb rehabilitation, and the timing and the clinically relevant movement data have a higher accuracy when the devices are paired.

Item Type: Journal Article
Keywords: Kinect v2, Monte Carlo Tree Search (MCTS), Myo Armband, FootPedal
Faculty: Faculty of Science & Technology
Depositing User: Dr Shabnam Sadeghi Esfahlani
Date Deposited: 30 May 2018 12:13
Last Modified: 29 May 2019 01:02
URI: http://arro.anglia.ac.uk/id/eprint/702694

Actions (login required)

Edit Item Edit Item