Virtual reality games for rehabilitation of people with stroke: perspectives from the users

  • Published on

  • View

  • Download

Embed Size (px)



    Virtual reality games for rehabilitation of people with stroke:perspectives from the users


    Health and Rehabilitation Research Institute, AUT University, Auckland, New Zealand

    Accepted March 2011

    AbstractPurpose. The purpose of this study is to evaluate the feasibility and users perspectives of a novel virtual reality (VR) game-based rehabilitation intervention for people with stroke.Method. Six people with upper limb hemiplegia participated in a 6-week intervention that involved VR games. A series ofeight progressively complex games was developed that required participants to navigate a submarine in a virtual oceanenvironment. Movement of the submarine was directed by forces applied to an arm interface by the affected limb. Outcomemeasures included assessments of arm function, questionnaires evaluating the intervention and a semi-structured interviewconcerning the participants opinion of the intervention.Results. All participants improved their performance on the games, although there were limited changes in clinical measuresof arm function. All participants reported that they enjoyed the intervention with a wide range of overall perceptions of theexperience of using VR. Three themes emerging from the interview data were: stretching myself, purpose and expectations ofthe intervention and future improvements.Conclusions. Participants found that taking part in this pilot study was enjoyable and challenging. Participants feedbacksuggested that the games may be motivating and engaging for future users and have provided a basis for further developmentof the intervention.

    Keywords: Hemiplegia, upper limb, rehabilitation, virtual reality


    Spontaneous recovery of movement function follow-

    ing a stroke normally completes by 6 months [1].

    However, further improvements in movement ability

    have been demonstrated even in chronic stroke

    following the application of rehabilitation interven-

    tions [24]. Despite this continued potential for

    recovery, patients motivation and adherence to

    treatment protocols are major barriers to recovery

    [5]. In a survey by Page et al. [6], it was found that

    almost 70% of stroke survivors were unlikely to

    participate in a specific treatment protocol even

    when it was proven to be effective. The major

    concerns expressed by patients and therapists were

    scheduling time, safety and availability of appropriate

    resources. The development of safe rehabilitation

    interventions that target motivation is likely to

    enhance patients time dedicated to treatment

    protocols and consequently facilitate the recovery

    of movement function.

    Virtual reality (VR) games aim to create fun

    environments that engage users and increase their

    motivation and have been used in a range of

    contexts, including rehabilitation [710]. The appli-

    cation of VR in movement rehabilitation affords a

    number of benefits over standard treatment options.

    VR provides a controlled environment in which users

    can acquire confidence in performing tasks that may

    otherwise compromise safety. Virtual environments

    can easily be adapted to the needs of the individual to

    provide a more target-oriented rehabilitation that

    focuses on the specific requirements of the user [11].

    The ability to provide controlled or augmented

    feedback is a further advantage of VR systems.

    Provision of feedback is critical for learning and

    rehabilitation [12,13], and visual or auditory feed-

    back can easily be manipulated in VR environments

    Correspondence: Dr. Gwyn Lewis, Health and Rehabilitation Research Institute, AUT University, Private Bag 92006, Auckland 1142, New Zealand.

    Tel: 64-9-921-9999. E-mail:

    Disability and Rehabilitation: Assistive Technology, September 2011; 6(5): 453463

    ISSN 1748-3107 print/ISSN 1748-3115 online 2011 Informa UK, Ltd.DOI: 10.3109/17483107.2011.574310






    ist T


    ol D



    d fr

    om in





    om b

    y U















  • to facilitate the learning experience [14]. Finally,

    repetition is a key ingredient in almost all areas of

    motor rehabilitation [15,16]; therefore, the capacity

    to provide an ever-changing environment that con-

    tinues to train the same underlying deficit is a major

    advantage over real-world situations.

    Several studies have assessed the use of VR

    technologies in people with chronic movement

    impairments following stroke. These have adopted

    a variety of custom-designed and commercially

    available systems incorporating head-mounted de-

    vices [10,17], large-scale video projections [18,19]

    and desktop computer or television displays [2025].

    Almost all of these studies have reported improve-

    ments in movement ability following VR training,

    both in terms of improved performance on VR

    training tasks [7,2527] and improvements in more

    standardised clinical rating scales [9,20,22,26].

