ECG Signal Construction From Heart Sounds via Single Node, Surface Acoustic Sensing

Area of Science:

• Biomedical Engineering
• Cardiovascular Technology
• Signal Processing

Background:

• Current electrocardiogram (ECG) monitoring is limited by intrusive hardware and stationary requirements.
• Existing phonocardiogram methods lack precision for crucial cardiac rhythm analysis and are susceptible to noise.
• There is a need for non-intrusive, mobile, and continuous cardiac rhythm monitoring solutions.

Purpose of the Study:

• To develop a method for generating ECG waveforms from acoustic heart sounds.
• To enable unobtrusive, long-term, low-cost daily cardiac rhythm monitoring.
• To validate the accuracy of acoustic ECG in capturing key cardiac metrics.

Main Methods:

• Utilized a wide bandwidth surface-acoustic-wave microphone placed on the neck to capture heart sounds via the carotid artery.
• Employed a cross-modal autoencoder, an advanced algorithm, for transforming acoustic heart sound signals into ECG waveforms.
• Conducted a 9-participant study to evaluate the generated ECG waveforms and metrics.

Main Results:

• Successfully constructed PQRST waveforms from acoustic heart sound signals.
• Demonstrated accurate determination of critical PQRST metrics from the acoustic-derived ECG.
• Showcased mobile acoustic ECG waveform construction during user ambulation (walking).

Conclusions:

• Acoustic heart sound to ECG transformation is feasible and effective.
• This technology paves the way for unobtrusive, long-term, low-cost daily cardiac rhythm monitoring.
• A single-node dry wearable device can provide valuable cardiac rhythm insights.