This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Serious Games, is properly cited. The complete bibliographic information, a link to the original publication on http://games.jmir.org, as well as this copyright and license information must be included.
Virtual reality (VR) interventions hold great potential for rehabilitation as commercial systems are becoming more affordable and can be easily applied to both clinical and home settings.
In this study, we sought to determine how differences in the VR display type can influence motor behavior, cognitive load, and participant engagement.
Movement patterns of 17 healthy young adults (8 female, 9 male) were examined during games of Virtual Dodgeball presented on a three-dimensional television (3DTV) and a head-mounted display (HMD). The participant’s avatar was presented from a third-person perspective on a 3DTV and from a first-person perspective on an HMD.
Examination of motor behavior revealed significantly greater excursions of the knee (
Differences in visual display type and participant perspective influence how participants perform in Virtual Dodgeball. Because VR use within rehabilitation settings is often designed to help restore movement following orthopedic or neurologic injury, these findings provide an important caveat regarding the need to consider the potential influence of presentation format and perspective on motor behavior.
Virtual reality (VR) has been used to shape motion in patients with various orthopedic and neurologic impairments (eg, low back pain, cerebral vascular accident) for a number of years [
Levin and colleagues have examined the effects of VR environment on motor behavior in both healthy participants and stroke patients [
Individuals with back pain and fear of movement due to perceived risk of harm or injury (ie, kinesiophobia) consistently avoid lumbar flexion [
We recruited 17 healthy young adults (9 male, 8 female) aged 18-35. Exclusion criteria included a history of low back injury, low back pain within the last 6 months, and any orthopedic, neurological, or visual impairment that would prevent participation. This study was approved by the Institutional Review Board of Ohio University, and written informed consent was obtained at the beginning of the session. Using a within-subjects design, participation consisted of standardized reaches to static targets in the real world (RW) and a round of Virtual Dodgeball using two different visual display types (ie, 3DTV, HMD). Each round of dodgeball consisted of three levels of difficulty. Between each level, the participant had to reach to static virtual targets presented at the same locations as the corresponding reaches performed in RW. This manuscript examines the joint excursions used to intercept the launched virtual balls during Virtual Dodgeball gameplay with two different visual display types.
Movement of light-reflective marker clusters attached to the head, upper arms, forearms, hands, trunk, pelvis, thighs, shanks, and feet were tracked using a 10-camera Vicon Bonita system sampled at 100 Hz. This optoelectric-based kinematic system can track the 3D coordinates of light reflective marker clusters attached to the participant with a spatial resolution of 0.1 mm.
The time-series joint angle data were derived from the 3D segment coordinate data using an Euler angle sequence of (1) flexion-extension, (2) lateral bending, and (3) axial rotation [
Participants reached at a comfortable speed holding a regulation dodge ball (24 cm diameter) with both hands. They performed reaches to each of three targets located in the mid-sagittal plane. Target locations were determined for each subject based on their hip height, trunk length, and arm length. The highest target was located such that the subject could, in theory, reach the target by flexing the hips 15° with the shoulder flexed to 90° and the elbow extended. Using the same shoulder and elbow joint positions, the middle and low targets could be reached by flexing the hips 30° and 60°, respectively. Using this individualized method of determining target heights allows for comparison of movement patterns across different individuals [
In brief, one full game was completed in each visual display type. The order of presentation of the visual display type was randomized and counter balanced such that half the participants played Virtual Dodgeball on the 3DTV first and half played Virtual Dodgeball on the HMD first. For each participant, the impact heights of the virtual balls were identical between the visual display types. During Virtual Dodgeball, participants competed against 4 virtual opponents and the object was to block or avoid virtual balls launched randomly by each of the 4 opponents. Participants earned points and cash rewards by successfully blocking launched virtual balls using a ball that they held in their hands or by avoiding a launched ball by ducking (see
Vizard software (WorldViz) was used to develop the virtual environment and control all presented graphics and audio stimuli, including the opposing team’s avatars. The six degrees of freedom kinematic data from the clusters of light reflective markers placed on the participant were streamed to the game environment at 100 Hz using Vicon Tracker software to allow for near real-time presentation of the participant’s avatar (39 ms latency). The MotionMonitor software was used to control bidirectional communication with Vizard, set game parameters and target locations, and record all kinematic data during the experimental testing session.
