A Closed-Loop Ear-Worn Wearable EEG System with Real-Time Passive Electrode Skin Impedance Measurement for Early...

Area of Science:

• Biomedical Engineering
• Neuroscience
• Wearable Technology

Background:

• Autism spectrum disorder (ASD) severity correlates with diagnosis age, necessitating early detection.
• Conventional EEG systems present challenges like bulkiness, wired electrodes, and high power consumption.
• Real-time monitoring of electrode-skin interface (ESI) impedance is often lacking in current EEG solutions.

Purpose of the Study:

• To develop a novel ear-worn wearable electroencephalogram (EEG) system for the early detection of ASD.
• To integrate continuous, long-term EEG recording and on-chip machine learning for real-time ASD prediction.
• To implement a passive ESI evaluation system to overcome limitations of conventional EEG.

Main Methods:

• Designed an ear-worn wearable EEG system with continuous, long-term recording capabilities.
• Incorporated an on-chip machine learning processor (180 nm CMOS) for real-time ASD prediction.
• Developed a passive ESI evaluation methodology using root mean square voltage to assess impedance.

Main Results:

• The on-chip ML processor has a minimal active area (2.52 mm²) and low energy consumption (0.87 µJ/classification).
• The passive ESI system effectively evaluates impedance by considering multiple physical factors.
• The system's performance was validated using the Old Dominion University ASD dataset.

Conclusions:

• The developed ear-worn EEG system offers a promising solution for early ASD detection.
• On-chip machine learning and passive ESI monitoring enhance the feasibility of long-term wearable EEG applications.
• This technology has the potential to reduce ASD severity through earlier diagnosis and intervention.