    However, there is a lack of in-depth and structured

    information on the users perspectives of these VR

    interventions. Such information could be used for

    future development of systems that more specifically

    meet the needs of the end users or indicate

    populations that respond more positively to VR-

    based rehabilitation.

    The aims of the current feasibility study were

    three-fold. First, we wanted to design a VR-based

    intervention to improve the upper limb movement in

    people with stroke. Up to 60% of people with chronic

    stroke are reported to have a residual deficit of arm

    function [28,29]. We aimed to develop a series of

    VR games that incorporated motor learning princi-

    ples of progression, feedback and goal attainment

    that encouraged independent control of upper limb

    joints. The second aim was to determine the effects

    of the developed intervention on arm function. The

    final aim was to determine the users perspectives of

    the intervention. This included how the intervention

    influenced their arm function, the usability of the

    system, enjoyment levels and any alterations that

    would make the system more appealing.


    A prospective feasibility study with both quantitative

    and qualitative data was conducted. The study

    involved a small sample to develop the game with

    feedback from participants, identify the potential

    impact and acceptability and inform future sample



    Six people with hemiparesis, following a single,

    monohemispheric stroke, participated in the study

    (see Table I for details). Participants were required

    to be at least 6-months post-stroke, have a residual

    deficit in upper limb function, have normal or

    corrected-to-normal hearing and vision and have

    no cognitive or perceptual deficits that impaired their

    ability to interact with a virtual environment or might

    increase their susceptibility to cybersickness. Tar-

    geted sampling of individuals was used to include

    people with a range of ability levels and experience

    with computers. Ethical approval for the study

    was received from the local Ethics Committee, and

    informed written consent was obtained prior to



    Participants were seated in front of a computer

    monitor with their impaired arm positioned at 458shoulder abduction, elbow at 908 and the forearmfully pronated (Figure 1). The forearm and wrist

    were secured in a padded fibreglass cast. The cast

    was mounted on a six degree-of-freedom load cell

    (67M50A, JR3 Inc, USA) and bolted to a height

    adjustable table. Application of forces and torques

    to the load cell controlled the movement of a

    submarine shown on the computer monitor, such

    that the submarine moved or rotated in the same

    direction as the forces and torques applied by the


    VR games

    The VR intervention consisted of a series of eight

    games in which the submarine was manoeuvred

    through an ocean environment. Three of the games

    required users to move the submarine in single

    planes of movement up and down, forward and

    backward or side to side to randomly selected

    targets. Users scored points for each target they

    reached within a 3-min game. The difficulty of these

    games was progressed by increasing the sensitivity to

    Table I. Participants characteristics.

    ID Age Gender




    side DASH

    S1 55 M 9.5 R 48

    S2 60 M 6.9 L 18

    S3 75 F 2.0 L 30

    S4 69 M 1.4 L 47

    S5 71 M 1.4 R 53

    S6 69 M 1.4 R 58

    DASH, disabilities of the arm, shoulder and hand (maximum

    100); M, male; F, female; R, right; L, left.

    454 G. N. Lewis et al.






    ist T


    ol D



    d fr

    om in





    om b

    y U















  • forces in the target plane, and then by increasing the

    planes of movement in which the participants could

    move the submarine. The remaining five games

    required the users to move the submarine in multiple

    planes of movement. Users accumulated points by

    moving the submarine to different targets or were

    timed in their completion of a set course. These

    games were again progressed by increasing the

    sensitivity to forces in specific planes. In this way,

    participants progressed through several levels of

    difficulty and received feedback in the form of scores

    or times at each of these levels.

    Intervention protocol

    Each participant was asked to attend three 1-hour

    sessions per week for a 6-week period. During that

    time, the participants were individually progressed

    through the games with the difficulty of each game

    being increased based on their performance. Three

    to four games were selected to be played in each

    session. Participants were given graphical feedback of

    their progress in each game after 3 and 6 weeks of the


    Outcome measures

    The feasibility of the VR games as a rehabilitation

    intervention was assessed by the participants atten-

    dance at the intervention sessions, performance

    (scores/times) in the individual games and progres-

    sion through the series of increasingly more difficult

    games. Arm function was assessed using the

    Fugl-Meyer Assessment [30] and the Box and Block

    Test [31].

