An Interactive Computer Game for Improving Selective Voluntary Motor Control in Children With Upper Motor Neuron Lesions: Development and Preliminary Feasibility Study

Abstract

Background: Computer game-based interventions are emerging in pediatric neurorehabilitation, as they can provide two key elements for motor learning-motivating environments that enable long-term compliance, which is particularly relevant for children, and augmented feedback for improving movement performance.

Objective: The overall aim of this study is to develop an interactive computer play for children with upper motor neuron lesions to train selective voluntary motor control and give particular attention to motivation and feedback. We also aim to determine features that make games engaging, investigate which sensory feedback modality is noticed the fastest during play, develop an interactive game, and evaluate its feasibility.

Methods: We identified engaging game features by interviewing 19 children and adolescents undergoing rehabilitation. By using a test version of the game, we determined the response times of 10 patients who had to react to visual, auditory, or combined feedback signals. On the basis of the results of these two subprojects, we developed and designed a game environment. Feasibility was studied in terms of the practicability and acceptability of the intervention among 5 children with upper motor neuron lesions.

Results: The game features deemed the most important by pediatric patients were strategic gameplay (13/29, 45% of answers) and choice (6/29, 21%). While playing the game, an acoustic alarm signal (reaction time: median 2.8 seconds) was detected significantly faster (P=.01) than conditions with other feedback modalities (avatar velocity reduction: median 7.8 seconds; color desaturation: median 5.7 seconds). Most children enjoyed playing the game, despite some technical issues.

Conclusions: The careful identification of game features that increase motivation and feedback modalities that inform children quickly led to the development of an interactive computer play for training selective voluntary motor control in children and adolescents with upper motor neuron lesions.

Keywords: augmented feedback; game therapy; mobile phone; motivation; rehabilitation; virtual reality.

Figure 1

Preliminary game environment and schematic illustration of feedback modes for testing different feedback signal modalities. (A) Game environment in which the avatar had to be navigated up and down. (B) Baseline mode where the avatar’s speed was reduced. (C) In the acoustic mode, the feedback signal was an alarm sound of increasing sound volume. (D) Illustration of the conversion from color to gray scale in the background color mode.

Figure 2

The upper part shows the organization of one trial, illustrating the start of the trial, the feedback initiation (ON) and increment, and the start of the next trial. The lower part of the figure illustrates the changes that occurred as soon as the participant responded to the feedback signal and how the reaction time was defined. RT: reaction time.

Figure 3

Game environment. The purpose of the game was to collect stars or coins and avoid the obstacles by moving the avatar up and down on the screen. (A) This screenshot from a game level shows a shield power-up in the top-left corner. After collecting it, the avatar was protected from obstacles for the next 10 seconds. (B) This level had a different theme, although the game elements stayed the same. (C) An example of challenges in one level. Each challenge required a different focus. (D) Players could use the collected coins to personalize the avatar.

Figure 4

Game controllers. (A) Electromyography-based control of the game via activation of the right tibialis anterior muscle (green cables). Mirror activity on the left side (red cables) would trigger the feedback signal. (B) ArmeoSenso with its three sensors (chest, upper arm, and lower arm). In this example, the game was controlled by elbow flexion and extension without simultaneously moving the shoulder (ie, shoulder abduction would lead to a warning feedback signal).

Figure 5

Boxplots of the reaction times for each feedback mode (1-4). In the baseline condition, only the speed of the avatar was reduced. In the other conditions, velocity reduction was combined with an additional feedback signal, that is, an acoustic alarm or a change of the background color. The colors represent individual participants. The small circles show the reaction time of each individual trial, whereas diamonds show the mean reaction time over all 5 trials per participant.

Figure 6

Participants’ ratings of enjoyment. Scores indicated how much they liked the game and whether they would like to play it again. Only sessions with the more appropriate system are shown.