In the 3DTV condition, a Samsung 1080p 240 Hz 3D Smart LED TV was paired with 3D shutter glasses providing an effective refresh rate of 60 Hz/eye. The participant viewed their slightly translucent avatar from a third-person perspective from a camera position 1.5 meters directly behind their avatar. The translucent avatar allowed for visibility of objects in front of the avatar. The FOV for gameplay with the 3DTV display was as follows: horizontal=50°, vertical=40°. In the HMD condition, the participant viewed their avatar from a first-person perspective that was projected using an Oculus Rift (Oculus Rift Developers Kit 2). From this perspective, the participant viewed their avatar and the environment from the position of the avatar’s eyes. The FOV for the HMD display was as follows: horizontal=100°, vertical=100°, and the refresh rate was fixed at 75 Hz/eye.
The game environment was an indoor basketball arena, with the participant positioned at the free-throw line on one side of the court and the four virtual opponents positioned on the free-throw line on the opposite side of the court. The opposing players moved 3 m fore-aft and 3 m left-right in a random order. Virtual balls were launched every 3.3 ± 0.3 seconds in a randomized order from each of the 4 virtual opponents. The opponent who was about to launch a virtual ball changed color 300 ms prior to launch to alert the participant. If the opponent turned green and the launched ball was yellow, the participant had to attempt to block the ball with the ball held in their hand (co-located with the virtual ball held by the avatar). If the opponent turned red and the launched ball was orange, the participant had to attempt to duck to avoid the ball. A large scoreboard was positioned at the opposite end of the arena (above the opponents) so that participants could track their performance and cash rewards earned. Sound effects were also incorporated, including crowd cheering, buzzers, referee whistles, and a duck quacking sound that occurred whenever an orange ball was launched. An instrumented participant engaged in virtual dodgeball with the HMD is shown in
A round of gameplay consisted of a basic practice level to introduce the scoring metrics and three game levels, each lasting approximately 2 minutes. There were two sets of 15 launched balls within each game level. The intended impact locations of the 15 launched balls were distributed to five impact heights (IH) that were determined by the participant’s height and the amount of lumbar flexion they used during the baseline standardized reaching tasks. For example, during Level 1 of gameplay, the participant could successfully block the virtual ball launched to IH4 (ie, the lowest impact height) simply by using the identical amount of lumbar flexion used in the standardized reaching task to the high target performed at baseline, whereas during Level 3 of gameplay, the participant could successfully block the virtual ball launched to IH4 (ie, the lowest impact height) simply by using the identical amount of lumbar flexion used in the standardized reaching task to the low target performed at baseline. The five impact heights used in gameplay were scaled to impact between the height of participant’s eyes (IH0=highest impact) and approximately their shins (IH4=lowest impact) across the three levels of gameplay (see
Performance was updated in real-time and displayed on the virtual scoreboard, and the participant was awarded progressively more for each successful block or duck at each level of play (Practice Level=1¢, Level 1=2¢, Level 2=5¢, Level 3=10¢). Successful contact for each highlighted ball presented between each set resulted in a bonus 25¢ reward. Conversely, the participant lost cash rewards for each failure to block or duck. Each player started the game with a cash balance on the scoreboard such that if they failed on every launched or presented ball, their cash balance would be zero. The average gameplay session lasted approximately 15 minutes.
Following each session, the participants rated their overall efforts using the NASA Task Load Index (TLX). The NASA TLX is multidimensional assessment that rates perceived workload across to assess system performance [
Participant instrumented and engaged in Virtual Dodgeball using the head-mounted display (HMD).
Methods for computing location of the impact heights (IH0-IH4) of the launched virtual balls for a single game level (left). The distribution of launched virtual balls across the 3 levels of gameplay is shown (right). The lowest impact height (IH4) for each gameplay level (1-3) was calculated from the lumbar spine flexion used to reach the high, middle, and low targets during the baseline standardized reaching tasks.
Because the games were played with both hands in fixed locations on the ball and joint excursions were nearly identical for the left and right limbs, analyses are restricted to the right side. First, time-series position vector of the right index fingertip was smoothed using a 41-point fourth-order Savitzky-Golay filter [
We calculated that we needed 14 participants to determine the within-subject effects of display type with 80% power, assuming alpha=.05, and correlation between measures of .5 and an effect size of
There was a main effect of Display Type on movement time (
Effect of 3D television (3DTV) versus a head-mounted display (HMD) on movement time for each impact height (IH).
As illustrated in
Effects of 3D television (3DTV) versus a head-mounted display (HMD) on hand position at target intercept for each impact height (IH) along the anterior-posterior (AP) axis, along the medial-lateral (ML) axis, and along the vertical axis.
