From sensing to clinical translation: advances and challenges in wearable cuffless blood pressure monitoring

Abstract

This paper aims to systematically synthesize state-of-the-art advancements in wearable cuffless blood pressure (BP) monitoring technologies, critically examining the sensing hardware, estimation algorithms and calibration strategies and identifying the core barriers that impede the translation from laboratory prototypes to clinically reliable cardiovascular monitoring systems. This paper adopts a comprehensive literature review approach, systematically analyzing emerging sensing modalities – including optoelectronic (PPG), mechanoelectric (triboelectric, piezoelectric, piezoresistive), ultrasonic and multimodal fusion systems – alongside the evolutionary trajectory of BP estimation techniques, from mechanism-driven models (PTT/PWV) through data-driven deep learning to the emerging physics-informed artificial intelligence paradigm. Clinical validation frameworks and calibration strategies are also critically assessed. The review reveals that while significant progress has been made in sensor sensitivity, signal processing and algorithmic accuracy, several critical barriers persist before widespread clinical adoption. These include motion-induced signal degradation, limited cross-population generalizability of algorithms, the absence of standardized validation protocols for cuffless devices and calibration drift over time. Physics-informed AI emerges as a promising paradigm for improving robustness, interpretability and generalization by embedding hemodynamic constraints into data-driven models. As a review paper, the findings are based on published literature rather than original experimental data. Future research should focus on adaptive calibration frameworks, rigorous multi-scenario validation aligned with ESH 2023 recommendations and hardware-software co-design approaches. The review provides actionable guidance for researchers, clinicians, and device developers by identifying key technological bottlenecks and outlining future directions, including endogenous autonomous calibration benchmarks and multimodal sensing integration, to accelerate the development of clinically viable wearable BP monitoring systems. This paper uniquely bridges the gap between sensing hardware and clinical requirements by providing an integrated analysis across the full technology chain – from sensors to algorithms to clinical translation. It highlights physics-informed AI as a critical emerging direction and proposes a forward-looking framework for all-scenario adaptive dynamic calibration, offering a comprehensive roadmap for the field.