Manipulation of visual biofeedback during gait with a time delayed adaptive Virtual Mirror Box

Area of Science:

• Rehabilitation Engineering
• Biomedical Engineering
• Human Movement Science

Background:

• Physical mirrors have aided phantom limb pain and movement control by creating a mirrored illusion.
• Limitations of physical mirrors include inability to sync with real-time gait and only showing parallel limb movement.
• Virtual reality offers a solution to overcome these limitations.

Purpose of the Study:

• To develop and evaluate a Virtual Mirror Box (VMB) system.
• To ensure temporal congruency between impaired limb movement and its mirrored virtual representation during gait.
• To adapt VMB for real-time treadmill use with delayed visual feedback.

Main Methods:

• Developed a VMB application using CAREN system's D-Flow software for real-time motion capture and mirroring.
• Implemented a continuous, calculated time delay for virtual limbs to match gait phase.
• Utilized dynamic morphing to align physical and virtual limb contact events, accommodating pathological gait variations.
• Tested on a trans-femoral amputee, comparing intact and mirrored limb joint angles (elbow, knee) via cross-correlation, RMSE, and correlation coefficients.

Main Results:

• The time-delayed adaptive VMB successfully created a symmetrical gait appearance in the avatar.
• Significant reduction in differences between intact and virtual limb joint angles observed (knee: 10.86° to 4.99°, elbow: 5.34° to 2.54°).
• Dynamic morphing showed minimal, non-significant changes in toe-off events compared to a fixed delay, with a slight, significant increase in initial contact events (p < 0.05).

Conclusions:

• Adaptive time delay in VMB extends its application to treadmill gait analysis and therapy.
• Dynamic morphing balances accurate mirroring with synchronization of gait events.
• Repeatable, asymmetrical gait patterns are expected to yield more precise mirroring in future applications.