Automated estimation of blood pressure using PPG recordings: an updated review

With the increasing incidence of hypertension, automated cuffless estimation of blood pressure (BP) is a high clinical need in an aging world. Recent years have witnessed a rapid growth of cuffless BP estimation techniques. Photoplethysmography (PPG) technology has been widely used in wearable sensors where artificial intelligence (AI) provides new potential for automated cuffless BP estimation. PPG signals reflect the volumetric changes in microcirculation, which are essentially associated with BP. Machine learning and deep learning algorithms enable PPG-based BP estimation to achieve high accuracy towards real-world application, with a gap in large-scale validation. This chapter starts with a brief introduction on cuffless BP estimation and PPG technology. The mainstream methods of PPG-based automated BP estimation are summarized with an analysis of the underlying physiological mechanisms. The advancement and limitations of existing algorithms are summarized. Further improvements in data, hardware, algorithms, and clinical validation are discussed as future directions of AI-enhanced PPG-based BP estimation. This chapter provides an overview of the state-of-the-art and serves as a reference for biomedical engineers and healthcare professionals.