There were significant interactions of Display Type by IH for on joint excursions of the ankle (
Interaction of 3D television (3DTV) versus head-mounted display (HMD) by impact height (IH) on the joint excursions of the ankle, knee, hip, spine, shoulder, and elbow.
Effects of 3D television (3DTV) versus head-mounted display (HMD) on the posture adopted at target intercept.
There was a significant interaction of Display Type and IH on displacement of COM along the AP axis (
Effects of 3D television (3DTV) versus a head-mounted display (HMD) on displacement of whole-body center-of-mass (COM) for each impact height (IH) along the anterior-posterior (AP) axis, along the medial-lateral (ML) axis, and along the vertical axis.
Analysis of the Task Load Index data (
Effects of 3D television (3DTV) versus head-mounted display (HMD) on NASA Task Load Index(TLX) scores.
The primary goal of this study was to determine the effects of display type on the joint excursions used while playing Virtual Dodgeball.
For virtual balls launched to IH1-4, hand position at target intercept contact was about 14 cm forward (AP axis) and 18 cm lower (vertical axis) in gameplay with HMD compared to 3DTV (
While differences in FOV do not provide an explanation for the differences in hand position or joint excursions, perhaps COM displacement can explain the changes in strategy between display types. The vertical displacement for the COM was greater in HMD compared to 3DTV across IH1-4. From an energetics perspective, lowering the height of the COM could result in a fundamentally more stable system. However, the forward displacement of the whole-body COM was also greater in the HMD condition compared to 3DTV. Thus, from the same energetics perspective, the greater forward displacement would not lead to a more stable system. It has been shown that COM displacement is changed in standing reaching tasks performed in virtual reality environments when the viewing angle of the participant’s avatar is altered [
It is possible that differences in joint excursions between the visual displays could be driven, in part, by differences in movement time to intercept the virtual launched balls. We have shown that joint excursions of the ankle, knee, and hip increase as movement time to target is reduced by half (ie, when participants move twice as fast to the target) [
Another potential contributor to the differences in observed movement strategies is the difference in refresh rates for the two display types. However, the kinematic input streams to the avatars in both display types is 100 Hz with the 3DTV, paired with the shutter goggles, having an effective update rate of 60 Hz/eye and the HMD having an update rate of 75 Hz. Thus the difference in refresh rates results in an absolute time difference of about 3.3 ms. Further, as both display types use an LED display, there should be no differences in persistence of the displayed images. Finally, according to Ware, the processing time for humans is approximately 166 ms and visual lags effect performance at about 200 ms [
The difference in movement strategies observed between the two display types could be due to the presentation of the avatar. Some investigations have reported that the sense of actual presence in VR was weakened when the avatar was viewed from a third-person perspective [
Finally, the results of this study provide further support for Virtual Dodgeball as an effective strategy to promote lumbar flexion. Importantly, the current findings also indicate that the clinical utility of Virtual Dodgeball may be enhanced with an HMD because it elicits more lumbar spine flexion, greater participant satisfaction with overall performance, and less frustration.
A limitation of this study is that it cannot assign differences in motor performance in these tasks simply to avatar perspective. The differences could also be due to the use of a 3DTV versus the HMD, or to differences in display of the avatar (ie, first-person versus third-person perspective). Accordingly, future studies are needed to carefully isolate the effects of perspective and display type.
The results of this study demonstrate that visual display type influences motor behavior in Virtual Dodgeball. These data are important for the development of virtual reality assessment and treatment tools that are becoming increasingly practical for home and clinic use. Because a primary goal of virtual reality within rehabilitation is often to restore movement following orthopedic or neurologic injury, it is important to understand how presentation of the avatar or, by extension, camera position will affect motor behavior regardless of the display through which it is presented (ie, 3DTV or HMD). Use of home devices such as the Kinect sensor to track and presents an avatar in a third-person perspective may result in very different motor behavior when compared to the same tasks being presented from a first-person perspective.
Video files.
The NASA-TLX survey instrument.
Supplementary tables.
three-dimensional television
anterior posterior
center of mass
field of view
head-mounted display
medial lateral
virtual reality
task load index
CRF and JST obtained funding for the research; contributed to developing the intervention and study protocols; reviewed, edited, and approved the final version of the paper; and accept full responsibility regarding the conduct of the study and integrity of the data. JST, CRF, MEA, STL, PEP, and SW have read and approved the final manuscript. This study was funded, in part, by the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of Health under Award Number R21AR064430. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
None declared.