Eye-Movement-Controlled Wheelchair Based on Flexible Hydrogel Biosensor and WT-SVM

Area of Science:

• Biomedical Engineering
• Wearable Sensors
• Assistive Technology

Background:

• Traditional rigid electrodes for biosensing have limitations in deformability and biocompatibility.
• Patients with restricted mobility require advanced assistive devices for independent navigation.
• Electrooculography (EOG) and strain signals offer potential for non-invasive control of assistive devices.

Purpose of the Study:

• To develop an eye-movement-controlled wheelchair prototype for patients with restricted mobility.
• To create a flexible, biocompatible hydrogel biosensor for collecting electrooculogram (EOG) and strain signals.
• To achieve high accuracy in recognizing various eye movements for precise wheelchair control.

Main Methods:

• Fabrication of a flexible hydrogel biosensor using conductive HPC/PVA hydrogel and PDMS substrate.
• Affixing the biosensor to the forehead to collect EOG and strain signals.
• Application of the Wavelet Transform-Support Vector Machine (WT-SVM) algorithm for classifying eye movement signals based on amplitude, duration, and interval features.

Main Results:

• The flexible hydrogel biosensor demonstrated low Young's modulus (286 KPa) and high breathability (18 g m^-2 h^-1), ensuring comfortable and conformal adhesion.
• The WT-SVM algorithm achieved an average eye movement recognition accuracy of 96.3%.
• The prototype successfully enabled precise manipulation of the wheelchair through recognized eye movements.

Conclusions:

• The developed flexible hydrogel biosensor is suitable for unobtrusive and accurate EOG and strain signal collection.
• The WT-SVM algorithm effectively classifies eye movements, enabling precise control of the wheelchair prototype.
• This eye-movement-controlled wheelchair system offers a promising solution for enhancing the independence and mobility of patients with physical disabilities.