Blood Pressure Estimation Based on the Timing of the Arterial Reflected Waves by Oscillometric Pulse Wave Decomposition

Recent innovations in oscillometric blood pressure (BP) estimation leveraging pulse transit time (PTT) show promise even when pulse amplitude is compromised. However, current PTT methods often require multiple sensors. This study introduces a novel approach for oscillometric BP estimation utilizing only a single sensor. Our method simplifies PTT measurement through decomposing oscillometric pulses into forward and reflected wave components. Notably, we found that the time interval between the forward and first reflected wave is shortest when cuff pressure is at the systolic level, while the interval between the first and second reflected waves is longest when the cuff pressure corresponds to the diastolic level. To validate our method, we extensively analyzed two datasets: the first comprised 150 arm oscillometric recordings from 10 healthy individuals, and the second included 425 wrist oscillometric recordings from 85 healthy individuals. Our analysis revealed minimal standard deviation errors for BP measurements: 7.03 mmHg (systolic) and 5.96 mmHg (diastolic) for arm recordings, and 7.98 mmHg (systolic) and 5.39 mmHg (diastolic) for wrist recordings. Across both datasets, the mean error for both systolic and diastolic BP measurements was merely 0.01 mmHg, underscoring the precision of the method. Comparative analysis against established methodologies further affirmed the superior performance of our approach, which also met the ANSI/AAMI/ISO81060-2 standard. While our dataset primarily comprises recordings from healthy individuals, further validation on diverse datasets encompassing challenging cases such as obesity, arterial stiffness, and atrial fibrillation is imperative to evaluate its applicability across various clinical scenarios.