Abstract

Continuous and non-invasive blood pressure (BP) monitoring has the potential to greatly improve hypertension diagnosis and management, along with enabling valuable personal health monitoring in the population at large. Existing methods rely on cuff-based sphygmomanometers, which are cumbersome and disrupt sleep. Despite extensive research utilizing photoplethysmography (PPG), which is integrated into many consumer wearables, errors in continuous BP estimation are generally considered unacceptable for general use. We developed a high-speed (390 Hz) Speckle Contrast Optical Spectroscopy (SCOS) system to measure the cardiac blood flow waveform simultaneously with the PPG signal at high temporal resolution. The system utilized high speed multiplexed detection of optical speckle patterns on the wrist and finger, enabling the extraction of novel features related to BP. In comparison to PPG alone, SCOS demonstrated a notable 31% improvement (p = 3.45 * 10-7) in systolic BP estimation when integrated into subject-specific machine-learning models. The resulting errors were remarkably low (systolic BP: 0.06+/- 2.88 mmHg, diastolic BP: 0.09 +/-2.14 mmHg) across a wide range of BP variations (range SBP: 89-284 mmHg). Importantly, this improvement was sustained several weeks later within a re-measured cohort, indicating highly robust BP predictions. Looking ahead, the use of SCOS for blood flow measurements holds the potential to substantially enhance BP estimations compared to conventional PPG-based methods.