Adaptive Heartbeat Modeling for Beat-to-Beat Heart Rate Measurement in Ballistocardiograms

Area of Science:

• Biomedical Engineering
• Physiological Monitoring
• Signal Processing

Background:

• Ballistocardiography (BCG) offers a non-invasive method for physiological monitoring.
• Accurate beat-to-beat heart rate measurement is crucial for sleep analysis and health assessment.
• Existing BCG methods may require complex setups or lack robustness in home environments.

Purpose of the Study:

• To develop and validate a novel method for calculating beat-to-beat heart rate from BCG signals.
• To enable unobtrusive, long-term heart rate monitoring within a home-use e-health system.
• To assess the accuracy and detection rate of the proposed method across various recording conditions.

Main Methods:

• Adaptive inference of heartbeat shape using hierarchical clustering from BCG signals.
• Detection of beat-to-beat intervals by matching the inferred heartbeat model to the signal.
• Validation using overnight recordings from 46 subjects in diverse settings (sleep clinic, home, different bed configurations, multiple sensor types).

Main Results:

• The developed method achieved a mean beat-to-beat interval error of 13 ms.
• An average of 54% of beat-to-beat intervals were successfully detected across recordings.
• The method demonstrated robustness in various real-world sleep measurement scenarios.

Conclusions:

• The proposed method provides a viable approach for accurate beat-to-beat heart rate measurement from BCG.
• This technique supports the development of unobtrusive e-health systems for sleep and health monitoring.
• Further research can optimize detection rates and expand applications in remote patient monitoring.