Motivation

Population ageing and prevalence of cardiovascular diseases (CVDs) have imposed enormous economic and societal burden worldwide. Emerging Low-cost and mobile health-monitoring devices has demonstrated significant promise in the prevention, prompt management, and treatment of CVDs. However, the functionality of wearable devices for interpreting and translating valid CVD diagnostic information remains limited due to the longitudinal degradation with ageing, complexity, and individualized characteristics of the cardiovascular systems.

The recent advances in hardware miniaturization and signal processing have enabled a variety of commercial wearable devices (e.g., Apple Watch, Oura-ring, etc.) as well as flexible epidermal sensors that support daily monitoring of basic vital signs, such as heart rate, blood oxygen level, and heart sound. These parameters, although being important health indicators, are insufficient for the assessment of vascular ageing and the progression of CVD. In fact, the signals captured by the wearable and flexible sensors, e.g., electrocardiogram (ECG), epidermal pulse waveform and photoplethysmography (PPG), already contain rich cardiac information; it is the lack of physiological models that prevents these devices from being a powerful diagnostic tool for CVD.

Highlights of our research

In this work, we introduce a stroke-volume-allocation (SVA) model that quantifies the arterial cushioning function, allowing wearable and flexible sensors to simultaneously assess arterial stiffness (AS), track blood pressure (BP) dynamics, and evaluate cardiovascular disease-related heart damage (CHD). Notably, the arterial cushioning function is compromised due to the thickening and stiffening of arterial walls, which is also associated with elevated BP and an increased risk of CVD. Therefore, to effectively measure and analyse this crucial function, we developed the SVA model, which calculates the ratio of stroke volume in arterial compliance during systole to the total stroke volume pumped by the heart in the same cardiac cycle.

By analysing the mechanisms of SVA in the circulation, we have successfully established the intrinsic physiological connections between SVA and BP, SVA and AS-cfPWV (a measure of arterial stiffness), and SVA and CHD. These relationships were further validated in a large hybrid database. Besides, to validate the feasibility of extracting SVA from wearable sensor signals, we conducted synchronized monitoring using ECG, PPG, and epidermal pulse sensors, accurately tracking dynamic SVA induced by deep breathing. Finally, our findings demonstrated that commercial wearable devices such as the Apple Watch and a PPG ring-based prototype of the INTS can effectively assess vascular ageing, track ambulatory BP, and provide cardiac risk diagnoses using the SVA model. These advanced wearable sensing methods empower users to detect CVD at an early stage and receive timely alerts, benefiting the biomedical community.

Looking ahead

Our work presents a promising way that can be widely deployed in particular the recently emerged novel flexible sensors for CVD assessment at a very low cost and with a short launch time. Considering CVD is the leading cause of death globally and closely related to population ageing, we believe the work can pave the way toward the acceleration of digital medicalization by showcasing efficient tracking of vascular dysfunction and giving examples of how next-generation advanced wearable medical systems can be achieved for CVD monitoring.