Electrode-Free ECG Monitoring with Multimodal Wireless Mechano-Acoustic Sensors.

Area of Science:

• Biomedical Engineering
• Cardiovascular Monitoring
• Wearable Technology

Background:

• Continuous cardiovascular monitoring is crucial for early cardiac event detection.
• Conventional electrocardiogram (ECG) systems with electrodes cause skin irritation and limit long-term wearability.
• There is a need for comfortable, non-invasive, long-term cardiac monitoring solutions.

Purpose of the Study:

• To develop an electrode-free system for continuous cardiovascular monitoring.
• To reconstruct ECG waveforms using synchronized phonocardiogram (PCG) and seismocardiogram (SCG) signals.
• To enable robust, daily-life cardiac monitoring compatible with the Internet of Things (IoT).

Main Methods:

• Utilized wireless mechano-acoustic sensors to capture synchronized PCG and SCG signals.
• Developed a deep learning model with specialized neural network components to fuse multimodal sensor data.
• Leveraged PCG for valvular event timing and SCG for mechanical context to identify cardiac intervals and patterns.

Main Results:

• Achieved a mean Pearson correlation of 0.96 between reconstructed ECGs and clinical ECGs.
• Obtained a Root Mean Square Error (RMSE) of 0.49 mV, indicating high accuracy.
• Demonstrated the system's capability for robust identification of systolic/diastolic intervals and pathological patterns.

Conclusions:

• The proposed electrode-free PCG-SCG fusion system enables accurate, long-term cardiovascular monitoring.
• Eliminating skin-contact electrodes enhances patient comfort and suitability for daily-life applications.
• The system is compatible with IoT infrastructure, paving the way for widespread adoption in remote cardiac care.