    Participants perspectives of the intervention were

    assessed using a post-intervention questionnaire and

    an interview. The post-intervention questionnaire

    was adapted from Merians et al. [9], who assessed

    the use of a VR intervention to improve hand

    function in people with stroke. Although the

    reliability and validity of the post-intervention ques-

    tionnaire have not been established, the questions

    were selected and modified from a validated and

    reliable questionnaire used by industry usability

    laboratories to evaluate user interfaces [32]. Minor

    modifications to the measure for our study involved

    the inclusion of terminology relating to the sub-

    marine game and were made to ensure the questions

    were specific to the intervention being evaluated.

    Each participant was also interviewed by an

    independent researcher after completion of the

    intervention. Interviews were semi-structured, with

    the researcher asking open-ended questions around

    topics including: what aspects, if any, of the

    programme that were liked or disliked; comfort

    during and after the sessions; comparison with other

    upper limb rehabilitation; willingness for future

    participation and any suggested improvements.

    Participants perceptions of change in arm move-

    ment and function during and outside session times

    were also canvassed. The independent researcher

    conducted all interviews, each taking 3045 min to

    complete. All interviews were recorded, subse-

    quently transcribed verbatim and rechecked for

    accuracy following transcription against the audio

    file. In keeping with qualitative descriptive methods

    [33], content analysis was applied [34]. Transcripts

    were coded line by line into codes, and these codes

    were grouped into categories of similar items. NVivo

    8 software (QSR International, Australia) was used

    to store the data, record coding and any memos

    associated with the interviews. Categories identified

    were compared between participants and were

    checked back against the data looking for statements

    that qualified, expanded, countered or supported

    them. In this way, themes that provided an explana-

    tion of participants experiences emerged from the

    data. Validation of the results was enhanced by

    rigorous checking back of emerging themes against

    the data, peer review of the results and member


    Figure 1. Set-up for the VR games. (A) The participants inserted

    their arms into the interface. (B) Forces detected by the load cell

    (incorporated into the interface) controlled the movement of a

    submarine through virtual environments displayed on a computer


    Perspectives on virtual reality games for stroke 455






    ist T


    ol D



    d fr

    om in





    om b

    y U















  • Results


    All six participants attended all 18 sessions over the

    6-week intervention period. The five most basic

    games were able to be completed by all participants.

    Three of the participants attempted all of the

    available games; however, the remaining participants

    were unable to control the submarine in the multiple

    planes required to complete the more complex

    games. In all games attempted, each participant

    demonstrated improvements in performance over

    time with regard to times or scores. As each

    participants programme was progressed individually

    by introducing more complex games and modifying

    the difficulty within each game, statistically mean-

    ingful descriptions of improvements over time

    cannot be made as progressions in difficulty routi-

    nely resulted in a subsequent decrement in scores or

    times. Example charts of performance for one

    participant are shown in Figure 2. This participant

    clearly shows progressive improvements in the games

    over time that are modulated by increasing the

    difficulty of the games.

    Arm function

    Table II shows the pre- and post-intervention Fugl-

    Meyer Assessment and Box and Block scores for

    each participant. One participant (S5) improved by

    five points on the Fugl-Meyer Assessment, but there

    were no other clinically significant changes in arm

    function in any individuals.

    Participants perspectives

    Results from the post-intervention questionnaire

    indicated that most of the participants enjoyed the

    intervention, believed that their arm function im-

    proved during the treatment period and rated

    the control and usability of the games as high.

    Individual responses for each question are shown in

    Table III.

    Three themes emerged from the interview data

    that appeared to capture the breadth of participants

    experiences. These were: stretching myself, pur-

    pose and expectations and future improvements.

    Each theme with contributing categories is described

    in turn. How these results relate to current literature

    on experiences of people using new technologies

    following stroke is canvassed in the subsequent

    discussion. Quotes in support of the analysis have

    been anonymised by use of random alphabetical

    identifiers and gender neutral lang...


View more >