Biomechanical mechanism of noninvasive plaque detection based on multi-sensor fusion

Plaque can induce stroke and other serious cardiovascular and cerebrovascular diseases. How to diagnose plaque early has important clinical significance. Recently, the conventional detection methods are mainly medical imaging methods such as computed tomography, and there is still a lack of portable nonimaging detection technology or devices that can be used at home. To achieve such plaque detection techniques, possible solutions are based on pulse wave sensors and blood flow sensors to extract plaque signature signals. Elucidating the relationship between the sensing signals of these two types of sensors and changes in hemodynamic parameters caused by plaque is the basis of developing accurate wearable continuous monitoring systems for plaque. In this study, based on the flow-solid interaction effect between the vessel wall and the blood flow, the stenotic vessels induced by plaques were modeled by numerical simulation software, and the distribution patterns of vessel deformation and blood flow velocity near plaques during the cardiac cycle were investigated in detail. By measuring and processing these two simultaneous dynamic signals, a preliminary method of estimating plaque size based on displacement, velocity, and their first-order derivative curves is developed, and the errors are all less than 9.5%. Meanwhile, to explore the relationship between the detected signals from multiple arterial sites and plaques, we investigated the response of carotid, brachial, and radial artery signals to different sizes of plaques using the block parameter model of vascular network, which provides a theoretical basis for the construction of a multi-sensor fusion for noninvasive plaque detection.