Lumen Segmentation and Motion Estimation in B-Mode and Contrast-Enhanced Ultrasound Images of the Carotid Artery in Patients With Atherosclerotic Plaque

摘要: In standard B-mode ultrasound (BMUS), segmentation of the lumen atherosclerotic carotid arteries and studying geometry over time are difficult owing to irregular shapes, noise, artifacts, echolucent plaques. Contrast enhanced (CEUS) improves visualization, but remains challenging varying intensities, CEUS-specific artifacts lack tissue visualization. To overcome these challenges, we propose a novel method using simultaneously acquired BMUS&CEUS image sequences. Initially, estimates nonrigid motion (NME) from sequences, intensity-based registration. The motion-compensated sequence is then averaged obtain single “epitome” with improved signal-to-noise ratio. segmented epitome through an intensity joint-histogram classification graph-based segmentation. NME was validated by comparing displacements manual annotations in 11 carotids. average root mean square error (RMSE) $112\pm 73~\mu {\rm m}$ . Segmentation results were against delineations images two different datasets, respectively containing (RMSE $191\pm 43~\mu ) 10 $351\pm 176~\mu From deformation fields, derived arterial distensibility values comparable literature. errors all experiments inter-observer variability range. best our knowledge, this first study exploiting combined for