Project 03 · 4 publications

Non-Contact Biopotential Sensing

Capacitive, non-contact electrode systems for acquiring ECG, EMG, and EEG through clothing and without skin contact or conductive gel — enabling truly unobtrusive long-term health monitoring and in-vehicle driver safety applications.

Non-Contact ECGCapacitive ElectrodesWearable HealthMotion ArtifactSmart Health
Overview

Biopotential Monitoring Without Skin Contact

Wet Ag/AgCl electrodes — the clinical standard for ECG, EEG, and EMG — require conductive gel, direct skin contact, and skilled placement. Non-contact capacitive electrodes go further: they couple to the body through one or more dielectric layers (clothing, air gaps, hair), making them compatible with fully unobtrusive embedding in garments, car seats, or small button-sized wearable modules.

This project developed the hardware, analog front-end circuit design, and motion-artifact mitigation strategies for practical non-contact biopotential systems — advancing them from bench demonstrations to wearable and automotive deployment scenarios.

Key challenge — motion artifacts: The small coupling capacitance of non-contact electrodes is highly sensitive to relative motion between the electrode and skin. Critically, the same triboelectric effect that powers harvesters corrupts biopotential signals. This project studied that crossover — characterizing, modeling, and mitigating triboelectric artifacts in non-contact wearable devices through analog feedback design and signal processing.

19Fabric layers (ECG through)
24 gNCMB-Button weight
ECG EMG EEGAll three signals acquired
Acquired Signals

Multi-Modal Biopotential Coverage

The non-contact sensing framework was validated across three distinct biopotential modalities, each with different amplitude, frequency, and clinical relevance:

ECGElectrocardiography

Cardiac electrical activity for heart rate, HRV, and arrhythmia detection. Detected through up to 19 fabric layers (polyester) and in-vehicle through a car seat — without skin contact.

EEGElectroencephalography

Brain electrical activity including alpha waves (8–12 Hz) visible during drowsy/relaxed states. Key indicator for driver alertness monitoring; detected through hair without conductive gel.

EMGElectromyography

Muscle electrical activity for motion characterization and rehabilitation monitoring, acquired without traditional wet gel — enabling both upper limb and eye-blink EMG detection.

Technical Approach

From In-Vehicle Monitoring to Wearable Integration

In-vehicle driver monitoring system.  The foundational system demonstrated fully non-contact, nonintrusive ECG, EEG, and eye-blink detection at a distance — requiring no physical contact with the driver's skin. Validated on a high-fidelity driving simulator, the system extracted HRV and alpha-wave drowsiness indicators in real time, establishing the core capacitive front-end architecture.

Motion artifact investigation.  A systematic study characterized how triboelectric charge accumulation at the electrode-skin interface degrades signal quality in non-contact wearable devices. Theoretical modeling and experimental validation provided the design guidelines — ultra-high input impedance front-end, analog DRL-style feedback loop — that inform the NCMB-Button system design.

Design and evaluation of non-contact wireless system.  Extends the driver monitoring architecture to a wearable wireless form factor. Detects biopotential through fabric as thick as three-layer jeans, with performance comparable to wet electrodes. Evaluates motion artifacts across different textile materials and motion types — providing design guidelines for wearable deployment.

NCMB-Button: multi-modal wearable system.  A compact (39 × 32 × 17 mm, 24 g) 3D-printed wearable module with ultra-high input impedance analog front-end, dual power supply management, and 150 mAh Li-ion battery. The feedback loop drives a driven-right-leg (DRL) style signal injection into the body to suppress common-mode noise and motion artifacts. Validated through 19 polyester layers — well beyond dry-contact electrode capability.

Publications
Smart Health · Vol. 11 · January 2019
A Wearable Button-Like System for Long-Term Multiple Biopotential Monitoring Using Non-Contact Electrodes
Xian Li, Ye Sun
Full journal paper describing the NCMB-Button: a button-form-factor wearable wireless non-contact biopotential system (39×32×17 mm, 24 g). ECG detected through 2.60 mm fabric with rapid recovery from motion artifacts via the analog feedback architecture. EMG signals detected from upper limb and eye-blinking. EEG alpha waves visible during drowsy conditions. Selected as one of six papers for journal publication from IEEE/ACM CHASE 2017.
IEEE/ACM CHASE · August 2017
NCMB-Button: A Wearable Non-Contact System for Long-Term Multiple Biopotential Monitoring
Xian Li, Ye Sun
Conference paper introducing the NCMB-Button. Demonstrates ECG and EMG detection through up to 19 polyester layers, with fast recovery from motion artifacts. EEG alpha waves detected during drowsiness. Packaged into a 39×32×17 mm 3D-printed box weighing 24 g, powered by a 150 mAh Li-ion battery with dual power-supply management circuitry. Selected for journal publication in Elsevier Smart Health.
IEEE BHI · February 2017
Design and Evaluation of a Non-Contact Wireless Biopotential Monitoring System with Motion Artifacts
Xian Li, Ye Sun
Reports the design and validation of a wireless non-contact biopotential monitoring system capable of detecting signals through fabric as thick as three-layer jeans, with performance comparable to wet electrodes. Evaluates motion artifacts of non-contact electrodes under different textile materials and motion types, providing key design guidelines for wearable non-contact biopotential detection.
IEEE EMBC · August 2016
Investigation of Motion Artifacts for Biopotential Measurement in Wearable Devices
Xian Li, Hui Huang, Ye Sun
Theoretical and experimental study of motion artifacts in non-contact biopotential measurement. Models triboelectric charge accumulation at the electrode-clothing-skin interface and its effect on ECG/EEG signal fidelity. Identifies design strategies — including ultra-high input impedance analog front-ends and feedback suppression circuits — for artifact mitigation in dry and non-contact electrode configurations.