Di Zhao, Zhichao Yang, Zhiheng Mo, Di Wu, Huijuan Li, Jianjun Yang, Longhui Zeng, Xiaonan Yang, Yuen Hong Tsang, Long He
The surveillance of blood pressure and glucose levels, which are pivotal physiological metrics, is imperative for the proactive management and prevention of chronic conditions, particularly cardiovascular diseases and diabetes. Conventional methodologies for their assessment frequently present limitations, including invasiveness, discomfort to the patient, and an inability to facilitate continuous monitoring. The advent of optoelectronic detection technologies has provided innovative approaches that effectively address these constraints. In this work, a WSe 2 /Si heterojunction photoplethysmography sensor was fabricated. The device demonstrates superior performance metrics, including a high detectivity of 3.65 × 10 13 Jones, a high responsivity of 494 mA/W, and a fast response time of 2.8/4.5 μs in the near-infrared region. The high performance of the detector enables the highly sensitive and noninvasive measurement of blood glucose with a sensitivity of 6.94 μA/mmol·L –1 . Machine learning algorithms were developed to further improve the precision of blood glucose measurement, resulting in a root-mean-square error of 0.52 mmol/L. Moreover, the accurate dynamic monitoring of respiratory rate, heart rate, and blood pressure was achieved by the device, highlighting its multifunctionality in health monitoring applications. This study demonstrates a feasible approach to designing novel, high-precision PPG sensors for real-time continuous noninvasive health monitoring.