Precise Opening Detection of Heart Aortic Valve Towards Noninvasive Continuous Blood Pressure Detection

This paper presents a novel earbud-based system for continuous, noninvasive blood pressure (BP) monitoring using precise arterial pulse transit time (PTT) estimation. The system integrates a low-noise inertial measurement unit (IMU) and a photoplethysmography (PPG) sensor, achieving highly synchronized data acquisition at 450 Hz with <1.1 millisecond synchronization error between the sensors. Key contributions include: (i) a rigorous validation of ballistocardiogram (BCG)-based proximal timing (B-point) estimation against ground truth B-point derived from synchronized electrocardiography (ECG) and impedance cardiography (ICG); (ii) development of a custom earbud prototype optimized for BP monitoring with high-precision arterial PTT measurement; and (iii) introduction of a B-point estimation algorithm leveraging multi-axis BCG fusion to enhance signal-to-noise ratio (SNR) and improve accuracy. Results show that BCG fusion reduces mean absolute error (MAE) in B-point estimation by 5.33% and 6.70% compared to using only the X- or Y-axis BCG signals, respectively. Furthermore, using the I-wave of the BCG signal as a proxy for the opening of the aortic valve reduces the MAE by 59.45% and 32.09% compared to using the H-wave and the J-wave, respectively. These findings establish a foundation for practical, wearable, and noninvasive continuous BP monitoring, with potential applications in real-world cardiovascular health management.