Assessment of human instantaneous arterial diameter using B-mode ultrasound imaging and artificial neural networks: Determination of wall mechanical properties

摘要

Wall artery viscoelastic properties (WAVP) are correlated with structural and functional state of the arterial system. An accurate estimation of these properties is achieved measuring wall instantaneous diameter and pressure signals. The aim of this work was to evaluate a new non invasive estimation method of the instantaneous arterial diameter (D), and consequently, WAVP. Ten common carotid arteries of hypertensive men were evaluated. D was calculated by using B-mode ultrasonic imaging and specialized software designed with Artificial Neural Networks. Instantaneous arterial pressure of all subjects was measured by piezoelectric tonometry. Arterial wall properties were evaluated using a linear autoregressive with exogenous input model. The new method, which determinates the arterial diameter, was compared respect to a specialized and previously validated method. Results showed no significant differences in all parameters derived of D (Bland & Altman test) and no differences in all the wall arterial mechanic indexes (p>0.05). For these reasons, the developed software based on Artificial Neural Networks was successful in determining the parameters associated with arterial diameters and it opens up the possibility of real time calculations of arterial wall mechanical properties because of its simplicity.