Heart rate detection method based on Ballistocardiogram signal of wearable device：Algorithm development and...

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Signal Processing

Background:

• Traditional electrocardiography (ECG) for heart rate monitoring is restrictive due to electrode requirements.
• Wearable devices offer non-invasive, convenient health monitoring, but Ballistocardiogram (BCG) signal acquisition faces robustness challenges.
• Accurate, untethered heart rate monitoring is crucial for continuous health assessment.

Purpose of the Study:

• To develop an accurate method for detecting heartbeat cycles using BCG signals, specifically for untethered monitoring during sleep.
• To improve the robustness and accuracy of non-invasive heart rate monitoring.

Main Methods:

• An innovative J-wave detection algorithm based on BCG signals was implemented.
• A bi-directional long short-term memory (BiLSTM) model was constructed for J-wave recognition after feature extraction.
• BCG signals from 28 healthy subjects in various sleeping positions were collected and analyzed.

Main Results:

• The J-wave recognition accuracy reached 99.67%, with a heart rate detection deviation rate of only 0.27%.
• The proposed method demonstrated higher accuracy compared to previous wearable device algorithms.
• Bland-Altman plots showed no significant difference between BCG and ECG heart rate results.

Conclusions:

• The developed method enhances the accuracy and generalization of BCG-based heartbeat cycle extraction.
• This approach provides a foundation for reliable, wearable-based, untethered daily health monitoring.
• The study validates the efficacy of the BiLSTM model for robust J-wave recognition in BCG signals.