A Skin-Tone Insensitive, Fully Integrated Smartwatch for Continuous Blood Pressure Measurement

Smartwatches commonly employ photoplethysmography (PPG) sensors for pulse measurement and health index computation, including heart rate (HR) and blood pressure (BP). However, PPG’s susceptibility to skin tone variations presents challenges to accuracy. This study presents a biomedical eddy current sensor (BECS)-based wristwatch that integrates a passive LC tank, inductance-to-digital converter ASIC, and wireless module fully embedded into the case back of the wristwatch. The BECS technology relies on magnetic coupling between the magnetic fields generated by the LC coil and the counteracting magnetic fields induced by the wrist artery, enabling the measurement of resonant frequency variations in response to arterial pulse signals. To enhance pulse measurement sensitivity, the optimal operating resonant frequency was determined by implementing a capacitor array and measuring pulse signals by human measurement to derive personalized design parameters. The proposed BECS-based wristwatch demonstrates pulse signal measurement insensitivity to skin tone compared to PPG-based modalities, as validated among subjects with different Fitzpatrick skin types. Importantly, the personalized BP algorithm was implemented based on HR and modified normalized pulse volume (mNPV), both derived from the measured pulse signals. The BP performance demonstrates the low mean absolute errors (MAEs) of $3.18~\pm ~2.61$ mmHg for systolic BP (SBP) and $3.64~\pm ~2.57$ mmHg for diastolic BP (DBP) among ten subjects, meeting the Association for the Advancement of Medical Instrumentation (AAMI) standard below $5~\pm ~8$ mmHg. The generalization of each personalized BP model was further verified through train-test split validation, with all results remaining within AAMI criteria. In conclusion, this study introduces a novel BECS-based wristwatch, highlighting its skin-tone friendliness and accurate BP monitoring, offering a promising alternative to optical modalities, thereby driving future smartwatch innovations.