An Electro-Oculogram Based Vision System for Grasp Assistive Devices-A Proof of Concept Study

Area of Science:

• Biomedical Engineering
• Neuroscience
• Rehabilitation Technology

Background:

• Amyotrophic Lateral Sclerosis (ALS) patients lose voluntary muscle control, impacting limb movement for object interaction.
• Existing grasp assistive devices often require muscle activity, limiting their use for individuals with severe motor impairments.
• Fine control over object orientation and width is crucial for replicating human-like grasps with assistive devices.

Purpose of the Study:

• To develop a vision system controllable via human gaze for grasp assistive devices.
• To enable individuals with ALS to control assistive devices using only eye movements.
• To enhance object manipulation capabilities for brain-controlled assistive technologies.

Main Methods:

• Utilized electrooculogram (EOG) signals to control a cap-mounted webcam's pan and tilt for object tracking.
• Implemented a signature extraction procedure to simplify algorithms and reduce system storage.
• Tested the system with ten healthy participants to evaluate performance in object parameter estimation.

Main Results:

• The gaze-controlled vision system accurately estimated object orientation and size within 22 ms.
• Achieved a combined accuracy exceeding 75% for determining grasp parameters.
• Demonstrated the feasibility of using eye movements to control a vision system for grasp assistance.

Conclusions:

• The developed gaze-controlled vision system offers a viable solution for ALS patients using brain-controlled assistive devices.
• Integration with grasp assistive devices can provide more natural object maneuvering capabilities.
• Future work includes expanding grasp types and further refining the system for clinical application.