Real time US-tagging combined with phase-based optical flow applied to 2D motion estimation of the carotid artery wall

Early detection of cardiovascular diseases (CVD) is possible by assessing the dynamic properties of the common carotid artery, i.e. the 2D (radial and longitudinal) motion of the wall layers. B-mode imaging is widely used, however this modality does not provide a high definition in the longitudinal direction (i.e. parallel to the probe) due to the aperture problem and the homogeneity of the tissues. To increase the definition in the longitudinal direction, we present here a specific beamforming method dedicated to estimate the 2D motion of the carotid artery wall. The generated US-tagged images present radiofrequency oscillations in both directions. The 2D motion is assessed with a specific phase-based optical flow method. The method was implemented in real time on a research echograph and applied in vivo on a healthy carotid over 3 cardiac cycles. The results show that our method is more accurate compared to classical block matching on B-mode images.