Building intelligent communication systems for handicapped aphasiacs.

Area of Science:

• Biomedical Engineering
• Assistive Technology
• Human-Computer Interaction

Background:

• Aphasia significantly impairs communication abilities in affected individuals.
• Existing communication aids may be costly or lack natural interaction.
• There is a need for accessible and intuitive communication tools for handicapped aphasiacs.

Purpose of the Study:

• To develop an intelligent system for basic communication tasks for handicapped aphasiacs.
• To integrate finger gesture recognition with a virtual keyboard for seamless interaction.
• To provide an affordable and natural communication solution.

Main Methods:

• Utilizing a 6-sensor data glove to capture finger bending degrees.
• Employing a finger language recognition subsystem with multiple regression analysis and radial basis function neural networks.
• Implementing a coordinate-indexed virtual keyboard for direct letter access.

Main Results:

• The system accurately recognizes finger gestures as language components.
• The radial basis function neural network enables fast and correct model construction.
• Users can access letters on the virtual keyboard at a practical speed.

Conclusions:

• The intelligent system offers a viable communication tool for handicapped aphasiacs.
• Continuous finger language input enables natural and affordable communication.
• This technology enhances the quality of life for individuals with aphasia.