GLIMPSE: Google Glass interface for sensory feedback in myoelectric hand prostheses

Area of Science:

• Prosthetics and Bionics
• Human-Computer Interaction
• Rehabilitation Engineering

Background:

• Providing sensory feedback to prosthesis users is crucial for effective control.
• Traditional tactile feedback methods have limitations in bandwidth and require extensive training.
• Augmented reality (AR) presents a novel alternative for delivering sensory information.

Purpose of the Study:

• To develop and evaluate GLIMPSE, an AR-based feedback system for prosthesis users.
• To assess the impact of AR feedback on task performance and user experience.
• To explore the feasibility of a wearable, self-contained feedback solution.

Main Methods:

• Developed GLIMPSE, a Google Glass application providing prosthesis state information (EMG, force, etc.) via AR and bone conduction audio.
• Tested the system with healthy subjects performing a modified clothespin task under conditions with and without feedback (FB vs. NFB).
• Measured task completion time, unsuccessful trials, and subjective feedback relevance.

Main Results:

• No performance difference between FB and NFB groups on simple tasks.
• AR feedback significantly improved performance on complex tasks requiring fine motor control.
• GLIMPSE feedback did not increase task completion time.
• Subjects developed unique strategies integrating AR feedback with intrinsic sensory information.

Conclusions:

• The GLIMPSE system is a feasible and functionally beneficial wearable feedback interface.
• AR feedback shows promise for enhancing prosthesis control, particularly in demanding tasks.
• GLIMPSE serves as a practical solution and a versatile tool for research in closed-loop prosthesis control.