A Wireless Fully-Passive Acquisition of Biopotentials.

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Sensor Technology

Background:

• Biopotential signals are crucial for monitoring organ function and diagnosing diseases.
• Current methods often rely on wired connections, limiting patient mobility and comfort.
• There is a need for unobtrusive, real-time biopotential monitoring solutions.

Purpose of the Study:

• To develop and validate a flexible, wireless, and fully-passive sensor for real-time biopotential signal acquisition.
• To assess the performance of the wireless sensor against traditional wired systems using deep learning analysis.
• To demonstrate the feasibility of the technology for clinical applications.

Main Methods:

• Fabrication of a flexible sensor on a 90 μm-thick polyimide substrate.
• Utilizing RF microwave backscattering with varactors for wireless signal transmission.
• Validation using emulated signals, electrocardiogram (ECG), electromyogram (EMG), and electrooculogram (EOG).
• Employing a deep learning algorithm to analyze signal quality and compare wireless vs. wired data.

Main Results:

• The wireless sensor achieved <3% discrepancy in deep learning accuracy for ECG and EMG compared to wired sensors up to 240 mm.
• Accurate tracking of horizontal eye movement (EOG) with high deep learning accuracy (93.6% training, 92.2% testing).
• Successful detection of biopotential signals as low as 250 μVpp.
• Demonstrated real-time, wireless, and fully-passive biopotential acquisition.

Conclusions:

• The developed flexible sensor enables feasible real-time, wireless, and fully-passive on-body biopotential acquisition.
• The technology shows high accuracy and minimal discrepancy compared to wired sensors.
• Potential for diverse applications in future clinical research and wearable health monitoring.