Viewing medium affects arm motor performance in 3D virtual environments

Abstract

Background: 2D and 3D virtual reality platforms are used for designing individualized training environments for post-stroke rehabilitation. Virtual environments (VEs) are viewed using media like head mounted displays (HMDs) and large screen projection systems (SPS) which can influence the quality of perception of the environment. We estimated if there were differences in arm pointing kinematics when subjects with and without stroke viewed a 3D VE through two different media: HMD and SPS.

Methods: Two groups of subjects participated (healthy control, n=10, aged 53.6 ± 17.2 yrs; stroke, n=20, 66.2 ± 11.3 yrs). Arm motor impairment and spasticity were assessed in the stroke group which was divided into mild (n=10) and moderate-to-severe (n=10) sub-groups based on Fugl-Meyer Scores. Subjects pointed (8 times each) to 6 randomly presented targets located at two heights in the ipsilateral, middle and contralateral arm workspaces. Movements were repeated in the same VE viewed using HMD (Kaiser XL50) and SPS. Movement kinematics were recorded using an Optotrak system (Certus, 6 markers, 100 Hz). Upper limb motor performance (precision, velocity, trajectory straightness) and movement pattern (elbow, shoulder ranges and trunk displacement) outcomes were analyzed using repeated measures ANOVAs.

Results: For all groups, there were no differences in endpoint trajectory straightness, shoulder flexion and shoulder horizontal adduction ranges and sagittal trunk displacement between the two media. All subjects, however, made larger errors in the vertical direction using HMD compared to SPS. Healthy subjects also made larger errors in the sagittal direction, slower movements overall and used less range of elbow extension for the lower central target using HMD compared to SPS. The mild and moderate-to-severe sub-groups made larger RMS errors with HMD. The only advantage of using the HMD was that movements were 22% faster in the moderate-to-severe stroke sub-group compared to the SPS.

Conclusions: Despite the similarity in majority of the movement kinematics, differences in movement speed and larger errors were observed for movements using the HMD. Use of the SPS may be a more comfortable and effective option to view VEs for upper limb rehabilitation post-stroke. This has implications for the use of VR applications to enhance upper limb recovery.

© 2011 Subramanian and Levin; licensee BioMed Central Ltd.

Figures

Figure 1

The virtual environment scene, viewed through the head-mounted display and the large screen projection during arm pointing tasks, consisting of six targets arranged in two rows in the ipsilateral, central and contralateral arm workspace. The subject's finger position was indicated by the blue dot (circle in centre of targets).

Figure 2

Overall endpoint (root mean square, RMS) errors (A), endpoint velocity (B) and trunk displacement (C) for the three groups: healthy (blue), stroke-mild (red) and stroke-moderate-to-severe (black). Data are overall mean (SD) values across all 6 targets and both media. Asterisks indicate significance. * p < 0.01, ** p < 0.005

Figure 3

Mean (SD) values across targets for total endpoint (root mean square, RMS) (A), vertical (B) and sagittal directional errors (C), endpoint velocities (D) and elbow extension ranges (E) for the three groups: healthy, stroke-mild and stroke-moderate-to-severe viewing the VE through the head mounted display (light blue squares) and screen projection system (orange circles). Asterisks indicate significance. * p < 0.01, ** p < 0.005

Figure 4

Vertical (top row) and sagittal (bottom row) directional errors for the 6 targets for the three groups: healthy (blue circles), stroke-mild (red) and stroke-moderate-to-severe (black). Subjects viewed the VE through the head mounted display (light blue squares) and screen projection system (orange circles). Data are mean values in each group. Targets: UC - upper contralateral, UM - upper middle, UI - upper ipsilateral, LC - lower contralateral, LM - lower middle, LI - lower ipsilateral.