Ultrasound-image-based displacement and strain maps of the human carotid artery using weighted-least-squares optical flow

Abstract

The mechanical displacements of the carotid artery wall during the cardiac cycle can be estimated using appropriate motion analysis methods. The waveforms resulting from such estimations can subsequently be used to quantify the elastic properties of arterial tissue. In this study weighted-least-squares optical flow (WLSOF) was used to produce radial and longitudinal displacement maps of extended segments of healthy and diseased arterial walls from B-mode ultrasound. Strain maps of the same segments were produced by compensating for the translational movement of the carotid artery walls and subsequently assessing the strain tensor by solving a nonlinear minimization problem with a modified version of the WLSOF algorithm. Displacement maps accurately estimated the motion of the healthy carotid artery and also revealed information about the motion of the soft tissues of the carotid artery which would otherwise be difficult to distinguish. Strain maps in the case of healthy carotid artery showed good reproducibility between different cardiac cycles, while in the case of atherosclerotic subjects additional validation is needed.