A Low-Power Wearable Stand-Alone Tongue Drive System for People With Severe Disabilities

Area of Science:

• Biomedical Engineering
• Assistive Technology
• Wearable Computing

Background:

• Individuals with severe disabilities often face challenges in controlling external devices.
• Existing assistive technologies may have limitations in power consumption and functionality.
• Tongue movement-based control offers a potential alternative for environmental interaction.

Purpose of the Study:

• To develop a low-power, stand-alone tongue drive system (sTDS) for environmental control.
• To enable users with severe disabilities to operate computers, smartphones, and wheelchairs.
• To reduce power consumption and extend battery life through on-board signal processing.

Main Methods:

• A low-power local processor integrated into the sTDS headset for real-time signal processing.
• Signal processing blocks include serial peripheral interface, magnetic interference attenuation, and machine learning classification.
• Implementation using a low-power FPGA and subsequent design in 65-nm CMOS technology.

Main Results:

• The sTDS significantly reduces wirelessly transmitted data by a factor of 64 (9.6 to 0.15 kb/s).
• The FPGA-based processor achieved 96.9% command detection accuracy, consuming 27 nJ per command.
• The integrated circuit processor demonstrated a 10x lower power consumption (0.43 mW) than the FPGA.

Conclusions:

• The proposed low-power sTDS effectively translates voluntary tongue movements into environmental control commands.
• On-board processing dramatically reduces power requirements, enhancing usability and battery life.
• The developed system offers a promising assistive technology for individuals with severe physical impairments.